...... 9 9 2
AVE XA04NO301 - 34 HSE Health &Safety Executive INIS-XA-N--030
S '4&SSeSSMMe1M'
E II N
I N T E R N A T I N A L C 0 N F E R E N C E
1 9 9 2 .00000 .... a 0 ...... 0 .... .0.0.00.0.
5 - 9 October 1992 Queen Elizabeth 11 Conference Centre, Westminster, London ...... 9 9 2
Page I- 7 Programme Session I What is risk assessment and what do we want from it?
I 0 Welcome Address Sir John Cullen 11-13 Opening Remarks Dr Bill Hunter 14 Brief Introductory Remarks Dr Kazutaka Kogi 15 Opening Remarks Dr Jim Brydon 16-17 Risk Assessment Dr Stanislaw Tarkowski 18-19 Opening Address Secretary of State for Employment 20-27 Overview of Risk Assessment John Rimington 28-31 Risk Assessment: A Consumer's Perspective Dame Rachel Waterhouse 32-35 An Employee/Trade Union Perspective to Risk Assessment Pekka 0 Aro 3644 Risk Assessment: An Employer's Perspective K C Williams 45-53 Risk Assessment - the Perspective and Experience of US Environmentalists Dr Ellen Silbergeld 54-63 The Transformation of the Industrial Economy into a Service Economy: the Function of Technology and the Role of Insurance in the Management of Risk Professor Orio Giarini
Session 2 The role of risk assessment in international policy
66-77 ILO Activities in the Field of Risk Assessment Dr Chandra Pinnagoda 78-87 The Role of Risk Assessment in the Work of the World Health Organization in Europe Dr Kees A van der Heijden 88-94 An OECD Perspective of the Role of Risk Assessment in Policy Development Dr Jim Brydon 95-99 Risk Assessment - A European Community Perspective Ron Haigh 100-112 Risk Assessment and the Environment Dr David Fisk 113-119 Occupational and Environmental Risk Assessment Problems in Central European Countries Needs and What has been Achieved - Poland as an Example Dr Tadeusz Sulkowski 120-130 Risk Assessment: A Regional Approach Milos Pale ek 131-138 Panel Discussion ...... 9 9 2
Session 3 Risk assessment in practice
140-152 General Approaches to the Risk Assessment of Chemicals Patrick Murphy 153-161 Risk Assessment of Chemicals - A Central European Perspective Professor Nadimir Bencko, Professor y6rgy Ungvdry 162-165 Pesticide Risk Assessment in the United States Dr Richard Hill 166-172 An Industry Approach to the Risk Assessment of Pesticides Dr Barry Thomas 173-182 Radiation Protection Standards - A Practical Exercise in Risk Assessment Dr Roger Clarke 183-188 Panel Discussion
Session 4 Risk assessment in practice
190-203 Food and Drinking Water Safety: Can Risk Assessment Help Us to Get Our Priorities Right? Dr H B Denner 204-209 Microbiological Risk Assessment and Public Health Dr Roger Skinner 210-220 Chemical Safety of Food and Drinking Water Dr Maged Younes - Dr Kees A van Der Heijden 221-226 Risk Assessment in the Context of EC Directives on Genetically Modified Organisms Dr Piet van der Meer 227-237 Risk Assessment for Federal Regulatory Decisions on Organisms Produced Through Biotechnology Terry L Medley - Dr John H Payne 238-244 Panel Discussion ...... 1 9 9 2 ...... 1 9 9 2
"NORg I I T IME ...... -12 sesslion I 4 00 ri Iii aix!
olIt pg-
"R MI
4IN III9 i 11
splonq iss;ol
52
S.- OR
05"
lol
pl g
","LlP E -IM1,111"n ...... 1 9 9 2
CONFERENCE PROGRAMME
Monday, October 1992
8.30 Registration and coffee
What is risk assessment and what do 00*0000900000000000 we want from it?
Chairman: Sir John Cullen Chairman, Health and Safety Commission, UK
9.30 Welcome address by Sir John Cullen. Brief introductory remarks by:
Dr Bill Hunter Director, Health and Safety Directorate, European Commission
Dr Kazutaka Kogi Director, Working Conditions and Environment Department, Internatio'nal Labour Office
Dr Jim Brydon Head, Environmental Health Safety Division, Organization for Economic Co-operation and' Development
Dr Stanislaw Tarkowski Director, Environment and Health, Regional Office for Europe, World Health Organization
10.00 Opening address by the Secretary of State for Employment
10.15 Overview of risk assessment
John Rimington Director General Health Safety Executive, UK
The paper will review the fundamental concepts of risk assessment before moving on to consider how quantitative information is obtained about different risks, and how structured risk assessment can be used to assist decision making in many different fields. It will discuss some general aspects of risk perception and consider bow different groups go about deciding which risks are acceptable and which are not.
11.00 Coffee
11.30 Risk assessment "users": their views and needs:
A consumer's perspective
Dame Rachel Waterhouse Member of the Consumers' Association and of the Health Safety Commission, UK
An employee/trade union perspective
Pekka 0 Aro European Project of te Finnish Industrial Workers
2 1...... 9 9 2
An employer's perspective
K C Williarns Vice President (Production) Exxon International, USA
An environmentalist perspective
Dr Ellen Silbergeld Professor of Pathology University of Maryland, Baltimore, USA
12.30 The transformation of the industrial economy into a service economy: the function of technology and the role of insurance
Professor Orio Giarini Secretaire General The Geneva Association, Switzerland
13.00 Lunch
The role of risk assessment in 00000000000000000 Internationalla 0 policy
Chairman: ProfessorLeszek Pacholski Chairman, Council of Labour Protection, Poland
14.30 Risk assessment - the international scene
The four papers in this session focus on the international role of risk assessment and what needs to be done to improve its contribution to priority setting and assisting developing countries.
Speakers Dr ChandraPinnagoda Chief, Occupational Safety and Health Branch, International Labour Office
Dr Kees A van der Heijden Director, Bilthoven Division European Centre for Environment and Health, World Health Organization
Dr Jim Brydon Head, Environmental Health Safety Division, Organization for Economic Co-operation and Development
Ron Haigh Head of Industrial Hygiene and Medicine Unit Commission of the European Communities
15.30 Risk assessment and the environment
Dr David Fisk Chief Scientist Department of the Environment, UK
Dr Fisk will discuss how risk assessment is being applied to environmental policy and suggest reasons for its apparent slow take-up. Taking man-made climate change as a case study, Dr Fisk will look at the pros and cons of extending risk assessment techniques over wide areas of government policy.
3 ...... 1 9 9 2
16.00 Tea
16.30 Central European countries' needs and what has been achieved
Dr Tadeusz Sulkowski Chief Labour Inspector, Poland
Dr Sulkowski will discuss the current state of the region's industrial infrastructure and the changes expected in the next few years.
Risk assessment: A local point of view
Milos Pale6ek Director, Occupational Safety Research Institute, Czechoslovakia
Mr Pale6ek will focus on Northern Bohemia, which is suffering the devastating consequences of long-term pollution.
17.30 Panel discussion of the day's proceedings.
RAPPORTEURS Jim Hammer President, International Association of Labour Inspection
Professor Jerry Jeyaratnam Head of the Occupational Medicine Division, University of Singapore
18.30-20.30 Cocktail reception at QEII Centre hosted by an Employment Department Minister...... 1 9 9 2
Tuesday, 6 October 11992
Risk assessment in practice 0000000000*0000000
Chairman: Dr Kerstin Niblaeus National Chemicals Inspectorate, Sweden
9.30 General approaches to risk assessment of chemicals
Patrick Murphy European Commission, Directorate-General XI
This paper will provide an overview of the process of chemical risk assessment and give an insight into how the process is applied in practice by scientists and regulators.
10.00 Risk Assessment of chemicals in Central Europe
Professor J,7adimirBencko Head of te Institute of Hygiene, Charles University, Czechoslovakia
Professor Gydrgy Ungviry Acting Director, National Institute of Occupational Health, Hungary
The authors will discuss how Central European countries have assessed the impact of chemicals on the environment and human health. They will focus on the reliance placed on limit values, and how this approach compares with those in the European Community and USA. Developing this theme, the authors will identify what help Central European countries may need as they review their approach to chemical risk assessment.
10.30 Assessment of pesticides in the USA
Dr Richard Hill Environmental Protection Agency, USA
Dr Hill will review the role of the Environmental Protection Agency in assessing pesticides, from initial evaluation through to the opportunity for public input and regulatory decisions. Looking ahead, the paper will discuss the consequences of possible moves towards international harmonization of pesticide risk assessment.
11.00 Coffee
11.30 An industry approach to the risk assessment of pesticides
Dr Barry Thomas Schering Agrochemicals Ltd, UK
Written from a European perspective, the paper will consider the industry's view of pesticide registration requirements and current practices in product development and product stewardship. It will also discuss what bearing international developments are likely to have on European standards and methodology.
5 ...... 9 9 2
12.00 Radiation Protection Standards: A Practical Exercise in Risk Assessment
Dr Roger Clarke Director, National Radiological Protection Board, UK
If an estimate is made of the risk per nit dose of radiation, then in order to set dose limits, an unacceptable level of risk must be established for both workers and the public. There has been and continues to be a debate about the definitions of "acceptable", "unacceptable" and "tolerable" and the attributing of numerical values to these definitions. This paper discusses the issues involved in the qualification of these terms and their application to setting dose limits on risk grounds. Conclusions are drawn about the present protection standards and the application of the methods to other fields of risk assessment.
12.30 Panel discussion
13.00 Lunch
Risk assessment in practice 000000*00000000000
Chairman: Dr Jim McQuaid Director, Strategy and General Division Health Safety Executive, UK
14.30 Food and drinking water safety: Can risk assessment help us to get our priorities right?
Dr H B Denner Chief Scientist (Food), Ministry of Agriculture, Fisheries Food, UK
Food and water borne risks can arise from different hazards whose relative importance varies from country to country. Can the formal practice of risk assessment avoid unnecessary use of resources while still minimising the overall level of risk?
15.00 Microbiological risk assessment and public bealth: The Department of Healtb Perspective
Dr Roger Skinner Department of Health
In this paper, te issues and problems concerning te application of formal microbiological risk assessment to public health, in particular food and water hazards, are explored and examples are provided.
15.30 Chemical safety in drinking water and food
Dr Maged Younes European Centre for Environment and Health, World Health Organization Dr Kees A van der I-Ieijden
Dr Younes will cover the challenges involved in assessing possible risks from additives and contaminants. He will focus on the problems related to adopting a safety factor approach to deal with uncertainty, and on the risk assessment process for food and on emerging issues in the risk assessment process for food and drinking water safety.
6 ...... 1 9 9 2
16.00 Tea
16.30 Risk assessment in decision making in research and development; use and release of genetically modified organisms
Dr Piet van der Meer Directorate for Chemicals and Waste Management, The Netherlands
Starting with a brief introduction to biotechnology risk assessment, the author will discuss how such assessment can help both maximise the potential benefits of biotechnology for human health and the environment, and minimise potential negative impacts. National and international examples will be used to illustrate the significant progress made already in this relatively new area of risk assessment.
17.00 Risk assessment for regulatory decisions on testing and release of organisms produced through biotechnology
Terry L Medley and Dr John H Payne US Department of Agriculture
This paper will discuss the statutory bases and intent of the laws and regulations for movement and release of organisms in the United States, and the complexity and uncertainty of ecological risk assessment as compared to assessment of human health risks. It will cover compliance with the National Environment Policy Act 1969 as a mechanism for documenting ecological risk assessments and assuring informed decisions.
17.30 Panel discussion
RAPPORTEURS: Dr Norman King Head of the Toxic Substances Division Department of the Environment, UK
Dr jaromir Skarkii Director, Safety Engineering Programme, Czech Environment Management Centre
Academician Professor Anatoly F Tsyb Chief, National Commission for Radiation Safety, Russian Academy of Medical Science
7 ...... 1 9 9 2
8 ...... 1 9 9 2
Monday, 5 October 1992
TFhat its risk assessinent and m hat do -%ve -%vant f romm it 09
0 0 0 0 0 0 0 0 0 0 0 0 0 a 0 ...... 1 9 9 2 Welcome Address Sir John Cullen Chairman, Health and Safety Commission
I am delighted to welcome you all to this Conference which is arguably the biggest and most wide ranging Risk Assessment Conference ever held - it is certainly the most international. Judging by the standing of the participants and their affiliations, it has all the ingredients to be a great success. It has been organized by the HSE and is a major event in both the EYSHH and in the UK's Presidency of the EC. I have no doubt that the conference win lead to increased awareness and use of Risk Assessment and subsequently to improved safety wherever it is used.
What makes the conference particularly topical is that, with the implementation of the European Community's Framework Directive, we see for the first time the concept of risk assessment built into European health and safety provisions. Yet the conference deliberations will cover much more tan just industrial health and safety - and will encompass many other countries than simply the members of the EC.
For evidence of this, we need look no further than the conference's co-sponsors, each of whom brings a special brand of expertise to the event. I am, therefore, particularly delighted to be sharing the platform this morning with representatives of each of our co-sponsors ... Dr Bill Hunter, Director of the European Commission's Health and Safety Conditions and Environmental Department at the International Labour Office ... Dr Jim Brydon, Head of the Environmental Health Safety Division at the Organisation for Economic Co-operation and Development ... and Dr Stanislaw Tarkowski, Director of the Environment and Health Division, World Health Organisation, Regional Office for Europe. Their presence here is of course the culmination of a major international collaborative exercise which stretches back nearly two years. From the outset, they and others in their organisations have played a major role in helping to plan the conference. Their participation in the International Steering Group was particularly useful. But here I must offer a special vote of thanks to all members of the Steering Group. Their breadth of expertise has helped us to compile a programme giving a unique overview of how risk assessment is used in a wide variety of areas. It is good to see so many of the group's members here today.
All in all, a great deal of time and effort has gone into organizing the conference. By the end of the week I am sure that we will all appreciate what a fine jobthe organizers and the Steering Group have done in producing this unique event.
With that, may I open this first session by inviting each of those alongside me to say a few words. Dr Hunter, perhaps you would like to start ...
I 0 - 1 ...... 9 9 2 Qpening Remarks Dr. Bill Hunter Director, Health and Safety Directorate European Commission
I would like to begin by congratulating the Health and Safety Executive on their initiative in organising this Conference. The Commission of the European Communities is particularly pleased to act as one of the co-sponsors, especially as the Community Directives have as their aim a reduction in the risks to workers safety and health.
The question of risk is often considered today in terms of hazard identification risk assessment, risk management or control and risk perception.
In order to be able to carry out a risk assessment the hazard has to be identified and information available on it. As an example, the Directive on classification packaging and labelling of chemical substances carries out evaluations of the hazard of these substances.
The fact that so many Community directives embrace the concept of risk assessment is itself an indication that it is a commonly accepted concept. Indeed, we are brought up to understand that there are risks attached to nearly all aspects of our lives. One must question whether this concept should not be introduced more formally into education at schools as it plays such an important role in all our lives. For example, in coming here today, either by public transport, private car, or on foot, we have all had to make risk assessments to avoid accidents.
Risk assessment is now an integral part of Community legislation, and has been for many years. A recent important example is the Framework Directive (89/391/EEC) on the introduction of measures to encourage improvements in the safety and health of workers at work which obliges the employer to implement measures to avoid risks, to evaluate the risks which cannot be avoided, and to combat risks at source.
Furthermore the employer has to evaluate te risks to the safety and health of workers, inter alia in the choice of work equipment, the chemical substances or preparations used and the fitting out of workplaces.
The concepts of risk assessment are also integrated into the individual directives based on this Framework Directive. For instance, the minimum safety and health requirements of workplaces (89/654/EEC) apply whenever required by the features of the workplace, the activity, the circumstances or a hazard. All these factors have therefore to be taken into account not only when a risk assessment is made but also when measures are taken to control the risk. It illustrates, also, that it is difficult to carry out a risk assessment independently of the measures required to control the risk.
The concept of risk control is perhaps best illustrated at Community level by the directive on the protection of workers exposed to carcinogens. In the first place the employer is required to reduce the risk by replacing a carcinogen with a less dangerous substance or preparation. When this is not technically possible the employer has to carry out manufacture and use in a closed system. When a closed system is not possible, the level of exposure has to be reduced to as
I I ...... 9 9 2 low a level as is technically possible, by the use of a series of measures which are spelled out in the Directive.
The easy recognition of the concept of risk assessment does not mean that it can be easily applied, in particular when it is governed by the force of law. It is the trigger for action, but it should not be used as an excuse for unjustified inaction.
It is already evident that there is a plethora of ways of carrying out a risk assessment. These range from simple "check lists" to more complex assessments involving general evaluations of the workplace and work rganisation. In many ways this is a reflection of the fact that a risk assessment has to be made for different circumstances. To draw a parallel, a geologist will have certain requirements when he buys a map of Great Britain, a car driver will have other needs, and a pilot will have yet others. There is no disguising the fact, however, that the social partners would like to have some guidance on how to carry out a risk assessment. However, the variety of occupations is so large that even experienced specialists are unlikely to be familiar with all the factors needed for a particular risk assessment. It may require a extensive literature search before such a specialist is able to visit and assess a given work situation, and this sort of exercise is likely to be repeated many times both in the Member States and elsewhere.
There are several ways in which to reply to such concerns. One idea which has been put forward to LO is to establish an up-to-date list of occupational hazards for workers in various common occupations which could serve as the basis for a risk assessment. Such a list could be based on the International Standard Classification of Occupations. Whether or not this approach is used, it is evident that we need a broad, orizontal method which covers all activities and types of risk and which also sets out the content and the procedure to be followed when making a risk assessment.
Finally, I would like to turn to the concept of risk perception. We are much more likely to accept a high risk activity wich we enjoy doing and wich gives us pleasure, than we are to accept a lower risk which is imposed on us. The risk of an accident from driving a car or playing soccer is far greater than from work, yet the public often perceive the risks as higher for safety and health at work. However, it is also important to realise that the public's perception of what constitutes a dangerous work activity may often be close to reality and in contrast with the public's perception of other risks. Thus, work activities such as agriculture, construction, fishing and the extractive industries are not only responsible for a higher rate of fatal accidents than other sectors, but are perceived to be so by the general public. It is for these reasons that the Commission made its proposals for directives in these areas.
In closing, I would like to say how much I would really like to be here for the five days of this conference. Unfortunately other obligations oblige me to be elsewhere. I look forward however, to hearing about the conclusions that you draw after five days hard work!
My Commissioner Mrs. Papandreou who is unable to be with us today, has asked me to tell you that she is very honoured at being asked to give the after dinner speech on Thursday of this week and that she looks forward to meeting you all.
1 2 ...... 1 9 9 2
Finally, it is my hope that this Conference will help to clarify many of the issues that I have raised. The question of risk assessment is not only of interest to us a but of considerable importance in achieving the ultimate goal of safety and health at work. I hope and trust that this Conference will mark an important step towards this goal, and I wish you every success with your efforts.
13 M ...... 9 9 2 Brief Introductory Remarks Dr K. Kogi Director, TRAVAIL, International Labour Office, Geneva
It is both an honour and a pleasure for me to welcome the participants to this Conference on behalf of Mr. Michel Hansenne, the Director-General of the International Labour Office as a co-sponsor and on my own behalf as a member of the Senior Advisory Committee of the Conference.
I am very pleased to witness a growing interest in international cooperation in the field of risk assessment as is evident from the distinguished patronage at the opening session. It is quite comforting to note that the WHO, the OECD and the Commission of the European Communities, as well as employers' and workers' circles, in addition to consumer groups are jointly involved in an endeavour to make a complete success of this Conference.
Risk assessment is increasingly recognised as a vital tool for the decision-makers in governments and industry to protect the health of workers, the general public and the environment. The Conference examines the ways in which risk assessment supports improving safety and ealth, in particular, at the workplace and in the environment. This get-together of experts from various fields will help us explore how the techniques might be improved or adapted to suit diverse needs and priorities throughout the world.
The Conference also stresses tripartism; that is the involvement of governments and employers and workers in the implementation of successful action programmes relating to risk assessment. Tripartism is unique to the ILO among the United Nations system in the establishment of social justice. I am very pleased to note that employers' and workers' views and needs in risk assessment receives prominence in the Conference.
Our experience in the world of work clearly shows that to have concrete results, practical local strategies must be established for the assessment and control of risks with a view to taking prompt action. International cooperation should aim at promoting such practical strategies in the local context.
Technical cooperation, particularly aimed at strengthening the capabilities of developing countries in matters related to risk assessment is another area of importance to the LO. Dissemination of information and training are key areas where assistance to developing countries is most beneficial. It is encouraging that the Conference deals with risk communication as one of its major themes.
The Conference, in particular, devotes time to examine the special problems of Central and Eastern Europe to help them prioritise and deal with risks posed by te ageing industrial infrastructure.
I am confident that this Conference will give us useful clues for a key to success of our joint efforts in protecting the health of workers, the general public and the environment.
May I conclude by congratulating the organisers and wishing you every success in your deliberations.
14 ...... 1 I...... 9 9 2 Opening Rem"ks Dr Jim Bydon Head, Environmental Health Safety Division Organization for Economic Co-operation and Development
It is a pleasure and an honor, on behalf of OECD and Bill Long, who cannot be present, to share the stage and share the welcome to you all.
OECD is pleased and proud to be a co-sponsor of this Conference. OECD, in its economic- and development-orientation, is involved in risk assessment procedures and policies in many ways. They vary from nuclear energy to transportation of toxic chemicals, which of course aie items considered by the Conference.
We believe that this Conference, as it is conceived and developed, is an excellent opportunity for experts and decision-makers in many fields to come together. We expect that they will be able to share their methods, problems and solutions. We expect tat this will lead to a greater awareness of the various ways risk assessment is carried out and will lead to a cross-fertilization of good ideas.
The Conference documents reveal that Risk Assessment eans different things to different people and that sophisticated methods are being used in some fields and being demanded in others. I look forward to a lively discussion and to some valuable conclusions at the end.
OECD congratulates Sir John Cullen, John Rimington, Sam Harbison and his staff on the preparation and rganisation of the Conference. The large and varied attendance here is a firm testimony to the value of this initiative by HSE.
15 ...... 1 I...... 9 9 2 Risk Assessment Stanislaw Tarkowski, MSc, DSc WHO Regional Office for Europe Copenhagen, Denmark
Mr. Chairman, Ladies and Gentlemen,
It is a great pleasure and honour to address you all on behalf of WHO on the occasion of this important international meeting in London. WHO is very pleased to be associated with it as a co-sponsor since the subject of the conference is of direct concern to our programmes related to environmental health. I would like to congratulate the Health and Safety Executive and to thank its director general John Rimington for taking the initiative for this conference.
Assessment and control of environmental hazards that may cause unacceptable risks to health are fundamental parts of programmes of protection of environment and public health. They are therefore integral parts of environmental health programmes within the frame of the WHO strategy Health for All, aiming at the protection of human health.
The public perception of risks to health has evolved through history and a level of risk once considered acceptable may not be acceptable today. For example, now that vaccine immunization programmes are available to prevent once-common diseases the public is not willing to risk outbreaks from infectious agents. Similarly, the growing knowledge of environmental health hazards, particularly related to chemical pollutants, and the continuous progress made in analytical methods and diagnostic procedures, leads us to expect with reasonable assurance that our surrounding environment and the products we use will not pose risks to our health and will maintain the environment is safe and healthy.
However, as the number and complexity of the environmental health hazards grow, so does the public concern as to their possible adverse effects. Governments and public health authorities are therefore under strong public pressure to prevent such risks and to establish procedures for their assessment and control.
Risk assessment is a basis for decision making. Its aim is not to justify a decision already made, or to find a cure for a lawsuit, but to provide a guidance on how to protect and enhance the public health.
The prevention and control of health risks imposed on people relies on available adequate information on health hazards for assessment of risk from exposure, and its evaluation in the context of other health risks encountered at the same time and options for regulations.
Risk assessment and risk management are in principle seen as two separate processes. Assessment of risk is basically a scientific process and an argument for separating these two tasks is that scientific analysis should be isolated from subjective judgements and political influences. A scientist often works best in separation in his ivory tower without immediate pressures of decision making on him. He needs to have a freedom to say "leave me alone until I have
16 ...... 9 9 2 the answer" A risk assessor is asked a legitimate question about a risk by public or their representative and must give an answer. He does not have the luxury of saying "come back next year when I will know more
Scientists are, however, involved in risk management through the results generated by their research and they are best aware that risk management demands that risk assessment be made with the highest possible accuracy and degree of confidence. At the same time they are the best to know that in risk regulation one is invariably dealing with incomplete and often inadequate scientific data upon which credibility of risk assessment and risk management depends.
The decision making must also consider the public propensity for taking various kinds of risk, which varies widely depending on a host of cultural, socioeconomic and psychological factors, as well as on the type and nature of the risk in question. Risk predictions are made i1i terms of statistical probabilities, which most people find difficult to visualize and understand. Public perception of risk is often markedly at variance with the estimates given by professional scientists. Meaningful communication on this subject between the scientists who generate the information, the public for whose benefit it is supposedly generated, and the decision makers who must make far-reaching decisions, is often insufficient and at some times even lacking. The decision makers endeavour to respond to pressures and demands from on the one hand an insufficiently informed public and on the other hand, from frequently biased interest groups.
WHO has been engaged in a whole series of programmes and activities concerning assessment and management of occupational and environmental health risks. The main role of WHO is to provide guidance on health criteria used in risk assessment, a basis for risks regulations. In doing so WHO in principle separates risk assessment from the process of risk regulation and leaves the final decisions regarding the magnitude of acceptable risk and the nature of control measures to national authorities as a matter of national policies.
WHO is, however frequently confronted with demands to assist and provide guidance on using risk assessment for solving specific problems concerning management of health risks. It has been our frequent experience that in such situations interlinking these two processes is conducted on the edge of capabilities of science, balancing between over and under conservative approaches seeing always a need to adhere to the principle of a prudence in prevention and control of environmental health risks.
Discussions continue on how to improve the accuracy of risk assessment and the credibility of risk management. This conference has adopted an agenda that hopefully will uncover new possibilities of making the process of risk assessment more coherent and risk management credible.
7 ...... 9 9 2 (Pening Address Secretary of State for Employment
Introduction
Ladies and gentlemen. It is a great pleasure for me to be here at the beginning of this historic Conference, and to add my welcome to those you have already received from Sir John Cullen and the various international bodies represented here.
A modern approach to health and safety
In opening this Conference I should like to revert for a moment to the European Symposium on Workplace Health and Safety which I had the pleasure of addressing in Paris on 9 September. Some of you may have heard my remarks on that occasion. But I make no apology for referring back to them today because I believe there is a vital message to be got across.
I see a need for all of us to re-examine the philosophy underlying our approach t health and safety - all the more necessary in a world in which the pace of change in industrial technology outstrips the speed at which we can legislate. The strategy for the future must include a critical look at the need for, and effectiveness of, regulatory proposals. All of us - Governments, regulators, industry, and citizens - need to be sure that regulations are soundly based and accurately targeted.
The penalties if they are not, are very large - extra costs, more bureaucracy, poorer competitiveness - but, above all, greater health and safety risks for people at work.
This is where risk assessment has a key role to play. The European Symposium underlined the value of risk assessment in ensuring that the real priorities are carefully weighed and that, first, the risks are identified and analysed to allow sensible, relevant and workable controls to be developed. Without proper risk assessment the legislation may not be right; will be difficult to apply on the ground; and will not achieve its health and safety objectives.
The task ahead
So this Conference has, in my view, a major task to perform. It falls to you, ladies and gentlemen, to take forward the development of risk assessment as a key input to all our future thinking on health and safety.
Of course, risk assessment as such is not new. It is what we all do in everyday life. But our approach tends to the instinctive rather than the analytical, to the assumed rather than the informed. This will not do at all for the sort of decisions that face us nationally, and internationally, in modern society. The potential risks - and the potential benefits - are very high. They must be properly assessed so that decisions can be taken, and defended, on a rational basis.
And we need to increase our capacity not just to assess the risks we already ave, but to predict new risks and their possible outcomes. In health and safety we have unfortunately tended to learn
18 ...... 1 9 I...... 9 2 from bitter experience. Most of the advances in health and safety in, for example, manufacturing industry and in mines came about through reacting to major accidents. Clearly, we will always need to benefit from hindsight. But increasingly it is foresight that we need.
This means a great deal of work to establish a common - and comprehensible! - language and understanding among experts, civil servants, regulators and industrial managers. The more that can be done to ensure that Governments are able to base decisions on the best possible assessments of comparable risks and benefits, the better.
If we can achieve this, it will go a long way towards improving our decision-making in many areas of society, and maximising the value we get from the resources we put into health and safety.
International aspects
For all these reasons, I warmly welcome the international nature of this discussion. It ranges much wider than the European Community, but it is a particular pleasure for me to be here during my term as President of the European Social Affairs Council. One of my principal aims is to ensure that health and safety is a priority for the Council of Ministers in Europe. We and other Member States are keen to see risk assessment techniques take their proper place in decision-making throughout the EC.
European delegates to the Conference will know that the European Community has adopted a Framework Directive on health and safety. This directive sets out for the first time the concept of risk assessment in European health and safety provisions. It paves the way for other more specific directives to be based on a requirement for risk assessment. The UK Government very much supports this and the benefits it brings of a common approach that transcends international boundaries.
On health and safety we are becoming more interdependent world-wide. The effects of major industrial accidents, and certainly of the overburdening of our environment, are international, if not sometimes global in scale. This means tat, unless we take preventive action, we may well export our problems to others less capable of controlling them, or even ignorant of their existence. This places on us clear duties of communication, as well as duties of restraint on self interest; and in order to communicate about risks, we need a common language of risk and of risk control.
Conclusion
This is the task before you. We have here today a remarkable range of experience and knowledge on this important and rapidly-developing subject. I believe this Conference will be a milestone in the development of risk assessment as a tool for practical decision taking. And as and when we have such a tool, and know how to use it, then I believe we sball be on the way to a healthier and safer environment.
I wish you all a successful Conference.
1 9 9 XA04NO302 Overview of Risk Assessment Mr. J. D. Rimington Director General, Health and Safety Executive
1. I want to begin by defining some terms. I shall then refer to a number of technical and other difficulties. Finally I shall try to set out why risk assessment is important and what its purposes are.
2. First, risk and risk assessment - what are they?
3. Risk is a subject of universal significance. Life is very uncertain, and we can achieve no object or benefit in it except by approaching nearer to particular hazards which lie between us and our objects. That approach represents acceptance of risk.
4. Risk assessment is a way of systematising our approach to hazard with a view to determining what is more and what is less risky. It helps us in the end to diminish our exposure while obtaining whatever benefits we have in mind, or to optimise the risks and the benefits.
Risk and hazard
5. The words "risk" and "hazard" are both French in origin. I believe the French language makes only a partial, if any, distinction between the two ideas. Perhaps "risque" originally conveyed the idea of a flirtation with danger for purposes of play, while "hasarder" meant something more fundamental.
6. But the English language, with its obsession with things or facts, and methods, found different meanings for the two words. "Risk" was probably used in its technical modern sense for purposes of conducting insurance business in the London market nearly three hundred years ago. The word implies a strict definition of a hazard to which the risk relates - for example, the loss of a ship at sea, and then some calculation of the chance of this happening, for the purpose of writing an insurance policy. "Hazard" was probably first used in English in its modern sense in relation to the game of golf, where it relates to a physical obstacle or trap into which the unlucky golfer may drive his ball.
7. So we may say that "hazard" represents a thing; risk, a calculation; and risk assessment, a method. Risk, we may say, is the chance that some thing adverse may happen. To be quite accurate, we must say that the concept of risk includes three linked components; first a probability, second an event and third the severity of the adverse effects attached to the event. These may themselves entail a further calculation of probabilities. Greater knowledge of these matters can affect our behaviour.
8. These adverse effects are given names. They are respectively "harm" and "detriment". It is convenient to employ the word "harm" in relation to something living, usually man or the natural environment. Thus "harm" is something we would seek to avoid even if no definite economic cost could be attributed to it, or if it were not possible to define and measure all its implications, e.g. emotional or aesthetic. It is in fact, something about which there can be
20 ...... 1 9 9 ...... 2 disagreement on ethical grounds - some people think certain effects "harmful" that other people do not, the area of agreement being greater in relation to man than in relation to nature.
9. The word "detriment" may be taken to apply to some form of economic loss, wich might indeed include a valuation of harm to living things but which might also include damage of a much wider kind, as for example from the accident at Chernobyl which rendered land uninhabitable. "Detriment" can apply either (1) to the quantum of damage which might be caused by the chance realisation of a hazard multiplied by the probability of this happening in a particular time period, or 2) to the quantum of damage attributable over a period of time to some existing source of continuous damage.
10. I will refer finally to one other word: "Consequence". This word is used to refer to effects flowing directly from a hazardous event, some of which may be mitigated by appropriate action such as the evacuation of the local population. The idea of "consequence" does not include whatever steps may be taken to reduce or contain the event itself.
11. So much for definitions. I come now to the second, and I fear much longer part of my disquisition. I now approach the difficulties, disputes and ambiguities which we encounter in making tose concepts useful allies in the impious task of achieving greater certainty in the world's affairs.
Acceptability of risks, or hazard
12. We accept many risks, that is to say, we approach many hazards, in our daily lives. Most of those who have studied the subject agree we are prepared to accept higher risks where the acceptance is voluntary. Indeed, one student(l) made calculations suggesting that people are prepared to accept sporting risks (such as skiing) roughly a thousand times greater than from involuntary hazards, for example food preservatives. If this is so, two reasons may predominate. First, voluntary acceptance of risk usually implies that the benefit of doing the risky thing accrues to the individual taking the risk. Second, an individual may think himself better able to calculate his chances of avoiding harm from a hazard he knows well.
13. So the acceptability of a risk is a very personal matter. And yet in modern society there are many species of risk that the individual can scarcely refuse - for example, the risks attached to living near some hazardous installation. In such cases, individuals are entitled to expect that steps will be taken for their protection; to know what those steps are, and what is the extent of the remaining risk. In relation to such "involuntary" hazards, a given level of risk could scarcely be thought fully "acceptable" unless it had been reduced its due proportion in to the kind of background level that individuals accept for life generally, and before they begin to consider taking particularly high risks for particular objects and benefits - as for example going skiing or suffering medical treatment. At this "background" level, the risks may not be negligible; but they will be low. We can call this level the region of acceptable, or broadly acceptable, risk.
Perception of risk
14. The public's perception of risk is a subject we shall be debating at length later this week. The
...... (1) S. Starr. "Social Behaviour versus Technological Risk", Science, 1969, 165: 1232.
21 ...... 9 9 2
problem of perception begins with the fact that when non-experts are asked how they rate risks, they gave different answers from experts.(2) When the reasons for this are examined, they seem to concern the fact that people generally attach less importance to probabilities and chances than they do to consequences. When, for example, people see a horseman approaching a high fence they may say "that's risky" meaning not so much that there is a high probability of his striking the fence, as that the consequence will be nasty if he does.
15. In other words, of the three components of risk which I referred to a moment ago, the public look first at the second and third components that is to say the event and the consequence, while the experts begin with the first, that is to say the probability. The attitude of the non-expert may partly arise because most people may think they know the probability better than in fact they do. There are studies(3) for example which seem to show that people under-estimate familiar risks and over-estimate unfamiliar ones, including those the media choose to portray more frequently or dramatically.
16. There may be another reason, belonging unfortunately to the vexed dimension of psychology. People have a tendency to detach particular objects of worry from teir place in reality. Nature offers us this facility for purposes of calculation and also for purposes of flight(4); but if we can neither calculate nor flee, if in short we think there is nothing we can do, we have the choice only of forgetting or of fretting - of allowing the thing too great a significance. It then becomes a scar on our mind's eye, a neurosis.
17. However, we must not make the mistake of over-estimating the irrational elements in public perception, or under-estimating the irresponsibility on the part of politicians, experts, businessmen, who have sometimes found it convenient to deny the existence of risks that palpably do exist. If the public sometimes seem to hanker after zero risk, who first taught them that zero risk is possible? Today, and this week we are confronting a real need for understanding and communication on this subject, for we cannot flee from all the hazards that surround us. Some we must tolerate - but which and to what extent? To know that better is the underlying aim of this Conference.
18. We must also acknowledge an ethical dimension in the public's attitude. There will always be some people who judge that certain hazards should not be entertained at all, no matter how low the risk: some people feel this, for example, in relation to nuclear power. And why do they feel it? Partly because of an association, however badly founded, with the bomb. There is a dread, too of the old fear conjured up by Dr. Faustus: the fear of mankind messing about with too much power. And there is its opposite, now being brought about by some of the ecological questions that confront us - that we may have too little power to restrain ourselves from irreparably damaging our only habitat, earth.
19. For all these reasons we must assume that we shall never achieve complete agreement about relative risk: the best we can do is to inject the greatest degree of objectivity we can into a subject
...... (2) P. Slovic, B. Fischhoff, S. Lichtenstein "Rating the Risks - The structure of Expert and Lay Perception, Environmental Impact Assessment and Risk Analysis" NATO-ASI Series Eds Springer-Verlog 1985. (3) S. Lichtenstein et al "Judged frequency of lethal events" Journal of Experimental Psychology: Human learning and memory 1978 4 5Z (4) R Slovic, B. Fischhoff, S. Lichtenstein "Facts and Fears: Understanding Perceived Risk" in R. Schwing and W. Albers (eds) "Societal Risk Assessments: How safe is safe enough?" Plenum Press NY 1980.
22 ...... 1 9 9 2 where fear and emotion will always play a large part. But to do that would be no negligible achievement. And the first step is to concentrate attention on those aspects of risk that the non- expert tends to neglect - the actual level of each risk and the benefits if any of running it. This is the heart of risk assessment. The political aspect
20. Of course, no one, Governments or otherwise, incurs risks except for some major benefit or unless by doing so, some greater risk can be avoided. The question of the balance of risk and benefit in imposing on people particular risks above the "background" level is one for society as a whole, not for experts. It is a political decision, which like others, needs to be taken in the light of the best possible information.
21. There is another reason why major risks undertaken by society for a benefit represent a political question, and that is because many such risks are of a nature where one group in society accepts most of the risks, but most of the benefits go to others. This is of course fundamentally the reason for what we call in the UK the "Nimby" syndrome - "not in my backyard". And risks may also be shifted through time so that they fall to future generations, giving rise to ethical questions. Like all questions of redistribution, these can only be settled through our political institutions. Tolerable risk
22. It seems evident that a considerable "grey area" exists between what is a normally acceptable or background level of risk, and levels we might for practical purposes regard as unacceptable. In this "grey area" people will accept risks in order to secure benefits. But in doing so, they may reasonably want to know what the nature and level of the risk is, so that they can compare it with the benefit; and to be confident that the risk is being controlled and where possible reduced.
23. It is this "grey area" that is the most interesting to us. For the principal benefits for which anyone will undertake a risk which he himself does not control are his livelihood, and the maintenance of his social infrastructure - for example, the production of electricity or the maintenance of food or water supplies on which all our comforts and lives depend.
24. In the UK we have called this "grey area" the region of "risk tolerability". Tolerability does not mean "acceptability". It refers to a willingness to live with a particular hazard so as to secure certain benefits and in the confidence that it is being controlled. To tolerate a risk means that we do not regard it as negligible, or as something we might ignore, but rather as something we need to keep under review and reduce still further if and as we can. To accept a risk by contrast, means that for purposes of life or work we are prepared to take it pretty well as it is.
25. There is no need for me to delve into all the implications of the concept of "tolerability". They are fully explored in the revised document "the Tolerability of the Risk from Nuclear Power Stations" which is before this Conference. Suffice it to say that the concept depends upon our ability to quantify and compare risks in such a way that people, and Governments, can make informed judgements about them, and to decide in particular whether the available benefits are worth the risk. It also provides a basis for legitimising social and political decisions to accept
23 ...... 1 ...... 9 9 2
significant risks from which some people receive greater benefit than others. It involves a statement of the standards and limits that society undertakes to apply as part of this process.
26. There is no implication that these standards will be the same the world over, for the reason that some societies may unhappily be forced to pay a higher price for benefits that others can take for granted. But the geometry, and the basic reasoning is the same wherever the citizens' voice determines the policy of the Government.
Risk assessment: basic terms
27. So far, I have spoken of standards and levels of risk as though it were self evident that the extent of any risk can be assessed in some objective manner and compared with other risks. But there are many difficulties about this which this Conference will need to explore, and which I had better begin to describe. Some of the difficulties arise from the need to give precision and analytical depth to something, namely the consideration of hazard, which is innately personal and instinctive. Risk is about fear, and fear is a shapeless and emotional fact about the human condition. In order to study it, we have to pretend that this is not so; and say instead, as we have done, that risk is about the probability of particular things happening. In order to analyse and measure that, we have to state precisely which things we have in mind, and to whom the risk attaches. We have to find proxies which express for comparative purpose matters which do not normally lie within the rational parts of our mind.
28. It is this that makes some experts despair about the problems of risk communication. To them, I will simply say, what's new? All thinking is about giving structure and predictive weight to things that we would otherwise settle by instinctive means. What's new is the need to apply to the subject of risk some of the methods by which the human race has given order and rationality to so many of its affairs. So now I come to the rules, the proxies that we need to apply to give this great subject a structure, first warning that nothing I can say has final value.
29. The first and cardinal structural point is that the measurement of risk depends upon complete clarity as to the hazard, whom it affects, and what consequences are in view.
30. Individual risk. To the individual the first question is always, what is the risk to me or my family from the realisation of the hazard in question. To answer this, we have to construct a hypothetical individual who is in some fixed relation to the hazard - let us say, the person most exposed to it, or a person living at some fixed point or with some assumed pattern of life. Other individuals can then, as it were "aim off" and reckon that they or their family have more or less risk than the hypothetical person.
31. The second step that is then necessary is to find a means of comparing the risk to which the hypothetical person is exposed with other risks that are within people's understanding - risks that they already accept and know about, like the risks of childbirth, or the annual risk of death from driving - both of these, by the way, are similar - of the order of I in 10,000 or I in 104 per event or per annum. Those are very useful datum points.
32. This, then, is "individual risk", the level of risk to an individual or for that matter to any group of individuals placed in a similar relation to some hazard. We must of course recognise that not all hazards will appear to individuals to be identical, either in themselves or in relation to any benefits they may perceive. People may regard the hazard of death in childbirth as very
24 ...... 9 9 2 different from the hazard of death from cancer due to a source of radiation that brings someone else a profit. We may have to decide that there is a general aversion to particular risks, and to factor this into any calculation. A statement of the comparative risk levels is nonetheless an important contribution to rationality.
33. Societal risk. Calculation of the risk of harm to an individual is not free from technical complication. But I now come to a subject of much greater conceptual as well as technical difficulty, namely the risk to society as a whole from particular hazards.
34. This is what the modern risk debate is all about. In recent years, we have become more and more conscious of major risks the extent of which we have few present means of assessing. Some of them are on an ecological scale - like the hazard of greenhouse gases. Some are from hazards of a continuing nature affecting the natural environment on a smaller scale - the risks from chemical wastes or residues for example. Some are concerned with the possibility of major events, such as nuclear accidents or other catastrophes resulting from the uses of technology. Some concern the risks of major accidents in connection with transport.
35. One might of course ask - why bother to assess these? After all we live not just with technological threats, but with political ones - of war, or of the breakdown of social order or from the movement of populations and so on. Let us nevertheless try to do so, since the risks from the uses of technology are risks which in principle we can control and about which we have choices. It is the fact of these choices that is the significant point. Risks from technology are risks flowing from the investment of resources; alternative investments, or self denial as to the benefits, are usually possible. The reason why we should assess the risks is to help us optimise our investments or withdraw from those that are potentially too harmful. What we are looking for is a rational basis for those decisions.
36. But there are difficulties about identifying the concept of societal risk and about measuring it. In te case of really big risks created by major investments and political decisions, we have to analyse all three elements that make up risk, remembering that there may be two sets of probabilities piled on each other - the probability of an event and the probabilities that particular consequences will flow from it. What in principle we are looking for is a quantity that can be compared at least roughly with the not benefits for society undertaking the risk or with the net benefits from running alternative risks. An example might be the risks and benefits from a nuclear programme as compared with those from generating power from fossil fuels.
37. Such a quantity of adverse consequences we have already referred to as a detriment. We can in principle categorise the detriment associated with some major hazard as falling into three categories (a) the price put on the loss of life (b) the cost of coping with the emergency, loss of plant destroyed or rendered unproductive and opportunities or other investments forgone and (c) the costs if they can be estimated of the disruption to social and political life. We can refer to those, if we like, as respectively the human, monetary and political costs of a major event should it occur. These (putative) costs have then to be combined with the probability that the event will occur to give the quantity "detriment".
38. So much for the geometry. It is when we come to assign money values to various factors in this equation that the difficulties arise. It becomes obvious that many conventions, some of them perhaps artificial, would have to be agreed upon to give an idea of the sums involved, including answers to questions about discounting the values through time. And this is before we consider
25 ...... 9 9 2
that there is not one, but many shapes that the consequences of an accident can take, and that each shape has a different probability. All the same, if the conventions could be agreed, the prize would be considerable. We would be in a position, at least roughly, to compare and contrast different investments and different social burdens, either of risk or of continuous damage from known, existing sources of detriment. Uses of risk assessment
39. In considering the uses of risk assessment we must not of course confine ourselves to these very large issues of major industrial hazards, global ecology and the like. On the contrary, any disciplined approach to hazard, for example hazards met with in work places, needs to be preceded by some form of identification and an assessment, however rough and ready, of the extent of the risk and the need for further precaution. This in fact is the present tendency of European legislation on worker safety. For example, European law, through the Seveso Directive lays down that operators of hazardous plant must make a safety report to the regulator; a report which is of course an expression of the hazards and if possible of the relative risks.
40. Such reports, recording the structure of the risks present in the plant, form a framework for later inspection both by the regulator and the operator. A good safety report is a classical expression of risk assessment, since it enables those responsible to prioritise their subsequent actions.
41. A further use of risk assessment is to enable one risk to be balanced against another for purposes of investment in safety. A recent evaluation of the risks on the London Underground showed for example that the hazard which was attracting the largest investment - fire - represented given existing precautions only I % of the remaining risk. But balancing of risks is important for other reasons also. It is no use in constructing a safe system to have ten strong points and one weak one. A very strong room with a weak lock on the door is a waste of all the money lavished on it. The balancing of risk in tis way is tus a feature of te design of all hazardous plant.
42. Again, the underlying principle is of risk assessment assisting us to prioritise action, on the basis that resources are not finite, and that many hazards need to be tolerated and controlled. We can expect this approach to govern the risk assessment of chemicals which will in turn govern the principles of control applying to each. In this as in other uses of risk assessment, it wil be important not to bias the approach by ignoring certain forms of risk because others are momentarily more fashionable, or of assuming that consequences whatever they are cannot be mitigated through proper forms of control. Risk assessment is about benefits and alternatives as well as about hazards. It is always about giving proper structure and weight to any detriments so that we can compare them with the benefits.
43. Talking about risk is of course one of the riskiest things one can do. There are so many experts about. May I conclude therefore by expressing the hope that my attempt to describe the achieved position on risk and risk assessment will not awaken sterile controversies about unimportant matters. The Conference hopefully, is about the progress we can now make, not an opportunity for further discussion about matters on which there is already sufficient agreement. So, finally, may I set out one or two questions on which I hope we shall be able to make progress.
26 ...... 1 9 9 2
44. First, I hope we shall all be able to get a clearer view of the concept of societal risk, that is to say how to structure and evaluate the detriment attached to sources of damage, actual and potential which are widespread and important enough for society to want to regulate.
45. Second. I hope we shall be able to agree on some of the principles that should attach to communication about risk, recognising the difficulty of the subject arising from acceptance of risk being ultimately a deeply personal as well as political matter on which full agreement will never be possible.
46 Tird I think it would be very convenient if we would make a start towards better understanding, even agreement as to some of the factors that will need to be used in quantifying and weighting risks, if the subject is to make real progress. For example, if it is true that society is more averse to catastrophes involving the death of many people at once than to an equivalent number of death seriatim from a single hazard, how should we express.this in our calculations? And can a value be given to life? And is there such a th ing as a "political risk" from a major event or should we ignore this aspect? Is there a world view about levels of risk, for example of death from particular major hazards, that are acceptable, unacceptable or tolerable under particular conditions?
There I conclude, with the wish that our Conference should at least be interesting and provocative.
2 7 1 9 10 2 XA04NO303 Risk Assessment: 'A Consumer's Perspective' Dame Rachel Waterhouse Member of the Consumer's Association and of the Health Safety Commission, UK
Many everyday choices which consumers make involve some form of risk. Different types of transport carry a different likelihood of accidents; financial services can affect the future well- being of the family; food can be a vector of health or disease. Frequently, however, evaluation of risk is not the main arbiter of choice. Speed and convenience are more likely to determine whether one flies from Aberdeen to London rather than travels by train or road. In financial services some investments may be viewed as risky but facilities such as current and deposit accounts are assumed to be safe havens for customers' money. Consumers expect the food they buy to be safe to eat.
Nevertheless, where there is considerable intrinsic public risk over which they have no direct control, consumers will expect that risk to be reduced to a practical minimum. Public transport must be licensed and subject to inspection. Electricity and gas suppliers must meet high standards of safety. Work activities that affect the public must be subject to legislation or Codes of Practice. Any danger of a major accident in nuclear, chemical, or other installations must be identified, closely supervised and emergency plans worked out and communicated to those in the neighbourhood likely to be affected. New authorities dealing with high profile commercial ventures where there is little safety experience, such as the Channel Tunnel Safety Authority, will undergo particular scrutiny. Control of new processes such as genetic modification needs to be clearly explained and both procedure and the results of risk assessments should be accessible.
The regulators in each of these instances represent consumers and are established to defend consumers' interests. For this reason they must have access to appropriate data and powers to extract information. In turn, they can expect the quality of these procedures to come under close public examination. In particular the Authority concerned must communicate to consumers the standards and criteria it has adopted to identify the minimum acceptable level of risk.
In 1986 Sir Frank Layfield, in his report on the Sizewell B Public Inquiry, said "the opinion of the public should underlie the evaluation of risk; tere is at present insufficient public information to allow understanding of the basis for the regulation of nuclear safety". This judgement should extend far more widely than the nuclear industry. Lack of information is widespread and too often both suppliers of goods and services and regulators have failed to give consumers any understandable indication of likely risks. The danger of major hazards was hammered home by the Seveso and Bopal disasters; public anxiety has been increased. Sir Frank Layfield's message has been adopted by the UK Health and Safety Executive in its paper on The Tolerability of Risk from Nuclear Power Stations: "final judgements about whether a given risk is tolerable are not matters for experts alone, but for te people who have to bear the risks, and who are therefore entitled to be given the best possible advice about them". Consumers strongly support this view.
This concept of tolerability of risk needs to be adopted across all sections regulated by safety authorities and some coherent approach developed. Consumers should not be asked, or expected, to accept one method of risk assessment for one activity and a different level for
28 ...... 9 9 2
another. Some generally accepted tolerability of risk criteria, involving a risk/benefit analysis, would enable consumers and regulators alike to accept a more consistent and realistic attitude to safety policy, avoiding the knee-jerk reaction that each individual accident and tragedy currently produces. Such reactions all too easily underestimate some of the longer-term but less visible risks, including many of those associated with inadequate diet, poor public health standards and environmental hazards. Not only individual consumers, but legislators and the media have been reluctant to adopt a consistent yardstick by wich to measure risk, to make comparisons of risk, to distinguish major from minor risks, or to determine the best use of public expenditure for the protection of people generally.
More directly, however, consumers exercise choice in buying goods and services in the market- place. With some products they expect a very low or minimal risk and in the UK over the years legislation has been introduced to protect consumers by banning the sale of certain unsafe goods and by requiring safeguards where appropriate. Consumers expect children's toys to be Aafe for children to play with; they expect the food they eat to do them no harm; they expect gas fitters to avoid all unnecessary risks in carrying out their work.
Many consumers rate choice very highly. They wish to have the choice to take risks. In many instances, however, the decisions of one person may put others at risk as well. This 'societal' risk occurs, for example, where a driver fails to change a worn tyre; where a motor-cyclist refuses to wear a crash helmet; or where motorists or their passengers prefer not to wear seat-belts. In each of these cases the cost in lives and injury has been demonstrated by statistical evidence to be so high that society has accepted legal enforcement of these safety measures. This in turn has necessitated the development of safety standards for the products themselves. Tyres, seat-belts and crash-helmets must give an adequate level of protection.
Despite tis, both motor-cyclists and cars remain hazardous modes of transport. Consumers are still able to choose their own make and model of vehicle. Not all information is, however, accessible in the public domain to enable consumers to make risk comparisons when taking a decision. It has taken many years to get even basic accident statistics collected and available according to make of car. Sometimes knowledge of weaknesses in particular model designs have been revealed by the tests of consumer organisations before the manufacturer (or for that matter the public authorities) acknowledged their existence. Subject to adequate overall standards, freedom to buy faster or smaller, safer or larger cars is the right of customers, but in making their choice they should have access to all known risk factors.
The point at which freedom to choose should be limited is a matter of political debate, especially where the potential damage is likely to be confined to an individual or group of individuals. Unpasteurised milk and products made from it may carry microbiological hazards which are destroyed by pasteurisation. In Scotland the sale of unpasteurised milk is prohibited. Elsewhere consumers may choose. In order to make this choice the products must be clearly labelled and information about risk, particularly to vulnerable groups, must be disseminated. There is no satisfactory method of assessing and conveying the level of risk, however, either in unpasteurised milk products or in eggs. Currently in the UK, Government advice is that pregnant women should not eat pAt6 or liver, but no indication exists of the level of risk associated with eating either food. Measures to control microbiological hazards in food, therefore, have rsted more upon scientific possibilities than upon risk/benefit analysis. Computer modelling programmes may help to remedy this situation.
29 ...... 1 9 9 2
Labelling of products either where hazards are not controlled or to support legislation that has been introduced is an important route for spreading information to consumers. Health warnings on cigarettes and on pharmaceutical products are normal. Care labelling and date labelling ave evolved over the years. Nutrition labelling is now required.
To convey complex information by labels is not always easy. Textiles, including furniture, are generally considered to be the most hazardous flammable products in private dwellings. Since 1988 in the UK use of a dangerous sort of polyurethane foam in furniture has been outlawed. The foam burns fast and produces thick toxic smoke - a lethal cocktail of carbon monoxide and cyanide. Polyurethane foam can now be manufactured with an additive that causes it to burn much more slowly. Cover material or interlining must also be match or cigarette-resistant. Permanent labels have to be attached to all furniture offered for sale to inform consumers of its fire-resistant status. Greater reliance must be placed on the banning of the dangerous substance: it will take time for labels to become familiar to consumers. Moreover, although the products are safer, there is no risk assessment, and there is always a danger that 'safer' may be regarded as 'safe'.
Because furniture is the type of product that may be expected to move freely in the European Single Market, UK consumers particularly welcomed the draft Directive on foam furniture which was first discussed in 1989. Safety regulations, of course, have often been used as barriers to trade in the past and UK consumers were anxious for a Community ban on the use of unsafe polyurethane foam. The draft directive was withdrawn in mid-1991, te main reason given being the variations in culture and attitude towards these particular hazards, and the need for further research, which will not be concluded until at least the end of 1994. Although not opposed by the UK government, this action caused considerable concern and alarm among UK consumers.
Many specific hazards are already covered by EC Directives. Electrical heating and cooking appliances are covered by the Low Voltage Directive; gas appliances by the Gas Appliances Directive. Attempts, however, to harmonize legislation for every product or potential product has proved difficult and time-consuming. In June 1992 the Council of Ministers agreed the General Product Safety Directive which will cover all those products for which no specific safety requirements have been included in any other Directive. It is designed to establish at Community level a general obligation to market only safe products and to provide information on any acceptable risks. Member states must provide systems for ensuring compliance and for dealing with dangerous products.
A safe product is defined as one which 'under norma 'Ior foreseeable conditions of use, including duration, does not present any risk or only te minimum risks compatible with the product's use, considered as acceptable and consistent with a high level of protection for the safety and health of persons'.
The Directive also requires that 'within the limits of their respective activities' producers sall provide consumers with the relevant information to enable them to assess the risks inherent in a product throughout the normal or reasonably foreseeable period of its use, where such risks are not immediately obvious without adequate warnings, and to take precautions against those risks'.
How producers will arrive at their risk assessments and how the information will be conveyed to consumers has yet to be determined. There are, however, several aspects of the Directive which
30 ...... 9 9 2
are welcome. The extension of the assessment of risk to the whole of the product's life appears to be innovative and important.
There is clearly scope within the Directive for distinguishing between those products which consumers are entitled to expect to be almost entirely free of risk - children's toys, food-stuffs, clothing - and those products which will inevitably be associated with risky activities - fast cars, power tools, motor-bikes. 'Compatible with the product's use' would suggest a possible scale of risk.
Furthermore if the Directive is to be fully implemented it will require better labelling, improved storage and stock control, better batch traceability, possibly better manufacturing standards and concern for the environment.
There has been strong resistance in the past to giving consumers information about risk on the basis of their incapacity to handle statistical data. Yet, large numbers of people engage in gambling and betting on horses or in card games and there is a high level of understanding of the nature of probability and of the means of assessing it. The odds on race horses are quoted in surprisingly complex ways: 2 to 1, and 3 to I are straightforward, as are odds to 2 or I I to 4; ordinary punters seem to have no difficulty in appreciating odds of 100 to or 100 to 6 and even 100 to 30, which represents a sophisticated difference between 11 to 4 and 7 to 2 Almost every High Street has a betting shop where these probabilities are assessed and money wagered in consequence every day.
Given the capacity of consumers to cope with information once their attention is engaged, the possibility of developing a scale of risk becomes extremely interesting. The concept is not strange in the health services, where a patient may be told that a particular operation has a 7 in 10 chance of success, but much research will be needed into whether an easily understandable system of units and notation could be devised, or whether a figure of probability 'p' would be more comprehensible. Within the EC a simple numerical form would have the advantage of international currency. If it became a mandatory European standard to support the General Product Safety Directive, consumers would soon become familiar with risk assessment and better able to use the information in making purchasing decisions. But it should go without saying that such information requirements must be properly tested on consumers first to ensure that they are both useful and comprehensible.
31 9 XA04NO304 An Employee/Trade Union Perspective To Risk Assessment Pekka 0. Aro Project Director European Project of the Finnish Trade Unions
Risk assessment does not easily lend itself to precise description. In some of its current forms of application, it is little more than a modern way of throwing bones or looking at tea leaves or the entrails of seep, like our ancestors used to do not so many centuries ago (the only real difference being perhaps that laboratory rats are cheaper and less messy than sheep). At its best, however, risk assessment can be used as a good tool to identify the needs for changes required to reduce hazards and to prevent accidents to humans or damage to the environment. This paper approaches risk assessment as a tool, highlighting the need and usefulness as well as the right of employees and their representatives to participate in all the steps involved in assessing the risks of the workplace and its environment.
NVhy should employees be involved?
The essential reasons for involving employees and their representatives in risk assessment can be summarized as follows.
Employees are in most cases first to be in direct contact with the substances, processes and technologies which they produce, use, store or transport. They are the most likely potential victims of an accident. Therefore, as part of the overall protection of their personal safety and health, employees have an immediate interest in any activity related to the safety of the workplace.
Employees at an installation have first hand experience of the operation and the risks involved in that installation. As a rule, they have a long term commitment to their place of work. Given the opportunity and facilities, they want to make this experience available for the safe operation of the installation. Employees are also in a position to recognize dangerous situations and to take action to prevent them or mitigate their consequences.
Employees and their families live in the community surrounding the workplace and have, therefore, a concern over the environmental well-being of that community. The safety of the workplace is inextricably entwined with that of the environment. This is a oncept which seems to be surprisingly hard to digest, given its overwhelming importance. Employees and their representatives must be involved in environmental considerations of their work and work-related safety activities must take the external environment into account.
A comprehensive risk analysis and assessment in the steel mill of Rautaruukki Oy in Raahe, Finland involved directly over 100 persons and identified some 700 different hazards. A crucial experience resulting from the exercise has been that safety thinking increased permanently in all levels of the personnel, particularly because the line rganisation and safety representatives of employees were directly involved at all stages, including decisions to take corrective action. Days lost because of sickness dropped from over 1,000 per one million hours worked down to 500
32 ...... 9 9 2
(which is quite low by international comparison of steel mills) in three years after the plant-wide search for risks was begun. The activity has continued in several areas with more detailed analysis. The key has been the involvement of employees and their representatives which has made it possible to solidify the good results.
How should the employee involvement be organised?
The type and scope of employee participation in risk assessment depends on, i.a. whether the assessment concerns the planning/development stage or normal operation of a process or technology. In assessing risks in an operating installation, the ways to include employees depend on the industrial relations practices and the line rganisation. Before normal operation the situation is slightly more complicated.
Before the object of risk assessment is integrated into the normal operation of a work place, the employee input could come from the intended users. If, for example, a machine is being developed for the internal use at the plant, those persons who will be requested to operate it should participate in assessing the risks involved. If the machine is to be transferred, the participation should come from the work place where the equipment is going to be operated.
Employee participation can meaningfully be arranged only through representative and cooperative mechanisms where the employees' self-chosen organisations, trade unions, act in the interest of their members. Safety committees and employees' safety representatives are the most commonly used structures which have also proved their appropriateness at the local level. They should also be used in risk assessment and in drawing conclusions on the basis of the assessment with a view to risk reduction and mitigation.
Assessment is always a political choice
A problem wich is frequently mentioned as a problem in risk assessment is its narrow focus on the technical properties of one isolated piece of equipment, substance or work area. Many of the commonly used risk assessment methods concentrate too much on individual aspects of risk. There is a need for more emphasis on interaction between the person(s) performing the work, the technology, processes and substances used for the actual work, as well as the surrounding environment, both at the work site and outside.
In 1978, the U.S. Supreme Court struck down OSHA's benzene standard, stating the OSHA had failed to address whether the risk of cancer was "significant". The Court didn't define "significant", but in a footnote stated that a reasonable person would surely find a risk of one in 1,000 to be significant, and a risk of one in 1,000,000,000 to be insignificant. Since then, the task has been to show that 30 years of exposure at lvels higher than those proposed lead to a risk of at least one in a ,000. Environmental levels of "significant" are more complicated.
Obviously, the request for a wider, more dynatnic assessment of casualties also entails substantially more complicated requirements placed on the assessment methods and the assessors. But then, assessment is always a social and economic - and therefore political - choice. This should be admitted and addressed, not papered over. The tolerance level of residual risk is of necessity a value judgement of the society or the enterprise and it is not primarily based on objective information.
33 ...... 1 9 9 2
The object of assessment should be as broad as possible without losing its focus. It should take account of new development of substances, processes and technologies, as well as new information about tem. Biological hazards are emerging as an example of a potential risk caused by production and use of organisms. They can not be excluded from the scope of safety activities simply because they are not mentioned in most existing legislation. Transport is another example of an important area which is difficult to approach. The difficulty - or novelty - should not be used as an excuse to avoid or neglect these areas.
The choice of criteria should be made more transparent. One possible approach is to adopt a policy of choosing assumptions which maximise the calculated risk. One would assume that, regarding toxic substances as an example, exposures in the human studies are at the low end of the range, that humans act like the most sensitive animal species, that dose response follows a linear pattern etc. For environmental effects, the object of assessment would be the risk to a hypothetical person standing for 70 years on the fence line in the prevailing direction of the wind. That way, risk assessment becomes a worst-case scenario.
Training and right to know
The more interactive, system-oriented risk assessment advocated here requires from employees a broad understanding of the process with which they work as a whole, not only at their specific work area. Only then can they contribute fully to the identification, analysis and assessment of the risks and to the conclusions and recommendations resulting from that. This broad understanding can only come as a result of both training and access to all relevant information. It is important to differentiate between the two.
Training should be organised for every employee before his/her undertaking normal duties and follow-up training should be provided regularly. The training of different levels of line management should be arranged at least partially together with workers to ensure understanding of everyone's functions in different situations. Safety personnel and representatives should be given access to specialised training on risk assessment, outside the work place when appropriate. For any training to be effective, employees and their representatives need to be involved in its planning, testing and revision. In addition to structured training, the employees' representatives should have access to any information which the employer has and which might have an impact on safety either at the work place or in the environment.
Assessment should lead to action
From the employees' point of view, analysis and assessment of risks are without meaning unless there is from the start an understanding that the results will lead to conclusions and action being taken on those conclusions.
Assessment should be followed by (a) determining the priorities and developing a plan of action; (b) implementing the plan and (c) exercising control to verify the results. This does not in everyday life of a work place happen as an orderly movement from one box of an organogram to the next, but more often different phases overlapping, sometimes all at the same time. This, too, should be accepted and noted in the choice of methods.
34 ...... 1 9 9 2
The structures used should be broadly and flexibly organised so as to be able to accommodate the fluid changes in their functions. One single core composition of the body which oversees the whole process is preferable to passing the task from one group to another. For specific tasks requiring particular skills, knowledge or practical experience, experts could be called in, from outside if necessary. Without the link to reduction and mitigation of risks, finding out about them doesn't have a lasting value. There has to be openness about the information produced. Any need for confidentiality is secondary to the safety of humans and the environment.
Various means could be used to simplify wide employee participation in the risk assessment. Fault-free analysis could be transformed into check-lists or questionnaires covering different hazard factors and asking for personal experience on them. The causes of secondary contributing factors to a hazard or risk could be analysed step by step so as to highlight deficiencies in the systems. Work-flow analysis could be used to divide the information on man-machine-environment interaction, in order to better understand the system as a w;ole.
Need for a credible concept
Risk assessment, as any safety related activity sould be based on a concept, such as the protection of humans, animals and physical environment (on-site and off-site) against hazards related to any on-site activity. The concept has to be credible and there has to be a real commitment by all levels of management towards it - and that commitment has to be perceived as real by everyone else. Building risk assessment into the line organisation's normal operating procedure, based on interactive communication and seeking the widest practicable participation of employees and their representatives is the best approach to solid results. The steel mill mentioned in the beginning has recuperated all costs of its analysis, assessment and subsequent changes many times over a brief period through reduced absenteeism.
The relevant authorities should develop support, information, regulations and control mechanisms which are based on the same or similar concept as the activities at enterprise level. Particularly, they should gather, process and disseminate information on new developments concerning substances, processes or technology, or issues like biological hazards.
35 9 XA04NO305 Risk Assessment: An Employer's Perspective K. C. Williams Vice President (Production) Exxon International, USA
Introduction
There is no question that a careful assessment of risk is essential for safe industrial operations. For that reason, a thoughtful analysis of the effectiveness of available risk assessment technologies is a prerequisite for responsible corporate decision making.
An "employer's" perspective on risk assessment cannot be constrained by any artificial restrictions which that term may imply. In reality, all those who are involved in the execution of an industrial enterprise: managers, regulators, the affected public, and especially those employees exposed to hazards, are necessarily partners in the assessment of risk.
The perspective of this paper is that of the oil and gas industry, in which the author's Organisation, Exxon Company, International, participates. The paper addresses what Exxon requires to assess and manage risk in its worldwide operations. The author is aware, however, through contacts with industry colleagues, that some of Exxon's initiatives are representative of similar actions being taken by others.
1992 is the European Year of Safety, Health and Hygiene, coinciding with the United Kingdom's Presidency of the European Council. It is also the year in which new "goal-setting" regulations covering safety in the U.K. offshore oil industry were put for-ward by the Health and Safety Commission. These regulations, based largely on Lord Cullen's recommendations following the Piper Alpha tragedy, set the pace for safety in the British North Sea and will significantly impact the safety of offshore oil installations worldwide. The requirement for risk assessment, using a systematic process of analysing and evaluating risk, is a key component of tis safety regime.
The nature and perception of risk
Risk, broadly defined as "the possibility of harm or loss", pervades all human activities, whether in the businesses we undertake to earn a living or in our recreational pursuits. Some risks we accept voluntarily, like weaving through heavy traffic or darting across a busy street. Many sports purposely incorporate risk taking, where the possibility of harm and the avoidance thereof, often through acquired skill, is seen by some as a pleasurable pursuit.
However, we also face risks created by others, risks not of our choosing, and some of these may arise from industrial activity. When risks are reduced to the lowest practical level that still preserves the essential benefits of such activity, they are usually viewed as reasonable.
The point to make, is that there are very few activities entirely free of risk and informal risk assessment is something we all do constantly, as individuals.
In Exxon, we define risk as the probability of a hazardous situation occurring and the associated consequences or impacts of that situation upon our own and contractors' employees, society, the
36 ...... 1 I...... 9 9 2 environment or physical assets. This means we consider the implications of risk in all that we do. Exxon's long-standing and uncompromising commitment to human welfare has resulted in safety policies and practices that go well beyond passive compliance with existing laws and regulations.
Need for enhanced risk assessment/understanding
Traditionally, industrial risk assessment has included safety engineering, operator training, inspections and other proactive initiatives, as well as industry-wide sharing of the lessons learned from major accidents. The investigation of such incidents has often resulted in better design codes, specifications, operating practices, and new regulations. Exxon's Alaskan oil spill, for example, prompted a re-examination of industry-wide practices and resulted in new regulations impacting the, marine transport of oil.
The increasing complexity of industrial activities and rapidly evolving technologies cause the nature of risk to evolve as well. Risk assessment today demands approaches which systematically and proactively allow the management of hazards at an acceptable level. We can no longer rely solely on traditional practices.
Judgment as to what constitutes an "acceptable risk" depends, however, on the perceptions of the individual or societal group involved. These perceptions can change over time based on individual and company experiences. Other factors influencing the acceptability of risk include whether it is taken voluntarily or involuntarily and who benefits, directly or indirectly, from the activity. Because risk acceptance is an ever-changing "social" process, it must include everyone involved - employees, industry groups, government and relevant sectors of society.
While we cannot totally eliminate risk in our lives, industrial risks can be managed, both in terms of the likelihood of hazardous events occurring and the consequential impacts should they unavoidably occur. A prerequisite to the successful management of risk is an increased understanding of the risks involved; this is the role of risk assessment.
Oil and gas industry risks
The oil and gas production industry is subject to a variety of risks, some familiar and obvious to everyone, others more subtle. In particular, the increasing complexity of operations has introduced the possibility of latent risks, that is, risks which may not be within our experience base nor intuitively obvious. A systematic assessment process can help expose these hidden risks.
Weather and environmental conditions are among the more familiar, externally imposed, risk sources. Oil and gas seem to be found at their most abundant in areas where extreme temperatures, high winds, Arctic ice and rough seas prevail or where deep water and challenging terrain test our ability to operate safely. More sophisticated weather forecasting, better structures and improved safety practices have aided the assessment and management of these sources of risk.
Other externally imposed risks, over which we often have no direct control, include ship/platform collisions and earthquakes. Even here, however, improved platform structural designs are now better able to withstand such events and emergency response capabilities help to mitigate the consequences.
3 7 ...... 9 9 2
Hardware failures, involving breakdowns of materials or equipment, are common and familiar in our daily lives. However, in the oil industry and most other industries they are rarely, by themselves, the root cause of significant safety or environmental incidents. Prudent operating procedures and established engineering practices, based on codes and standards, and past experience, together with safety redundancy, help to assure this.
Studies and incident investigations conducted within the oil industry and elsewhere clearly indicate that the more significant sources of risk to safe and environmentally sound operations involve people. Human errors, which may occur at a worker and management levels, predominate as contributors of incidents. Weather and environmental or hardware conditions may act in conjunction with errors in decision-making, actions taken or other human behavior to produce an incident.
The "Challenger" space shuttle and Piper Alpha tragedies resulted from such event chains and clearly illustrate that we cannot depend on technology and engineering alone to control risk. More attention must be given to continuous safety training, including unannounced drills that thoroughly test the available emergency response capability. In addition, more research needs to be done on critical man-machine interfaces, such as the effect of human sleep-wake cycles and other ergonomic impediments to safe work behaviour. Finally, we must share our research findings with our employees and one another.
People in an operating company and its interfacing organizations work within a "safety management system", whether or not formally identified as such, which defines their standards of safe behaviour. Weaknesses or omissions within this managing system, such as poor documentation of process equipment changes, create potential for latent failures to occur. These may "bait the trap" for the initiation of human errors,
Exxon7s formal risk assessment
In Exxon, we believe the development and use of a more structured, disciplined and comprehensive safety management system, or "Operations Integrity Management Framework", as we call it, assists us in further minimising operational and decision-making errors and expose risks. It allows us to continue managing our increasingly complex business in a safe and environmentally sound manner.
No approach can totally eliminate industrial hazards, however, they can be better managed and the impact of incidents reduced when the risks are fully assessed. There are many methods for achieving this assessment. In Exxon, we see risk assessment as first and foremost a 64questioning" process, which may range from a simple conversation about the risks involved in a routine operational task to more sophisticated investigations of an entirely new production concept. We encourage this questioning of risk at a levels of our operations by designers, suppliers, constructors, operators and managers, to name a few.
38 ...... 1 I...... 9 9 2
The risk assessment process we use is the same irrespective of the analytical tools used to support it (Figure 1). The five questions we ask ourselves are these:
1) What can go wrong with this design, process or operating procedure? That is, let's identify the potential hazards.
2) What sequence of events would cause these hazards and how likely are they to occur? Risk assessment professionals call these "causal and frequency analyses".
3) What are the consequences or impacts of the hazards occurring?
4) How can the causes be prevented?
5) How can the impacts of a hazard, however unlikely, be reduced or mitigated, should it occur?
Prevention of the causes of hazards and reduction of their consequences, the last two questions in our process, represent areas where an operating company can identify and exert the most influence upon risk.
Hardware v. software
Looking first at cause prevention or elimination, this has traditionally been achieved by hardware means. Incorporating tolerance of a wide range of upset conditions into equipment design and specification of materials is typical of this approach.
Hardware remains important. However, experience has shown us that a broad range of
4 6software" issues are also fundamental to reducing risk. These include, among others, selection and training of employees, including supervisors and managers, choosing contractors based on competency and experience, not just price and social-economic parameters, and implementing an effective drug and alcohol policy.
Of primary importance, however, is the implementation of a formal, structured and proactive system of safety management.
Exxon7s "operations integrity management framework"
Reference was made earlier to Exxon's "Operations Integrity Management Framework" or O1MF. This concept clearly defines corporate objectives and expectations for all facets of safety management. It is aimed at incident prevention through effective and pre-emptive management of the causes of risk. Although not driven by the Cullen Report recommendations, Exxon's OIMF is consistent with the safety management principles Lord Cullen outlined following his investigation of the Piper Alpha tragedy.
OIMF consists of eleven principal elements forming a disciplined and structured safety management system (Figure 2 Management, leadership, commitment and accountability is the first element in Exxon's system. It extends right up to Exxon's Board of Directors, where a Public Issues Committee, with both inside and outside directors, has oversight responsibility regarding the company's policies, programs and practices in the areas of safety, health and environment.
39 ...... 1 9 9 2
The management leadership element provides the driving force for other elements covering systems, practices and procedures in the following areas:
- risk assessment and management - facilities design and construction - process and facilities information and documentation - personnel and training - operations and maintenance - management of change - third party (contract) services - incident investigation and analysis - community awareness and emergency preparedness
The last element, operations integrity assessment and improvement, encompasses the critical need for regular reviews of operational performance relative to the safety, industrial health and environmental expectations outlined within the OIMF. These assessments are conducted by multi- disciplinary teams which include expertise from outside the organizational unit under review.
Each of these integrity management elements must have clearly defined objectives, responsibilities, standards and procedures, as well as an ongoing process of self-assessment and self-improvement to ensure that objectives are being met. These goals and requirements are determined by the organizational group involved and the level of assessed risk. Simply stated, Exxon's OIMF is not an inflexible corporate directive. It is a framework within which different business and operating units can address their particular risks with different priorities and action plans.
The ultimate goal of Exxon's OIMF is to make the identification of every kind of risk, and proactive responses to them, an integral part of day-to-day business throughout the organization. This process of continuous improvement will help achieve "operational excellence" - leading to better managed, better controlled, more efficient and, ultimately, more profitable operations.
Emergency preparedness
The final question in the risk assessment process is one that must be addressed regardless of how many risk prevention measures have been implemented. When safety or environmental incidents occur, despite sound risk management, how can the impacts be reduced or mitigated?
Emergency preparedness is always necessary and Exxon spends significant time and resources to be ready to respond to unexpected operational situations involving safety of personnel or the public and protection of the environment. Our emergency response training has been developed to include scenario-based drills, which make use of escalating incident conditions and breakdowns in the command structure. It involves participation by external agencies such as public fire services, coast guard, industry oil response teams and others. Exxon also has access to emergency supplies and expert personnel, who are immediately available from around the world to assist in an emergency situation.
40 ...... 1 9 9 2
Risk assessment tools
It should be clear that within Exxon, the act of "questioning" is seen as fundamental to any risk assessment process. However, numerous and increasingly powerful tools are available to address the questions posed in the risk assessment process. Correctly applied, these tools can effectively supplement, but not replace, what we learn from direct "hands-on" human experience.
Exxon makes use of a suite of risk assessment tools, each with its own strengths and limitations. Hazard identification is carried out using design reviews, checklists, Hazard and Operability Studies HAZOP) and Failure Mode and Effects Analysis (FMEA). Mathematical models are used to examine hazard consequences, such as fire impacts, explosion overpressures, the fate of oil spills and wave loadings on platforms. Event and Fault trees are employed to determine hazardous event sequences. Risk matrices are developed to examine specific risk scenarios. Operations that benefit from the use of numerical risk estimates utilise quantitative risk assessments. Each of these tools has a role to play in the risk assessment process.
Quantitative risk assessment
Today, there is much discussion of the use and value of Quaw Lative Risk Assessment QRA) often to the exclusion of other methods, especially those of a more qualitative nature. Exxon's experience with a broad range of worldwide operations indicates significant value in utilising a diverse spectrum of risk assessment techniques. Indeed, it appears unsound to rely only on one methodology.
The discipline imposed by QRA is its greatest benefit and Exxon supports expanded use of the technique within the oil industry. QRA also provides a means to compare alternative risk management strategies. However, the exclusive use of QRA as an absolute measure of risk is suspect at this time. We share the common concerns about the quality of data available to input to QRA tools and the caution that is required to ensure that reality is reflected in the analysis.
However, these concerns should not prevent effective use of this tool for comparison studies. To improve future analyses, Exxon participates, with other members of the oil industry, in the development of improved reliability and failure rate data bases.
Qualitative risk assessment
As previously indicated, Exxon strongly encourages the questioning of risk at all levels of its operations. This includes a range of assessment tools, many of an intuitive nature, such as operational checklists and "hazard hunts" at operating sites. These are needed to reflect the capabilities and experience of the personnel involved. Furthermore, care must be taken not to limit involvement in risk assessment to only those persons with specialized risk analysis skills. Such a limitation would undervalue the intuitive input of experienced platform employees and installation managers, and subvert the original intent to broaden the understanding of risk.
There can be no "black boxes". People at all levels need to be fully involved in risk assessment. Risk understanding, which is crucial to the successful management of risk, comes from such direct involvement. In practice, this means providing opportunities for workers to participate in such activities as the development of design basis documents, the design of operating procedures, the determination of safety controls and the ongoing daily assessment of hazards in their work place, among many others.
41 ...... 1 9 9 2
Whatever risk assessment techniques are used, it is important to realize that they are just tools and no more. They cannot make decisions. They can only support the judgment of those individuals who do make decisions - operational employees, supervisors and management, as well as the societal judgments of regulators and the public. In Exxon, we enhance the capability of our staff to apply sound judgment to risk management decisions through extension formal training and assignment rotations which expose them to many different types of risk scenarios worldwide. Societal risk assessment
Risk management decisions are also strongly influenced by other considerations, such as social and corporate values and the risk perceptions of employees and the affected public-. In order for individuals or society, through regulatory agencies or direct public hearings, to agree on an acceptable level of risk, they must first understand the risks involved. As observed earlier, the assessment of risk is a social process and one that must, necessarily, involve everyone who is exposed - directly or indirectly - to the risk in question.
The allocation of adequate time to conduct a risk assessment and the availability of credible data are also essential. Rushing trough a risk assessment or seeking to use the results as an absolute measure of risk, when this is clearly not justified by the quality of the input data, does little to build commitment to risk management solutions. Summary
• Industrial risks cannot be eliminated, but they can be assessed and managed at levels acceptable to individuals and society.
• Risk assessment is an essential prerequisite to risk management and demands a clear understanding of the risks involved. This means supplementing our knowledge of past incidents with an assessment process which enables potential hazards to be uncovered or predicted.
• Exxon believes this process should involve intense questioning of risk at all organizational levels, and is best supported by a diverse blend of quantitative and qualitative techniques. Many of these techniques should be relatively unsophisticated, so they can be understood by a wide range of people with firsthand knowledge of risk exposures.
• Where the use of rigorous quantitative techniques is necessary, it is essential that people with a broad spectrum of differing skills be fully involved. This assures understanding, validity and shared ownership of both the results and the process by which they were achieved.
• Human errors, occurring at all worker and management levels, are often enabled or initiated by failures and omissions in the underlying safety management systems. Prevention of these system breakdowns may be enhanced by greater attention to non-hardware, "people" issues and implementation of more structured, disciplined and proactive approaches, such as Lord Cullen's Safety Management System and Exxon's Operations Integrity Management Framework.
• Lastly, improving our assessment and management of risk in no way reduces the need for well prepared emergency response plans, which must be regularly tested through realistic scenario-based drills. Mitigating the consequences of hazards, foreseen or unforeseen, is essential to achieve the lowest practical level of risk to individuals, society and the environment.
42 ...... 9 9 2
4
z cn 0 n 0 w < n z Fn - cnD >-I - :)W i WZ (n < < a < > < z w Z Ww Er < 112 U- a.
z z C) U) O 0 0 z >LU z< a.F- < < < C U)z LIJ II N LL LI) z < p M rr MZ w < C) LLI U) w F- Lij Cl) < Cc
z l) 0 w Dw cc a D w < U) (1) z p w< < 0 C) z U) 0 U) >-, -J U) 43 ...... 9 9 2 w > w 0 Z LLI W4 w cc w LU (-) a- >< 0 0 w 06 I-- z C/) F- Z w z z w >- LL, 0 EE - U2 :>LLI U) 0 C/) cr D LL Er LLI CLLU --W EL cn C) Z u):E cn w C/) 0 LLI - CL LU C) cc 0 2 ::) z LU C[: 0 LU (f) o (D F- C) CO 0 < U) < Z LU C-6 06 U L) W z z U) - < (.D z (f) Z Z Z z < 0 LU LJ U) 0 U) :E (-) F--- z w < LU - LL] U 0 w 00 .6 -I -r U Z M X F- - z = - < LLJ < X C/) 0 C) ZZ ;;z C) :> (D c a- LLI UJ >. 7' (-D < 0 Cr LL CE < W w - U- 0 LLI Cc: CC Cf) od I--- U) Lu n Lu O z -6 U) 0 Z 0 U) z > CO < - L-Ljj :;, Lu f- 7, b w U) F- z 0 .5 Cc: z LLI (/) LU U) < Z LLJ < Z Lu Q U) U) 2i (D Z ::) (-D w 0 LU :E rr 0 < C) m (f) < ccC) Y 0 0 CC Lu Z U) rr LL w r) < =i= C) 0 ir L 06 F-- 0 z- =F LLI < LU C/) E > W rr I-- (.D Lu 2 =) < r--j E 0 Z < < LJ.J C) < 44 9 XA04NO306 Risk Assessment - the Perspective and Experience0 of US Environmentalists Dr Ellen Silbergeld Professor of Pathology, University of Maryland, Baltimore, USA Introduction Since 1981, risk assessment has formed the methodological basis for much public policy related to occupational and environmental chemicals in the US. Risk assessment rose to prominence out of public concern over the potential contribution of chemical exposures to cancer and out of public frustration with delays in regulation by the Environmental Agency (EPA) and the Occupational Safety and Health Administration (OSHA). In addition, pressures from industry required EPA and OSHA to provide a scientific rationale for specific regulatory decisions. The past decade provides a convenient body of evidence upon which to consider the value of risk assessment as a method for reaching public policy decisions. This paper will provide such an evaluation; from the environmentalist perspective. My criteria include: the efficiency, adequacy, clarity, enforceability, and public acceptability of regulation during this period. In addition, the scientific validity of risk assessment is of concern to environmentalists, as it is to others. Risk Assessment: the early years In 1979, an interagency committee of the US government proposed guidelines for the identification and assessment of chemical carcinogens (see OTA, 1987 for a history of carcinogen risk assessment in the US). These have served as the source of subsequent science policy for environmental regulation in the US since that time. Prior to 1979, regulation of toxic chemicals had generally been based upon one of three principles: technology; risk: benefit balancing; or banning. The technological basis for regulation was predicated on the availability of control technology (Portney, 1990) or the limits of analytic chemistry (as in the Safe Drinking Water Act) (Freeman, 1990). However, a technological basis is inherently unsatisfactory, when it limits risk reduction by currently available technology. No further action can be taken in those cases where relatively high levels of pollution remain even after controls are imposed, or where analytic chemistry is insufficiently sensitive to measure contamination at the point of release. The risk: balancing approach to regulation, embodied in statutes regulating pesticides, incorporates an assumption that such a balance could be found. Put in other words, this assumed that there were levels of exposure where risks were very low, or nonexistent, such that the benefits of continued use were clear. In the early 1970s, it was generally assumed that pesticides were nontoxic to nontarget species at some dose, either because of the nature of their toxic effects (socalled threshold effects), or because of qualitative differences in species response (socalled selective toxicity). Some statutes empowered regulators to use the option to ban the production or use of a substance entirely based solely upon the finding of hazard. The Delaney amendments to the Food, Drug and Cosmetic Act, sections of the old Clean Air Act, and FIFRA permitted agencies to prohibit releases and uses entirely without the need to calculate the extent of the hazard, which is the function of risk assessment. 45 ...... I 9 9 2 The convergence of new statutes, new public health concerns, and new developments in science in the mid-1970s caused many environmentalists, and others, to reconsider the appropriateness of these approaches to regulating all types of toxic chemicals. The new statutes, such as the Toxic Substances Control Act, required more explicit risk balancing; new initiatives, by EPA and OSHA to regulate a broader range of industrial chemicals, provoked strong response from the regulated community; and scientific developments supported a central role of mutagenesis in the induction of cancer (OTA, 1987). Moreover, public concern over cancer heightened by President Nixon's declaration of a "war on cancer" in 1971, coinciding with the creation of the EPA and OSHA. One element in the "war on cancer" was disease prevention; after considerable controversy, the _public health community had succeeded in translating the scientific consensus on the role of cigarette smoking in carcinogenesis (Doll and Peto, 1981). The search for other identifiable chemical etiologies of cancer was advanced by the development of relatively simple assays for mutagenesis, including the Ames test. Prevention, as part of public health policy, assumes the obligation to act prior to the induction of disease or death. To prevent chemical- induced disease, action must be taken prior to information on human response. The 1979 proposals on risk assessment were thus premised on the following: chemical exposures were a significant contribution to overall incidence of cancer; preventing chemical-induced cancers required identification of potential human carcinogens prior to human exposure on such a scale that epidemiological studies could demonstrate plausible associations between exposure and disease; the evaluation of potential carcinogens should be quantitative as well as qualitative, in order to base rational regulations under the risk balancing statutes (OTA, 1987). Debate continues on the quantitative contribution of chemical exposures to human cancer (see, for instance, Perera and Boffetta, 1988; Davis, et al, 1990; Doll and Peto, 1981). It is unlikely that epidemiological studies can completely resolve this issue, given problems in exposure assessment and the likelihood of complex interactions among chemical exposures, genetics and socalled lifestyle factors (including smoking and diet). Methods for the identification of carcinogenic chemicals have been applied for over a decade, using animal models, structure: activity analysis, and sort term in vitro tests. As this information increased, regulatory agencies needed criteria for judging the results of these tests and assays, examples of which can be found in the decision rules of the US National Toxicology Program (NTP) and the International Agency for Research on Cancer (see Huff, et a, 991, for a discussion of these rules and their science base). However, as discussed above, for regulatory purposes, qualitative methods were insufficient to support regulation, particularly after the US Supreme Court struck down OSHA's benzene standard because of an insufficient science based rtionale for risk assessment and reduction. From 1979 to 1987, in the US regulatory agencies developed more specific principles upon which to base quantitative risk assessment for chemical carcinogens. These principles were bounded by several important policy decisions that were generally accepted in the 1980s: first, they were to operate in the absence of human data, relying upon results from experimental research; second, they were to provide quantitative estimates of dose at level of risk that were deemed politically acceptable; and third, they were to be presumptively conservative, tat is, protective of human health in the case of uncertainty. These are policy decisions. They have largely shaped the continuing debate over quantitative risk assessment in the US (see, for instance, Silbergeld, 1991). 46 ...... 1 9 9 2 Risk Assessment in Theory and Practice, 1980-1990 Although risk assessment has been utilized widely in US policymaking since 980, the environmental community remains divided over its acceptability as a policy tool. For some the mere use of risk assessment is an admission that some amount of risk i§ acceptable; to these and other advocates, te imposition of any risk is unlawful under certain statutes (such as the old Clean Air Act) and unethical in most circumstances Doniger, 1978). This is a powerful argument, and the American tradition of rejecting unequal treatment under the law challenges the allocation of individual risks in a way that inevitably affects some person(s) more tan others (and others not at a, given the strict interpretation of risk assessment as a probability estimate). Other environmentalists (including this author) have supported the appropriate use of risk assessment for its potential to increase the efficiency of government response in preventing disease. Environmentalists ave generally supported a strict interpretation of risk assessment guidelines at te level of hazard identification (e.g., the criteria that support the designation of a chemical as carcinogenic in animals) and have resisted modifications in the methods for dose: response calculation and dose: exposure extrapolation from those that build in elements of conservatism. In theory, environmentalists have strongly argued for continuing the qualitative assumption that animal carcinogens, under any circumstances of dosing, are probably human carcinogens. The most controversial provisions of quantitative risk assessment are the default assumptions for quantitating exposure and risk at low levels, in the range generally accepted by policymakers (that is, between 1:10,000 to 1:1,000,000). The major default assumption is based upon the molecular biological theory of cell mutation, the socalled single-hit theory of mutagenesis. This assumption is the basis for the linearized multistage model of arcinogenesis, which provides the rationale for the statistical approaches used to derive unit risk estimates from experimental dose: response data (Huff, et al, 1991). Because policy decisions require estimation of dose beyond the range of feasibly obtainable experimental or human data, inference rules must be used to extrapolate to the range of concern. Since 1980, the EPA has relied upon the assumption that carcinogens at molecular doses increase the probability of cancer by some amount greater tan zero, and that in the low dose range increments of dose are associated with proportional increases in risk (OTA, 1987). It is admitted that this assumption, being generic, does not necessarily reflect chemical-specific mechanisms of action. Environmentalists and others have defended the default assumptions, and the rules that cover extrapolation from experimental data to estimated human dose, to provide conservative buffers for uncertainty and possible variation in sensitivity within the human population (Perera and Boffeta, 1988; Huff, et al, 1991). Critics, particularly those who generally favour less stringent or frequent egulation, have objected to the general application of these default rules (Nichols and Zeckhauser, 1986). They have challenged the relevance of animal data for particular types of carcinogens on the basis of purported species differences in metabolism, target tissue response, or inherent sensitivity; they have objected to the extrapolation principles, such as dose scaling; and they have rejected the general applicability of the linearized model to all carcinogens (Ames, et al, 1987). Over the past 7 years these objections have arisen in the context of proposals to regulate specific chemicals, such as methylene chloride, formaldehyde, perchloroethylene, and the dioxins (Robinson and Paxman, 1992). In no instance have the critics presented compelling scientific evidence to justify alternate approaches to risk assessment. Critics have also objected to the bias towards conservatism that operates in risk assessments (Nichols and Zeckhauser, 47 ...... 9 9 2 1986); however, as pointed out by Finkel 1989) and Allen et al 1990) it is not clear that these rules always result in gross overestimates of risk. In principle it is unreasonable to object to the admission of "better science" in risk assessment. Relevant data on specific chemicals may be very useful in judging the applicability of the general default rules. The problem is that we do not really know what these data are, or how we would incorporate them into alternate methods of risk assessment. The critics of risk assessment more often than not are in disagreement with the hazard identification stage, and they do not accept the NTP bioassay as a reliable method of identifying carcinogens (Ames, et al, 1987; Huff, et al, 1991). It is also not always recognized that risk assessment is primarily a policy tool, not the complete method of testing scientific hypotheses. Public health policy and clinical medicine have similar relationships to basic science: both are practical disciplines that demand, at some point, action in the face of residual uncertainties. Has risk assessment improved public health? This question may be answered in several ways; I shall focus on the perspective of process. Has the use of risk assessment improved policymaking, in terms of efficiency, speed, and acceptability? The record over the past decade does not support a positive view, although it is difficult to distinguish problems in the process from political interferences (Lash, 1984). Despite the confidence of some that risk assessment could be insulated from the political and economic parts of policymaking (Ruckelshaus, 1984; also NRC, 1984), it is in practice impossible to separate risk assessment and risk management from each other (Silbergeld, 1991). It would be difficult to claim that the US process of policymaking is efficient or expeditious. One problem with the adoption of risk assessment as a policy tool is the difficulty in terminating data accumulation, even temporarily. While research continues it does not reach incontrovertible certainty. Risk assessment per se provides no guidance as to how much information is sufficient for decisionmaking. For most of the chemicals of great public concern in the 1980s - lead, dioxin, formaldehyde, Alar, ethylene dibromide - no final regulatory action has resulted after years of arguments over data and data analysis (Robinson and Paxman, 1992; Roach and Rappaport, 1990). Risk assessment can encourage the use of research as a delaying tactic, given the high stakes of most regulatory action in the US. Because it is impossible for EPA to issue interim rules for a particular chemical, each step in the process has increasingly higher stakes for all parties concerned. The history of EPA dealings with the dioxins is illustrative. By the end of the 1970s, EPA had concluded its first evaluation of the health and ecological hazards of 2 3 7, 8-tetrachlorodibenzo-p-dioxin (TCDD) in connection with regulations under the Clean Water Act. However, in 1981, this process was suspended after an intervention by the Dow Chemical Company with Dr. John Hernandez, then Deputy Administrator of EPA In 1983, the second wave of environmental controversies over dioxin erupted, with the revelations of contamination in communities in Missouri, Illinois, and New York (Reggiani, 1980). The strong provisions of the hazardous waste cleanup statute, Superfund, required EPA to take action at these sites, but it did not have a standard. The agency's hand was forced by the Centers for Disease Control, whose scientists used risk assessment principles to calculate public health guidance for evacuation of contaminated communities (Kimbrougb, et al, 1985). 48 ...... 1 9 9 2 EPA reopened its process of risk assessment for the dioxins shortly afterwards, and in 1986 the first health assessment document (HAD) for TCDD was published. Using standard methods of quantitative risk assessment, the EPA proposed an extremely low unit risk for lifetime exposure to TCDD 0016 fg/kg/day (a unit risk is the lifetime daily dose associated with an increase in cancer risk of 1:1,000,000). This HAD immediately raised political controversies as more and more dioxin-contaminated sites were discovered, in New York and New Jersey. In addition, it was found that municipal waste incinerators could be sources of dioxins from the incomplete combustion of certain precursor materials in the presence of halogens (Denison and Silbergeld, 1988). The Environmental Defense Fund petitioned and later sued the EPA, under the Toxic Substances Control Act, to regulate the ongoing releases of dioxins and related compounds from a range of sources, including chemical production, industrial waste disposal, and solid waste incineration. An important part of EDF's petition was the novel proposal to treat the dioxins and furans as a class of similarly toxic chemicals, based upon structural fit to the socalled dioxin or Ah receptor. This proposal has been adopted by many national regulatory agencies, in the socalled toxic equivalency factor approach. The inclusion of municipal incinerators as sources of dioxins and furans was controversial, and regulation was delayed as EPA undertook another assessment of the risks of dioxin in 1988. EPA's first proposal was to average the risk estimates produced by its own scientists, those at CDC, and at other national agencies (Finkel, 1988), but scientific objections to this method encouraged the agency to adopt a more science-based approach. The conclusions of that review were basically a restatement of the 1986 risk assessment. However, a new political obstacle to regulation under the 1988 risk assessment was the revelation that pulp and paper mills could also generate substantial dioxin and furan inputs through the use of chlorine bleach processes. The paper industry, facing very strict regulation on its discharges and waste disposal practices (including land farming of its solid waste), forced EPA into another reconsideration of the risk assessment. The third risk assessment reevaluation is ongoing t the EPA at the present time (April 1992). Enormous pressure has been brought on regulators and scientists involved in this activity. The chlorine industry contributed funds for the convening of a scientific meeting at the world renowned Banbury Center in October 990 a public relations firm retained by the industry then disseminated misleading information on the nature and conclusions of this meeting (see Bailey, 1992). This process of assessment and reassessment has not encouraged public confidence in risk assessment. Moreover, the highly sophisticated use of risk assessment methodologies by the waste management industry, seeking permission for siting its incinerators and landfills, has increased public suspicion of the validity of the numbers because of the ease with which a large and profitable industry of consulting risk assessors can produce a growing pile of documents purporting to estimate the precise risks of such facilities. Public suspicion has often been attributed to public ignorance of the nature of risk and the technical steps of risk assessment (Davies, et al, 1987; Viscusi and Zeckhauser, 1992). If this is the source of the public's reaction, then it is important to consider how, and whether, this can be overcome. Investing government decisionmaking in a process inaccessible to public understanding is contrary to principles of American government (Ruckelshaus, 1985). On the other hand, the public may not be so much ignorant as it is cannily skeptical of a method that does not seem able to generate stable estimates of risk. If this is the case, unless the method can be greatly improved and a broader consensus generated as to its scientific validity, then it will 49 ...... 1 9 9 2 not be possible to persuade the public to believe something that industry, environmentalists, and academic scientists periodically attack in both theory and practice (Silbergeld, 1991; Zeckhauser and Viscusi, 1990, Finkel, 1989 Ames, et al, 1987). Alternatives to Risk Assessment The drive to risk assessment, as noted above, arose from several sources, including scientific advances, pressures upon agencies to specify risk estimates more precisely and with greater documentation, and the desire of the public for more decisions on toxic chemicals. The practical application of risk assessment does not seem to have answered these needs: risk estimates may appear to be more precise, but very few - industry, regulators, scientists, or the public - actually believe their precision. Scientific advances in our knowledge of carcinogenesis make the process more complicated and open up more possible models to guide extrapolation. Finally, the last decade has seen fewer regulations rather than more, as compared to the 1970s. There are alternatives to risk assessment. First, we could return to the technology-based or the banning approach to regulation. Both of these approaches, as discussed above, are based upon a qualitative finding of risk (hazard identification), which triggers either the application of best available control technology or an outright ban or restriction upon use. Commoner 1990) has argued that only bans have significantly reduced environmental risks, citing the consequences of EPA actions to ban DDT and other organochlorine pesticides, and the drastic reductions in the allowable use of lead in gasoline. International agreements to ban the production and use of ozone-depleting chemicals are more recent examples of this approach, as is the OECD experiment in multinational approaches to risk reduction for socalled "sunset" chemicals, whose risks are sufficiently great that no quantitative calculation of risk is necessary to justify concerted action. Second, we could adopt simpler rules for risk estimation. The approach used in the Netherlands, to apply a safety or uncertainty factor to all types of toxicants, avoids the need to select mechanism-based approaches to classes to toxicants, based upon endpoint and assumed mechanism of action. An advantage of this approach is that it restores priority for chemicals that may not be carcinogenic, but highly toxic for reproduction of the nervous system; arguably, these types of chemical risks have been overlooked in the US in the past decade (for neurotoxins, see NRC, 1992). However, this approach does not reduce the problems of determining levels (no-effect or lowest observed effect) to which to apply safety or uncertainty factors (Roach and Rappaport, 1990). Third, we could utilize novel tools of risk reduction, that avoid the burden of setting point estimates for standards and guidelines. The approach embodied in California's Proposition 65, largely written by EDF, sidesteps the entanglements of full risk assessment as the path to reaching specific standards. Risk assessment is used to trigger disclosure provisions, rather than to specify control actions. Under Proposition 65, industries and other sources must disclose to the public when they release or otherwise expose people to chemicals known to cause cancer or reproductive effects in humans or animals. Te requirement of disclosure is different from an overt, enforceable requirement to bring these exposures down to specific levels below which a specified increase in risk is estimated to occur. A recent review of implementation of Proposition 65, by California's state government, concludes that it is an efficient and productive mechanism for risk reduction (Book, 1992). Several examples of product reformulation, to avoid disclosure, have already occurred (Smith, 1990). 50 ...... 9 9 2 A similar approach operates through the Toxics Release Inventory (TRI) in the US. TRI requires industrial sources to report annually their releases of toxic chemicals into air, water, and surface environments. This provision in the Superfund amendments, written by environmentalists, has elicited risk reduction measures by industry in the absence of point estimates of risk on a chemical-by-chernical or release-by-release basis. Because of industry reaction to disclosures, EPA established the socalled 33/50 initiative to reduce release of high-priority toxic chemicals by 50% from 1990 to 1995 (EPA, 1992). The success of this program depends upon two factors: first, the disclosure provisions of the TRI program are mandatory and accessible to computer- literate citizens; and second, most releases are generated by a small and identifiable segment of industry (see Table 1). About 10% of reporting facilities release more than 75% of the high priority toxic chemicals (EPA 1992). If the goal of halving these releases within years can be accomplished without specific regulations, then the need for quantitative risk assessments will be greatly reduced. It remains to be seen te extent to which such innovative approaches can be used in other areas of risk reduction. Summary The rise of risk assessment in the US was supported and encouraged by environmentalists in the late 1970s. However, in practice, risk assessment has not proved to expedite regulatory actions to reduce risks, and public acceptance of the techniques of quantitative risk has decreased over the past decade. A review of the paralysis induced in EPA by TCDD exemplifies the scientific and political problems with risk assessment. In addition, the exclusive focus of risk assessment upon cancer has probably resulted in lost opportunities to reduce human exposure to other types of toxic substances, and, by implication the incidence of noncancer disease and disability. Alternatives to the use of risk assessment include a return to technology-based or "pure" risk approaches, where quantitation of risk is not mandatory. The new Clean Air Act of 1990 to some extent exchanges the "pure" risk approach for technology-based regulation, with more specification of technology-forcing (Robinson and Paxman, 1992). Although no political consensus has been reached to repeal the remaining "pure" risk provisions of the Food, Drug and Cosmetic Act, it is unlikely that extending a "pure" risk strategy for industrial chemicals would be politically popular. Another alternative is the use of risk assessment as a semiquantitative signal for action, using disclosure as a means of stimulating action without specification of the precise standard to be reached. This approach has clearly resulted in relatively rapid response by industry, to avoid disclosure, but it is ultimately limited in its efficacy to reduce risks. There are limits to the impact of information on human behavior (Viscusi and Magat, 1987), and there may be cases where semiquantitative actions will not reduce risks to the level deemed acceptable. Attempts to reduce exposure to tobacco smoke may have reached their plateau of efficacy in the US. Nevertheless, this approach is preferable to the handing over of regulatory decisionmaking to a technological elite, whose members alone can understand the increasingly arcane basis and data analysis of quantitative risk assessment (Ruckelshaus, 1985; Silbergeld, 1991). Environmentalists today as in the past (Hays, 1987) accept the challenge of integrating public health goals into broader commitments to a just and workable democratic society. 51 ...... 1 9 9 2 Table 1. The Toxics Release Inventory in Action: Effects of Disclosure upon Industry Behavior* Cheniicals Targetted for Reductions Percent of total releases by Under the 33/50 Program Companies in 33/50 Program Percent benzene 73 cadmium and compounds' 49 carbon tetrachloride 90 chloroform 88 chromium and compounds 42 cyanides 79 lead and compounds' 60 mercury and compounds' 93 methyl ethyl ketone 56 methyl isobtyl ketone 62 methylene chloride** 62 nickel and compounds 59 tetrachloroethylene 56 toluene 45 trichloroethane 43 trichloroethylene 43 xylenes 51 *From EPA, 1992. **Chemicals also identified in OECD "sunset chemicals" risk reduction program, 1992. 52 ...... 1 9 9 2 Acknowledgements The research in this paper was supported in part by grants from the Robert Wood Johnson Foundation and the Charles E. Dana Foundation, to the author while a member of the Toxic Chemicals Program of the Environmental Defense Fund. The author thanks Karen Florini, David Roe, and Robert Yuhnke of EDF, and Kevin Tonat of Johns Hopkins, for continuing opportunities to discuss these and other issues related to risk assessment. References 1. Silbergeld, E. K. 1991) In: Mayo, D., and Hollander, R. (eds) Acceptable Evidence: Science and Values in Risk Management, New York: Oxford University Press, pp 99-114. 2. Ruckelshaus, W. 1985) Issues in Si. Technol. 1: 19-38. 3. Office of Technology Assessment 1987) Identifying and Regulating Carcinogens. Washington, DC: US GPO. 4. National Research Council 1992) Neurotoxicology: Identification and Risk Assessment. Washington, DC: NAS Press. 5. Nichols, A., and Zeckhauser, R. 1986) Adv Applied Microecon 4 55-82. 6. Lash, J. 1984 A Season of Spoils. New York: Grossman. 7. National Research Council 1984) Risk Assessment in the Federal Government: Managing the Process. Washington, DC: NAS Press. 8. Hays, S. 1987) Beauty, Health, and Permanence: Environmental Politics in the US 1955-1985. Cambridge: Cambridge University Press. 9. Finkel, A. 1989) Columbia J. Environ. Law 14: 427-467. 10. Finkel, A 1988) Risk Analysis 8: 161. 11. Portney, P. R. 1991 In Portney, P. R. (ed) Public Policies for Environmental Protection. Baltimore: Johns Hopkins University Press, pp 27-96. 12. Freeman, A. M. 1991) In Portney, P. R. (ed) Public Policies for Environmental Protection. Baltimore: Johns Hopkins University Press, pp 97-149. 13. Davies, J. C., Covello, V., and Allen, F. (eds) Risk Communication. Washington, DC: Conservation Foundation, 1987. 14. Viscusi, W. K., and Magat, W. A. 1987) Learning about Risk: Consumer and Worker Responses to Hazard Information. Cambridge MA: Harvard University Press. 15. Ames, B. N., Magaw, R., and Gold, L. S. 1987) Science 236: 271-277. 16. Zeckhauser, R. J., and Viscusi, W. K. 1990) Science 248: 559-564. 17. Bailar, J., et al 1988) Risk Analysis 8: 485-497. 18. Allen, B., et al 199?) Risk Analysis 8: 531-542. 19. Doniger, D. 1978) Ecol. Law Quart 17: 233-316. 20. Roach, W. S., and Rappaport, S. M. 1990 Amer. J. Ind. Med. 17: 727-753. 21. Robinson, J. C., and Paxman, D. G. 1992 Amer J. Ind. Med. 21: 383-396. 22. Book, S. A. 1992) Testimony before Senate Committee on Governmental Affairs, March 27, 1992. 23. Bailey, J. 1992) Wall Street Journal, February 20, 1992. 24. Smith R 1990) Wall Street Journal, November 1, 1990. 25. US EPA 1992) EPA's 33/50 Program Second Progress Report. Washington: US EPA, February 1992 (TS 792A). 26. Huff, J., Haseman, J., and Rall, D. 1991) Ann Rev Pharmacol Toxicol 31: 621-652. 27. Perera, F., and Boffetta R 1988) JNCI 80: 1282-1293. 28. Doll, R., and Peto, R. 1981) JNCI 66: 1191-1208. 29. Davis, D. L., et al 1990) Lancet 336: 474-481. 30. Commoner, B. 1990) The Closing Circle. 31. Kimbrough, R. D., et al 1985 J Toxicol Environ Health. 32. Denison, R., and Silbergeld, E. K. 1991) In: Travis, C. and Hattemer-Frey, H. (eds) Municipal Waste Incineration Risk Analysis, Boca Raton, FL: CRC Press, pp 275- 293. 33. Reggiani, G. 1980) In: Kimbrough, R. D. (ed) Halogenated Biphenyls, Terphenyls, Naphthalenes, Dibenzodioxins, and Related Compounds New York: Elsevier, pp 303-371. 53 ...... 1 9 9 XA04NO307 The Transformation of the Industrial Economy into a Service Economy: The Function of Technology and the role of Insurance in the Management of Risk Professor Orio Giarini Secretaire General The Geneva Association, Switzerland 1. The transformation of the Industrial Economy into a Service Economy: The function of technology and the role of insurance Technology has been at the core of the Industrial Revolution, by producing new materials, new products and processes and by increasing manufacturing productivity by means of specialization and economies of scale (including higher production's speed). A key difference between the industrial economy and the service economy is that the first one gives value essentially to products which exist materially and which are exchanged, whereas value, in the service economy, is more closely attributed to the performance and real utilization (in a given time period) of products (material or not) integrated in a system. Whereas during the classical economic revolution, the value of products could be identified essentially with the costs involved in producing it, the notion of value in the service economy is shifting towards the evaluation of costs incurred with reference to obtained results in utilisation(l). The real change towards the service economy stems precisely from the fact that services are becoming indispensable to make products and services fulfilling basic needs. Services are no longer simply a secondary sector, but they are moving into the focus of the action, where they have become indispensable production tools to satisfy the basic needs and to develop the wealth of nations. The insurance industry is a typical example: until two decades ago, everybody, including people in the insurance industry, accepted that insurance policies covering e.g. life risks, or material damages, were a typical secondary product in the traditional economic sense and that they could only expand once the basic needs were satisfied by material production. However, during the years following 1973, when the growth to GNP in the world dropped from an average of 6 to less than 3 % per year, the overall sales of policies continued to grow at I % or 2 % per year above the growth of GNP. If insurance consumption were of secondary importance, the slow down in other activities and in particular in manufacturing would have produced more than proportional reduction in the sales of insurance, according to Engel's law. Considering the economy as a "service" economy also allows to better appreciate the contributions made by contemporary technology: the latest technological advances have their ...... * Secretary General and Director of the "Geneva Association" (International Association for Risk and Insurance Economics Research). (1) See Orio Giarini and Walter R. Stahel, "The Limits to Certainty - Facing Risks in the New Service Economy", Kluwer Academic Publishers, 1989, 164 pp. 54 ...... 1 9 9 2 greatest impact on systems concerned with the communication and organization of information, which is exactly what is needed in order to better manage the development of the present day economies. All this is quite different from the direction which technology had taken during the classic Industrial Revolution, when all that appeared to matter was how to investigate and improve the stages of production which transformed raw materials into finished products. 2. Uncertainty and te vulnerability of systems The notion of system becomes essential in the service economy. Services produce positive results or economic value when they function properly. The notion of systems (or functioning) requires to consider real time and the dynamics of real life. And whenever real time is taken into consideration, the degree of uncertainty and of probability which conditions any human action becomes a central issue. Unfortunately, the notion of vulnerability is generally misunderstood. To say that vulnerability increases parallel to the increase of the quality and performance of modern technology might seem paradoxical. In fact, the higher level of performance of most technological advances relies upon a reduction of the margins of error that a system can tolerate without breakdown. Accidents and management mistakes still happen, even if less frequently, but their effects have now more costly systemic consequences. Opening the door of a car in motion does not necessarily lead to a catastrophe. In the case of a modern airplane, it will. This shows that systems functioning and vulnerability control become a key economic function where the contributions of economists and engineers must be integrated. In a similar way, problems of social security and savings for the individuals have to take into account vulnerability management at a personal level. - Thus, the notions of risk and management of vulnerability and uncertainty become characteristic of the service economy. 3. The notion of risk in the Industrial Revolution and in the Service Economy: Entrepreneurial (commercial) versus Pure Risks Modern technology has been at the source of the increasing benefits but also of increasing risk management problems in many ways. In particular, there is a shift of emphasis from the traditional entrepreneurial risks to pure risks of the "insurable" type. Risks today are becoming concentrated at levels where the vulnerability is such that the overall uncertainty of the economic process increases. The problem of environmental hazards, which very often is linked with the question of transportation and storage of dangerous materials is part of the same type of risks and vulnerabilities that our modern society has to face. The risk taking attitude was not studied in detail by the first great economists: it was rather taken for granted in a given cultural environment. A widespread process is now going on in economics for reconsidering some basic concepts, where the fundamental point is the need for a better understanding of the conditions and reasons for modern economic risks and uncertainties that enable the human entrepreneurial talent and 55 ...... 9 I...... 9 2 creativeness to meet the present challenges in a more successful way. The worldwide discussion on Risk Management is a sign of this process. The connotations of the notion of risk in the service economy cover a much larger ground than the notion of risk current in the Industrial Revolution. In the latter case, the key risk normally referred to is the so-called entrepreneurial or commerical risk; in the service economy, it is extended to the so-called pure risk. The entrepreneurial risk is one where the people involved in an action can influence its goals and the wa y the action develops by deciding to produce, to sell, to finance, etc. The pure risk is out of reach to those involved in an action. It depends on the vulnerabilities of their environment or of the system they are working in, and it will materialize by accident and by hazard. This notion of pure risk is strictly linked to the notion of the vulnerability of systems and its relevance is distinctive of the service economy and for correctly understanding the role ofinsurance. One of the great differences between neoclassical economics and the new service economy is that not only the "entrepreneurial" risk is taken into account (as in the case of Frank Knight), but that the notion of the economically relevant risk is extended to include the notion of pure risk. The notion of risk, globally, has therefore two fundamentally different but complementary connotations. For any important economic endeavour, the consideration of both notions of risk is today on an equal strategic level (again linked to the notion of systems and of vulnerability). An appropriate risk management action needs to identify each category of risks in its own right, evaluate how they interract and design an appropriate global risk management strategy. The demarcation line between pure and entrepreneurial risks is the notion of "moral hazard"(2). This notion has long been understood by insurers wen tey ave to face damages produced by those suffering them with the purpose of making money out of them. For instance, the case of somebody burning his own home or building to collect te insurance: such cases concern more than 20 percent of fires(3). Economists look at this notion from another point of view as a spin-off of their studies on economic incentives: moral hazard is equivalent to studying the negative results of incentives. One important case concerns the level of social insurance for unemployed people who might stop looking for another job if the level of compensation is too high(4). Many economists who have dealt with public policy are entering into the field of moral hazard (=negative effects of incentives) and could profit from the old experience of insurers in this field(5). ...... (2) Among many other references, the subject has been presented in an Annual Lecture of the "Geneva Association" by Joseph Stiglitz ("The Geneva Papers" nr. 26, January 1983). (3) See "The Invisible Bankers", by Andrew Tobias, Pocket Books, New York, 1982. (4) There is a delicate economic and social trade-off problem here: a high level of compensation for unemployment might be socially desirable. But it is also desirable not to use working taxpayers' money beyond a certain level which might favour unemployment, by eliminating the need to find a job. (5) "The Pure Theory of Moral Hazard", by Joseph Stiglitz, op. cit. 56 ...... 1 9 9 2 4. Technology and the growth of pure risks Modern technology is at the source of the increasing management problems of pure risks in many ways: a) Increases in the economies of scale have been mainly due to progress in technology. The gains in productivity throughout the period of the Industrial Revolution were enormous, but the increased concentration of production also increased its vulnerability to small disturbances. This is the area where risks and vulnerability are increasingly of the "pure", insurable type: they need a statistical universe of occurrences in order to assess probabilities in a given time and space dimension. Lower frequency and higher size of risks make them more and more difficult to manage. b) Specialization has been a key factor for industrial progress, but an excess of specialization has today resulted in systems that are increasingly interdependent and vulnerable, leading to a high growth of consequential losses (losses deriving from the non-functioning or malfunctioning of a system). Furthermore, specialization can reduce the adaptability to changing market conditions of a machine or installation, and can impose more severe maintenance and repair requirements that may be difficult to implement under some operating conditions. Gains from specialization may be partly offset or even outweighed by the lack of flexibility that results. c) Operating reliability has made great progress due to advances in technology. However, minor variations and small accidents in one component can lead to disasters in a complex system, even if these accidents occur less frequently due to the higher operating reliability. d) The quality of many products has been greatly improved by modern technology. However, this same improved quality for a specific task may increase the problem of its recycling when a product is thrown away. The human and economic environment, as Alfred Marshall puts it, is much more like a biological process than a mechanical one. An improvement in one sense may introduce disequilibria in another: this is the lesson brought home by the problems of pollution and hazardous waste management control. These examples have in common a shift of emphasis from the traditional entrepreneurial risks to pure risks of the "insurable" type. We can thus expect to find a reflection of these developments in the practice of insurance business, which in fact is passing through a period of quantitative and qualitative change and development unequalled in its long history. 5. The quest for performance and liability risks Risk is now becoming concentrated at levels where the vulnerability is such that the overall uncertainty of the economic process increases. How many Boards of Management today dream of the decision possibilities experienced twenty years ago? Consumers are also reluctant to become increasingly consumers of "risk". The unique situation in the field of product liability and malpractice in the United States although amplified by a specific legal environment, starts to have its effects on other parts of the world. This is a typical trend of demand in the service economy: the consumer is more and more conscious that tools and products which exist for given purposes and even experts are only of value when the results of their "utilization" is positive. The fact that their utilization might give negative results is refuted and gives raise to requests 57 ...... 9 9 2 for compensations) A product liability is a great issue in the United States where litigation has led in some cases to extremely high and even excessive compensations. Chemical and pharmaceutical companies have a special problem(7) in this area. Doctors, lawyers and other experts are sued in court for "malpractice" and have to compensate their clients if found guilty(8). At the European level, a Directive(9) is the result of ten years of discussions and preparations to manage the expanding phenomenon of the increasing perception by the public that producers of economic wealth have to be liable for delivering a 4 4product" yielding negative results. Once again, in the contemporary economy, it is the "performance" which has economic value, which counts, rather than the simple "existence" of a product or service. The problem of environmental hazards, which very often is linked with the question of transportation and storage of dangerous materials is part of the same type of risks and vulnerabilities that our modern society has to face(10). 6. Insurabifity and risk management If the risk management activity is to be fully integrated into economic practice, the notion of insurability must become a key took for economic and managerial analysis. Insurability(ii) means essentially that a risk of the pure type can be rationally managed (be it or not insured in actual practice). In other words, it concerns the fact that the main risk characteristics, i.e. size and frequency, combine in such a way that the risk is predictable and economically manageable within a reasonable level of probability. Insurability should become one of the main factors to be considered for defining the effective efficiency levels of economies of scale and of productivity. And this on two grounds: - The prevention and security costs associated with the vulnerability of a given system; - The possibility to rationally (economically) control the level of risk in the case of any adverse occurrence. The supply-demand system in risk management is described in the following table, where the economic demand to cover a risk is given in terms of levels of vulnerability: whatever the source ...... (6) See S. Shavell, "Accidents, Liability and Insurance", Harvard Institute of Economic Research, 1979, Discussion Paper nr. 685 "Liability, Insurance and Safety Regulation", The Geneva Papers on Risk and Insurance nr. 43, April 1987 (special issue). See also The Geneva Papers on Risk and Insurance nr. 56, July 1990, on "Comparative Studies in Liability and Compensation". (7) See the book by Arthur Hailey, "Strong Medicine", Pan Books, London, 1985. (8) See R. Jackson and J. Powell, "Professional Negligence", Swee & Maxwell, London, 1982. (9) The European Community, Directive on Product Liability, 1985. See also "Liability Issues", The Geneva Papers on Risk and Insurance nr. 61, October 1991. (10) See "Transportation, Storage and Disposal of Hazardous Materials", Papers from a Conference at I.I.A.S.A., Laxenburg (Vienna), edited by H. Kunreuther, Wharton School, University of Pennsylvania, Philadelphia, 1986, and "Hazardous Waste Management", The Geneva Papers on Risk and Insurance nr. 51, April 1989. (11) See in particular Baruch Berliner, "Limits of Insurability of Risks", Prentice Hall, New York, 1982; also the special issue of "The Geneva Papers" on "Limits of Insurability of Risks" nr. 39, April 1986. 58 ...... 1 9 9 2 or the origin of the risk and damage, the demand-side is interested in adequately managing or covering its levels of vulnerability. From the supply-side, risks are covered by homogeneous classes, defined by homogeneous origins of the risks. Table: Supply and Demand for Risk Cover (Risks of "insurable", pure type) Risk of The system is catastrophic E destroyed and cannot level be replaced ------The system is Important Risks destroyed and can be replaced ------Average ...... The system is damaged Risks and can be repaired 4------The system is slightly Minor damaged and normal Risks maintenance measures SUPPIY are adequate A B C D X Various types of risks by origin Risk covered LLLJ It is very important to re-examine in depth what the notion of homogeneity of risk classes may mean when: - An industrial risk today might simultaneously refer to an insurance or insurable cover(12), which is offered from different risk classes. A production in a chemical plant might stop and produce consequential losses; this production stop might be linked to the identification of a product which is likely to produce damages when used. In this case, it falls under product liability. Further, an industry might be obliged to spend money to recall these products. It is obvious that in a situation in which one single accident produces opportunities of losses in so many different directions, the insurance management of such a problem should strive to integrate as far as possible various risk classes (in this case consequential losses, recall practices and product liability together). ...... (12) The notion of insurable and of pure risks goes beyond what is actually underwritten by insurance companies. For instance, Felix Kloman (Business Insurance, October 19, 1987, pp. 3 and 23) estimates that by 1995, selfinsurance techniques will account for half of the total cost of risk financing in commercial and industrial activities. 59 ...... 9 9 2 Technology very often means specialization and today, when one wants to cover for instance those risks associated with space activities, one finds that each and every space programme has a different vulnerability configuration. A third, very important aspect, is the question of the time of reference related to various classes of risk and in particular those with catastrophic connotations; in other words, the problem that in order to reconstruct a usable universe of risks, it is more and more necessary to consider time units not simply on the basis of one year of activity as is normal practice in the economic system, but often on periods of 5, 10, 20 years or more. This also is very often the effect of specialization and increase of technological vulnerability. This very important issue has a lot to do with fiscal policies. The key economic question here is: which is the correct period of time (how many years) during which an economic activity should be taken into consideration? This requires a massive and fundamental dialogue and research between risk managers and fiscal economists. Up to now, when insurable risks were not an essential tool in managing the overall risks of an enterprise, of an individual or even of a society, a client could accept to adapt to the classification proposed on the supply side. Once the insurable risks become higher or more important for an industrial company, the profession of risk manager arose up, which essentially represents an effort to define the demand side, not in terms of one type of risk but in terms of one level of vulnerability. This is parallel to a movement where more and more risks, in the industrial sector, enter the upper categories (important and catastrophic risks) and when it becomes even more of a problem to understand, estimate and underwrite risks related to new technologies. Moreover once an industrialist analyses his risk control demand, in terms of protection and quality control of his product from te production line to the consumer, he becomes more and more aware of the problem of how to integrate a fire policy with a consequential loss, with a product liability, with a recall protection, etc. At this stage, the insurance supplier needs greater qualifications to understand the system at risk to be insured at its various vulnerability points and integrate adequately the specific ("vertical") insurance possibilities. In a purely supply oriented situation, the actuary is more than adequate to define rates within a homogeneous category. Then, when real demand is defined, the engineer and the economist have to step in. It should be clear that it will always be necessary to define homogeneous risk classes, at least up to a reasonable level, in order to rationally cover and protect pure risks. The task of te "pure risk" manager consists therefore in matching two different logics: the one referring to the definition of the vulnerability level of a specific economic system which is a clear characteristic of the service economy, and the other referring to the indispensable technique of diluting and controlling risk occurrences. This example which is close to the experience of the insurance sector is once more significant of a key aspect of the service economy, and for all classes requiring a professional management of risks. 60 ...... 1 9 9 2 7. Examples of research and activities in risk management and insurance The development of the situation which we have tried to describe in the previous points has stimulated the insurance world to develop its professionalism and its research basis to evaluate vulnerabilities in various industrial sectors and the way to cope with them. The Geneva Association (International Association for Risk and Insurance Economics Research) has made a series of studies which are schematically presented in the following table: On the one side, they concern specific analysis of industrial sectors utilising new technologies or being under full transformation. On the other side, there is a series of studies on specific key issues concerning different industrial and economic sectors referring to problems of liability and of business interruption (consequential loss). Table 2 The Industry Insurance Interface P P S P Risk Management in Consequential Introduction to West European losses in the lability -0. industrial Recall Companies chemicalEuropean industry economics P F e studies n] Comparative costs of liability A- claims in differ- ent couuntries P U U P onsequential Second step specific Economic losses of losses production Updating of study on liability -ctors of computer systems in liability and SE, 1988 computer Computer losses technology utilization P Containersi economic losses in 1988 P osses due to LNG P/C P osses in ing industry C Robots and insurance C Superconductivity and insurai P C limahc Rsks C 61 ...... 1 9 9 2 The Geneva Association also publishes a quarterly on risk and insurance practice. In the issue of July 1992 (nr. 64), one can find an updating of key issues in risk management today. The GENEVA PAPERS on Risk and Insurance ISSUES AND PRACTICE RISK MANAGEMENT STUDIES 1. RISK MANAGEMENT TODAY Rethinking Risk Management by H. Felix Kloman 299 Toward a Holistic Approach to Total Risk Management by Yacov Y.Haimes 314 Comparative Product-Life-Cycle Confrontation of Risks by William W.Lowrance 322 Risk Management inthe United Kingdom - A Personal Retrospect by Dennis Farthing 329 Risk Management from a Technological Perspective by Vernon Leslie Grose 335 Taking Aim at Environmental Risks: Questions of Feasibility and Desirability by Adam M.Finkel 343 The Past and Future of Loss Financing by James V.Davis 355 (contd. last cover page) Les Cahiers de Gen6ve No. 64 (17th Year) Die Genfer Hefte July 1992 62 ...... 9 9 2 Is Good News No News? by Judith Lichtenberg and Douglas MacLean 362 Education in Risk Management A European View: Past, Present and Future by Gordon C. A. Dickson 366 11.STUDIES ON CLIMATIC CHANGE An Analysis of the Climate Change Issue by J. L. Rasmussen 371 Greenhouse Effects on Natural Catastrophes and Insurance by Gerhard A. Berz 386 Insurance Implications of Climatic Change by Andrew F. Dlugolecki 393 Sea-Level Changes and Forecasting Floodrisks; by M. J. Tooley 406 111. MANAGING RISK IN THE CATERING INDUSTRY Case Study: The Management of Risks to Employees and Consumers of the Catering Industry in the United Kingdom and France by Alan Gordon 415 63 ...... 1 9 9 2 64 ...... 1 9 9 2 Monday, 5 October 1992 The role of risk assessiment in ]International Policy 65 ...... 9 XA04NO308 ILO Activities in the Field of Risk Assessment Dr C Pinnagoda Chief, Occupational Safety and Health Branch, Working Conditions and Environment Department Introduction Control of chemical risks has become a matter of increasing concern for the occupational safety and health activities of the ILO. Its chemical safety programme focuses attention on a series of comprehensive measures to promote national level action for effective control of chemical risks aimed at the protection of workers and consequently of the general public. The supply and use of information based on risk assessment for exposure control and the application of practical improvement measures through improved risk management and training continue to receive increasing importance. The important and direct role of employers and workers in the assessment of risks and planning and implementation of workplace improvements is an identified need. Risk assessment is a component of the chemical safety programme of the ILO that places significant emphasis on: the setting of international standards that lay down basic principles in chemical risk assessment and control; national level action to promote coherent policies and programmes on safety in the use of chemicals and prevention of industrial disasters (major hazard control); technical advisory services through the development of guides and training programmes that focus on workplace participatory approaches; collection and dissemination of practical information; and technical co-operation assistance to developing countries. International labour standards International Labour Standards (called Conventions and Recommendations) are adopted by the International Labour Conference characterised by its tripartite composition of delegates from governments and employers' and workers' organisations. Of over 350 standards adopted so far, more than half deal with safety, health and working conditions. The Occupational Safety and Health Convention, 1981 (No. 155) and its accompanying Recommendation (No. 164) that lay down the foundation for the establishment of a coherent and comprehensive policy at national and enterprise level for prevention of occupational accidents and diseases; and the Occupational Health Services Convention, 1985 (No. 161) and the Recommendation (No. 171) for the establishment of occupational health services which incorporate the provisions for monitoring the health of workers and the working environment are basic instruments. These instruments provide an important basis for national and enterprise level action to establish criteria for risk assessment and promote safety in the use of chemicals. Other examples, in general, include: - Asbestos Convention, 1986, (No. 162) and Recommendation (No. 167); - Occupational Cancer Convention, 1974, (No. 139); - Benzene Convention, 1971, (No. 136) and Recommendation (No. 144). 66 ...... 1 9 9 2 The Chernicals Convention The Chemicals Convention, 1990, (No. 170) and Recommendation (No. 177) provides for safety in the use of chemicals at work and also extends its scope to environmental considerations. Safety in the use of chemicals is ensured in these instruments through their focus on principles for classification and labelling of chemicals, provision of chemical safety date sheets (CSDS), minimising exposure of workers to hazardous chemicals, provision of information and training to workers, determining the rights and duties of workers, assigning the responsibilities of suppliers, employers and exporting countries and soliciting the co-operation of all parties. The responsibility for risk assessment of chemicals becomes the duty of suppliers (whether manufacturers, importers or distributors) as they have to ensure that: - chemicals have been classified in accordance with systems and specific criteria based on the type and degree of their intrinsic health and physical hazards and a search made for available information, or be assessed for such chemicals that have not been classified; - all chemicals are marked so as to indicate their identity; - chemicals classified as hazardous which they supply are then labelled and CSDS are prepared for such hazardous chemicals and provided to employers who are the usual customers. Furthermore, suppliers of chemicals which have not yet been classified should identify these chemicals, and assess the properties on the basis of a search of available information and then determine whether they are hazardous in order to convey the relevant information to the customers. The component authority of each country, or a body approved or recognised by the competent authority, in accordance with national or international standards has to establish: (a) systems and specific criteria appropriate for the classification of all chemicals (Note - the Convention specifies that, the hazardous properties of mixtures can be determined by calculation and in the case of transport, systems and criteria shall take account of the UN Recommendations on Transport of Dangerous Goods. The Convention further specifies that the classification system and their application shall be progressively extended); (b) requirements for marking of chemicals so as to indicate their identity and for labelling or hazardous chemicals in a manner which is easily understandable to workers. (Again, in the case of transport, due account should be taken of the UN Recommendation on Transport of Dangerous Goods); and (c) criteria for preparation of CSDS for hazardous chemicals. (See Annex I for details.) Thus, the Convention provides for the competent authority to exercise its mandate to ensure that hazardous chemicals used are subject to assessment of hazardous properties in accordance with national or international standards. The term "use of chemicals at work" means: - any work activity which may expose a worker to a chemical, including: - production of chemicals; - handling of chemicals; - storage of chemicals; - transport of chemicals; - disposal and treatment of waste chemicals; 67 ...... 1 9 9 2 - release of chemicals resulting from work activities; - maintenance, repair and cleaning of equipment and containers for chemicals. The Convention further specifies that the information about the intrinsic hazardous properties of chemicals are utilised at the workplace to ensure safety in the use of chemicals at work. The employers have a responsibility to ensure that, all chemicals have been labelled or marked and that the CSDS have been compiled for hazardous chemicals; the CSDS are also to be made available to the workers concerned. If chemicals have not been labelled or marked, or CSDS have not been provided, employers should obtain the relevant information from the supplier or from other sources, and should not use the chemicals until such information has been obtained. Employers should ensure that when chemicals are transferred to other containers, the contents are shown so as to make their identity known to the workers; any hazards associated with their use and any safety precautions to be observed should also be clearly shown. Employers should involve occupational hygiene measures for minimising the risk by controlling the exposure of workers to hazardous chemicals; assessing their exposure; maintaining records of exposure for prescribed periods and ensuring that these records are accessible to the workers and their representatives. (See Annex 2 for details.) Practical workplace approaches provide for employers to assess the risks from the use of chemicals at work, and to protect workers against such risks by appropriate means such as the choice of chemicals, the choice of technology, the use of adequate engineeering control measures, the adoption of working systems and practices and adequate occupational hygiene measures, and the provision and proper maintenance of personal protective equipment and clothing. These should be supplemented by precautions to limit exposure of workers to hazardous chemicals, to provide first-aid, and to make arrangements to deal with emergencies. (See Annex 3 for details.) Participatory approach and co-operation at the level of the enterprise is governed by specific provisions whereby employers, based on results of risk assessment, have to: - inform the workers of the hazards associated with exposure to chemicals used; - instruct the workers how to obtain and use the information provided on labels and CSDS; - use the CSDS along with information specific to the workplace, as a basis for the preparation of instructions to workers; and - train the workers on a continuing basis. Workers and their representatives shall have the right to: - information on the identity of chemicals, precautionary measures, education, and training; - information contained in labels and markings; - CSDS; and - any other relevant information. Annex 4 enumerates further details. The criteria for the preparation of CSDS for hazardous chemicals contain essential elements of risk assessment and hazard evaluation. They should contain essential information as listed in Annex . 68 ...... 9 9 2 Related ILO activities ILO activities are formulated and implemented to promote and to strengthen action at the national level in the field of occupational safety and health. Activities in this field, notably technical co-operation assistance are known as the "International Programme for the Improvement of Working Conditions and Environment" (PIACT). Within the framework of PIACT a series of activities are undertaken to reinforce national and enterprise level actions for preventing accidents and diseases caused by the use of chemicals and promote good management of chemicals utilising also the results of risk assessment. The support is provided usually to member States by various means of ILO action. These include Conventions and Recommendations, codes of practice, guides and manuals, technical advice and information, and technical co-operation projects. Particular mention should be made of the following: The implementation of principles set out in the chemicals Convention and Recommendation will provide for the establishment of basic principles for national policies for the promotion of chemical safety at workplaces. The Convention provides a binding obligation on the member States that ratify it. These instruments give guidance to all member States irrespective of whether they ratify the Convention or not. Each member State ratifying the Convention shall make an annual report to the ILO on measures which it has taken to give effect to the provisions of the Convention. These reports provide the supervisory mechanism for monitoring the implementation and feedback. The ILO activities for chemical safety carried out within the PIACT programme include: preparation of codes; guides and manuals; provision of advice and information; rganisation of seminars and workshops; technical co-operation projects, and action aimed at promoting the implementation of principles set out in the Chemicals Convention and Recommendation. A Code of Practice adopted by a tripartite meeting of experts in April 1992 provides further guidance on the safety in the use of chemicals with specific reference to their classification, labelling, transport, storage, handling and use, and waste disposal. It will assist member States in the formulation of appropriate national legislation and the establishment of a national infrastructure to deal with chemical safety. It may also be used as a training manual. Thus, the Code will assist ILO constituents to have access to essential information related to risk assessment. Harmonisation of Chemical Classification and Labelling Systems The Chemicals Convention and Recommendation require national criteria and systems to be established for the classification of chemicals according to their intrinsic hazards. The 1989 International Labour Conference adopted a resolution concerning the harmonisation of systems of classification and labelling for the use of hazardous chemicals at work. A harmonised international classification and labelling system will help eliminate contradictions between the transport and use oriented systems, avoid duplication of efforts in chemical risk assessment and thus reduce the use of laboratory animals, facilitate the elaboration of hazard communication tools in many languages, and facilitate international trade by reducing the need for notification and relabelling. In response to the resolution, the ILO has taken a leading role in promoting the harmonisation of existing classification and labelling systems in a manner that contributes to maintaining the 69 ...1 ...... I...... I9 ...... 9 2 highest level of protection of workers and the public in general while facilitating international trade. In order to share the burden of this work and to establish efficient coordination among the national, regional and international bodies administering existing systems, the work of harmonisation will be carried out within the framework of the UNEP/ILO/WHO International Programme on Chemical Safety (IPCS). Following completion by the ILO of a review of the major existing national, regional and international systems to evaluate the task of harnionising these systems, and upon proposals made by the ILO a Coordinating Group for the Harmonisation of Chemical Classification Systems (CG/HCCS) has been established within the framework of IPCS (January 1992). The Coordinating Group includes as its core, the organisations presently cooperating in the IPCS (ILO, WHO, UNEP), the OECD and the UN Committee of Experts on Transport of Dangerous Goods (UN CETDG). The FAO is invited to join this group. Very strong links are established with those responsible for already established systems (e.g. Canada, the Commission of the European Communities, and the United States of America). The UN CETDG is expected to represent the views of other organisations involved in the transport of dangerous goods such as IMO and ICAO. The ILO provides the secretariat for the Coordinating Group. Participation of concerned international industry, employer and worker organisations to extended meetings of the Coordinating Group will be ensured. The main objective of the CG/HCCS will be to catalyse the development of a globally harmonised classification and hazard communication system for chemicals, according to an established workplan. This implies a role of effective coordination and overseeing of activities undertaken by international, regional and national bodies who express an interest in working on specific aspects of harnionisation. The CG/HCCS will eventually function within a restructured and strengthened IPCS as recommended by the UNCED Preparatory Committee. It should represent a model for setting up additional coordinating structures to address the other five action programmes proposed in the UNCED agenda for achieving an environmentally sound management of chemicals worldwide. The Coordinating Group, through appropriate consultations, will set-up priorities and identify focal points responsible for given activities. It will assist the focal points to establish working parties which may be composed of several national agencies, research institutions, members of the private sector and other interested bodies. Each working party will be responsible for ensuring te completion of specific task(s) assigned. Te Coordinating Group, through its secretariat, will assist the focal points in their tasks. It will communicate to the focal points the planned activities, establish procedures for liaison, coordination and reporting, and initiate actions. The final product of any harnionisation activity for a given hazard classification or hazard communication element will be a set of harmonised criteria, methods or guidelines to be presented as specific IPCS Recommendations or Guidelines. Procedures for international adoption could be developed through an eventual intergovernmental forum for chemical risk assessment and management as endorsed by UNCED in its Agenda 21. As examples of recent harnionisation activities, an OECD Clearinghouse including the Commission of the European Communities, Sweden and the USA is close to completing its work on harmonising criteria for "acute oral toxicity" and "danger to the environment". At its first meeting of the Coordinating Committee 23 March 1992 a proposal that the ILO becomes the focal point for the task of harmonising classification criteria for physical hazards was endorsed. 70 ...... 1 9 9 2 The secretariat of the CG/HCCS has prepared a Work Programme for the period 1992-97 which defines the different elements for prioritisation of the work, provides a set of general principles, a procedure and a timetable. The scope for inter-agency collaboration has received an added impetus with the adoption by the 45th World Health Assembly (May, 1992 a Resolution on IPCS which calls upon the Director- General of the WHO, among other things: to review the current arrangements with the Executive Heads of 110 and UNEP, as well as with representatives of other organisations that might participate in the International Programme in the future, in order to determine the changes that would be required for its expanded role, including the function of secretariat for an intergovernmental forum on chemical safety, as recommended in proposals to be presented to governments at the United Nations Conference on Environment and Development; to take steps to ensure that in expanding the chemical risk management tasks of the Programme, the scientific quality and integrity of the work on risk assessment are fully protected; and to report to a future session of the Executive Board on the expanded International Programme, particularly in relation to the enhanced role of WHO with its partners in the implementation of the decisions of the United Nations Conference on Environment and Development for environmentally sound chemical risk management." The International Co-ordinating Committee (lCQ of the IPCS took note of this Resolution and the Recommendation of UNCED for an expanded and strengthened role for the IPCS in the context of chemical risk assessment and management. The ICC formed a Working Group comprising officials of the ILO, WHO and UNEP to identify the chemical assessment and management activities as a pre-requisite to formulating action. Prevention of Industrial Disasters (Major Hazard Control) Following various consultations and series of actions in response to recent major accidents, the ILO published (in 1988 a manual on Major Hazard Control for general use in all countries which have industrial activities with major accident potential. This Manual reflects the experience of several Western European countries which practise major hazard control systems. A Code of Practice on the Prevention of Major Industrial Accidents was published in May 1991. The main components of a major hazard control system described in the Code are: identification of major hazard installations; assessment of major hazards and their control by the management; emergency planning; safety report; and siting and land-use planning. Nuclear hazards and those of military nature as well as transportation of hazardous chemicals are excluded from the scope of the Code. It is expected that the International Labour Conference in 1993 will adopt new international standards for Prevention of Industrial Disasters. These will provide for the establishment and periodic review of a coherent national policy for the protection of workers, the public and the environment against the risk of major accidents. 71 ...... 1 9 9 2 The identification and analysis of hazards and the assessment of risks is a key element in the proposed instruments for arrangements at the level -of the major hazard installation. Identification of major hazard installations; notification following this identification; preparation of safety reports; general protective measures; accident reporting and inspection are other key provisions addressed in the proposed instruments. The ILO continues to assist member States by providing advisory services for establishing national systems for the prevention of industrial disasters. ILO action relating to major hazard control has placed particular emphasis on technical cooperation activities. The aim of these activities is: to stimulate action to install a national system for identification of potentially hazardous industries; for risk assessment and control of these industrial activities; and for emergency operations in case of major accidents. The ILO has executed technical co-operation projects in India, Indonesia, and Thailand aiming at establishing a national unit for major hazard control advisory services to the industries and government inspectors. These also include strengthening the chemical inspection capabilities of inspectorates, and training of management, supervisors, and workers concerned. Future action Recent experiences in the ILO relating to chemical safety and major hazard control activities demonstrate that effective services in occupational safety and health have some common features. Successful action programmes need to be based on local traditions and culture and on local practice. It is necessary to provide positive guidance and feedback for better safety and management and to facilitate active participation of employers and workers. It is necessary to learn from these success stories and provide support for organising workplace action linked to local solutions. It is to be expected that priorities for action may vary according to the extent of problems and available means of action. For the future there is, therefore, a need to continue to promote the development of comprehensive chemical safety and risk assessment programmes directed towards the needs of countries in all regions, and the effective implementation of such programmes through concerted action at national, regional and international levels. In response to the identified need for the exchange of experiences and information to promote practical safety and health measures at the workplace, the collection and dissemination of priority information concerning safety and health legislation, training courses, guides, and reports will continue to be expanded. Requests concerning work hazards, chemicals, and preventive measures by governments and employers' and workers' organisations continue to increase both in number and in complexity. Particularly important are requests for information concerning safety and health legislation, practical solutions to safety and health problems, and training materials. On the basis of the experience gained by the ILO, technical assistance will continue to be provided to developing countries for the establishment and improvement of safety and health information centres. 72 ...... 9 9 2 Future ILO action in chemical safety related to risk assessment will support the harmonisation of existing classification systems, the establishment of a common scheme for labelling, the collection and dissemination of chemical safety data sheets for use by employers and workers and the development of practical training programmes. The ILO will take a leading role within the enhanced IPCS activities through the CG/HCCS. In particular, technical assistance in strengthening national level regulatory measures, safety and health surveillance programmes, and organising training activities, based on international assistance and tripartism, will play an important role in future action programmes for promoting safety in the use of chemicals in all occupations. Finally, it should be recalled that chemical safety and major hazard control systems could function effectively only with parallel development and it should be emphasised that risk assessment is a key to the success of effective measures for the protection of the workers, the public and the environment against chemical hazards. 73 ...... 1 9 9 2 ANNEX I ILO Convention No. 170 concerning Safety in the Use of Chemicals at Work (Cited as the Chemicals Convention, MO) PART III CLASSIFICATION AND RELATED MEASURES Article 6 CLASSIFICATION SYSTEMS 1. Systems and specific criteria appropriate for the classification of all chemicals according to the type and degree of their intrinsic health and physical hazards and for assessing the relevance of the information required to determine whether a chemical is hazardous shall be established by the competent authority, or by a body approved or recognised by the competent authority, in accordance with national or international standards. 2. The hazardous properties of mixtures composed of two or more chemicals may be determined by assessments based on the intrinsic hazards of their component chemicals. 3. In the case of transport, such systems and criteria shall take into account the United Nations Recommendations on the transport of dangerous goods. 4. The classification systems and their application shall be progressively extended. Article 7 LABELLING AND MARKING 1. All chemicals shall be marked so as to indicate their identity. 2. Hazardous chemicals shall in addition be labelled, in a way easily understandable to the workers, so as to provide essential information regarding their classification, the hazards they present and the safety precautions to be observed. 3. (1) Requirements for marking or labelling chemicals pursuant to paragraphs I and 2 of this Article shall be established by the competent authority, or by a body approved or recognised by the competent authority, in accordance with national or international standards. (2) In the case of transport, such requirements shall take into account the United Nations Recommendations on the transport of dangerous goods. Article CHEMICAL SAFETY DATA SHEETS 1. For hazardous chemicals, chemical safety data sheets containing detailed essential information regarding their identity, supplier, classification, hazards, safety precautions and emergency procedures shall be provided to employers. 2. Criteria for the preparation of chemical safety data sheets shall be established by the competent authority, or by a body approved or recognised by the competent authority, in accordance with national or international standards. 3. The chemical or common name used to identify the chemical on the chemical safety data sheet shall be the same as that used on the label. 74 N ...... 1 9 9 2 ANNEX2 ILO Convention No. 170 concerning Safety in the Use of Chemicals at Work (Cited as the Chemicals Convention, D90) Article 12 EXPOSURE Employers shall: (a) ensure that workers are not exposed to chemicals to an extent which exceeds exposure limits or other exposure criteria for the evaluation and control of the working environment established by the competent authority, or by a body approved or recognised by the competent authority, in accordance with national or international standards; (b) assess the exposure of workers to hazardous chemicals; (c) monitor and record the exposure of workers to hazardous chemicals when this is necessary to safeguard their safety and health or as may be prescribed by the competent authority; (d) ensure that the records of the monitoring of the working environment and of the exposure of workers using hazardous chemicals are kept for a period prescribed by the competent authority and are accessible to the workers and their representatives. ANNEX3 ILO Convention No. 170 concerning Safety in the Use of Chemicals at Work (Cited as the Chemicals Convention, L990) Article 13 OPERATIONAL CONTROL 1. Employers shall make an assessment of the risks arising from the use of chemicals at work, and shall protect workers against such risks by appropriate means, such as: (a) the choice of chemicals that eliminate or minimise the risk; (b) the choice of technology that eliminates or minimises the risk; (c) the use of adequate engineering control measures; (d) the adoption of working systems and practices that eliminate or minimise the risk; (e) the adoption of adequate occupational hygiene measures; (D where recourse to the above measures does not suffice, the provision and proper maintenance of personal protective equipment and clothing at no cost to the worker, and the implementation of measures to ensure their use. 2. Employers shall: (a) limit exposure to hazardous chemicals so as to protect the safety and health of workers; (b) provide first aid; (c) make arrangements to deal with emergencies. 75 ...... 1 9 9 2 ANNEX4 ILO Convention No. 170 concerning Safety in the Use of Chemicals at Work (Cited as the Chemicals Convention, MO) PART V. DUTIES OF WORKERS Article 17 1. Workers shall co-operate as closely as possible with their employers in the discharge by the employers of their responsibilities and comply with all procedures and practices relating to safety in the use of chemicals at work. 2. Workers shall take all reasonable steps to eliminate or minimise risk to themselves and to others from the use of chemicals at work. PART VI. RIGHTS OF WORKERS AND THEIR REPRESENTATIVES Article 8 1. Workers shall have the right to remove themselves from danger resulting from the use of chemicals when they have reasonable justification to believe there is an imminent and serious risk to their safety or health, and shall inform their supervisor immediately. 2. Workers who remove themselves from danger in accordance with the provisions of the previous paragraph or who exercise any other rights under this Convention shall be protected against undue consequences. 3. Workers concerned and their representatives shall have the right to: (a) information on the identity of chemicals used at work, the hazardous properties of such chemicals, precautionary measures, education and training; (b) the information contained in labels and markings; (c) chemical safety data sheets; (d) any other information required to be kept by this Convention. 4. Where disclosure of the specific identity of an ingredient of a chemical mixture to a competitor would be liable to cause harm to the employer's business, the employer may, in providing the information required under paragraph 3 above, protect that identity in a manner approved by the competent authority under Article 1, paragraph 2 b). 76 ...... 1 9 9 2 ANNEX5 ILO Recommendation No. 177 concerning Safety in the Use of Chemicals at Work (Cited as the Chemicals Recommendation, MO) CHEMICAL SAFETY DATA SHEETS 10. (1) The criteria for the preparation of chemical safety data sheets for hazardous chemicals should ensure that they contain essential information including, as applicable: (a) chemical product and company identification (including trade or common name of the chemical and details of the supplier or manufacturer); (b) composition/information on.ingredients (in a way that clearly identifies them for the purpose of conducting a hazard evaluation); (c) hazards identification; (d) first-aid measures; (e) fire-fighting measures; O accidental release measures; (g) handling and storage; (h) exposure controls/personal protection (including possible methods of monitoring workplace exposure); (i) physical and chemical properties; 6) stability and reactivity; (k) toxicological information (including the potential routes of entry into the body and the possibility of synergism with other chemicals or hazards encountered at work); 6) ecological information; (m) disposal considerations; W transport information; (q) regulatory information; (p) other information (including the date of preparation of the chemical safety data sheet). 77 ...... 9 XA04NO309 The Role of Risk Assessment in the Work of the World Health Organization in Europe Kees A. van der Heijden and Richard M. Stern World Health Organization Regional Office for Europe European Centre for Environment and Health Bilthoven Division Abstract The World Health Organization, through its Headquarters in Geneva (WHO/HQ), and its Regional Office for Europe (WHO/EURO) in Copenhagen, has the responsibility for providing national governments with advice on formulation and implementation of public health policy globally and in Europe, respectively. Globally, the major areas for health related risk assessment/management is the provision of adequate and safe drinking water and food and control of vector borne and parasitic disease. In the industrialized countries of Europe, a wide number of issues are dealt with which require the development and application of risk assessment and risk management tools and strategies. Primary areas of application are in monitoring trends and status of public health, harmonization of issues of chemical safety, development of criteria documents for environmental pollutants, and providing decision support and technical cooperation, especially in the area of development policies and environment management and their potential health impact. An emerging concern is the need for the introduction of these methodologies in the Countries of Central and Eastern Europe, and harmonization of approaches used by international and intergovernmental organizations and the Member States. One of the first steps towards the management of the environment as a resource for health in Europe, the mandate given WHO/EURO by the European Charter for Environment and Health, (Frankfurt, 1989), has been the creation of the European Centre for Environment and Health (ECEH) with support from the Netherlands and Italian Governments. The initial task of ECEH is a description of the current state of the environment and the current state of public health in the European Region, using harmonized methodologies for information gathering. The production of this report, "Concern for Europe's Tomorrow", provides the basic elements of a unified regionwide approach to priority setting for the risk assessment and risk management process. Introduction One of the roles of the World Health Organization (WHO) is to provide information for and advice on the formulation and implementation of public health policy among the Member States of the United Nations. Underlying this WHO activity in the area of public health since 1977 has been the concept of Health For All By The Year 2000 (HFA). In the European Region, the Regional Office for Europe (WHO/EURO) promulgated 38 HFA Targets in 1984 to define the programme of work in the public health arena while in the area of environmental health, since 1989, The European Charter On Environment and Health has provided the operational framework. Within WHO, the formal process for managing risks has been defined as consisting of four phases: hazard identification, risk assessment, risk evaluation and priority setting, decision 78 ...... 1 9 9 2 making. Because its role is limited to an advisory capacity, the use of (quantative) risk assessment ((Q)RA), and the development of risk management (RM) strategies and practices as tools for health promotion is extremely context dependent, especially within each of the four major "Chapters" of the programme of work of WHO/EURO. The areas encompassed by these chapters are: The Basic Requirements for Health (health services, disease prevention and quality of care), Lifestyles and Health (health promotion, health policy), Environment and Health (multisectorial policies, monitoring, control) and Appropriate Care (primary health care, resource coordination), and exemplify the range of WHO activities. The use of Q)RA by WHO falls into two broad categories: monitoring of (national) health status worldwide, and of specialized HFA indicators in Europe (e.g. collection of national data concerning health outcomes, especially mortality, morbidity, and incidence statistics) and the development of criteria documents and guidelines (e.g. establishment, by expert concensus of values for toxicity and of no (adverse) effect levels for specific substances, and the man'agement and evaluation of epidemiological studies). Risk assessment, as such, can take on many forms, only some of which are systematized and have the characteristics of the formal probabilistic approach used for example by the chemical industry in loss prevention. Examples of Risk Assessment by WHO Globally Global issues of health are dealt with by the WHO Headquarters (WHO/HQ) in Geneva, Switzerland. The health challenge of coming decades has been seen to require concerted action by many individuals and organizations directed and coordinated by WHO/HQ globally and each of the six Regional Offices, with the assistance of Collaborating Centers locally. Because of public pressure, the environment has recently been placed high on the political agenda of all nations, and it is felt that environmental problems have reached new dimensions, coming to the point where the deleterious effect of environmental degradation on health is becoming evident. An independent Commission on Health and Environment was convened by WHO in 1990 to prepare a report, "Our Planet, Our Health", which is essentially a global assessment of the health consequences of environmental change, recently presented at the U.N. Conference on Health and the Environment in Rio de Janeiro, Brazil. An underlying philosophy of this report is "The maintenance and improvement of health should be at the centre of concern about the environment and development. Yet, health rarely receives high priority in environmental policies and development plans ... despite the fact that the quality of the environment and the nature of development are major determinants to health". Within this context, the primary use of RA is simply to enumerate health effects, first globally to establish major priorities for programmes, and then to perform local analysis as an aid in national policy development and implementation. The WHO report continues "...the most immediate problems in the world are ill health and premature death caused by biological agents in the human environment: in water, food, air and soil... The problem is most acute in developing countries where 25 million infants or children die every year from diarrhoeal diseases, largely as a result of contaminated food or water; over a million people die from malaria each year and 267 million are infected; hundreds of millions suffer from debilitating intestinal parasitic infections". In this respect, the basic need is to improve the ability of WHO to convince national governments to introduce the systematic use of validated techniques of QRA to plan for resource allocation for programme development and implementation. 79 ...... 1 9 9 2 It is timely that WHO examine the use of (Q)RA and RM methodologies by international and national organizations involved in development programmes which have health consequences, the need for which has recently been documented: ineffective planning of anti-malarial programmes which concentrated on protecting the environment and led to a significant increase in the spread of the disease, and exacerbation of schistosomiasis and other water-related diseases through agricultural development projects which greatly increased the area of impounded water and hence vector habitats, are two examples of areas where the use of QRA techniques which included health could have been expected to significantly effect development policies thereby leading to a reduction in human disease and suffering. Although infectious and parasitic diseases are much less prevalent in Europe than in the developing countries, according to WHO, other "Serious environmental health problems are shared by both developed and developing countries, affecting: hundreds of millions of people who suffer from respiratory and other diseases caused or exacerbated by biological and chemical agents, including tobacco smoke, in the air, both indoors and outdoors; hundreds of millions who are exposed to unnecessary chemical and physical hazards in their home, workplace, or wider environment (including 500,000 who die and tens of millions more who are injured in road accidents each year)." Towards this effort, WHO/HQ has developed a major International Programme on Chemical safety (IPCS), and a global programme on accident prevention and disaster preparedness. Originating in decisions made at the Stockholm Conference in 1972, PCS is a joint programme of the International Labour Organization, The United Nations Environmental Programme and WHO, set up to specifically provide assessment of the risk to human health and to the environment by chemicals, thereby providing the scientific basis on which Member States could develop their own policies involving chemical safety. Other objectives are: promotion of methodology for risk evaluation; promotion of technical cooperation between member states; promotion of effective international cooperation with respect to emergencies involving chemicals; support of national programmes for prevention and treatment of poisonings involving chemicals; to promote manpower training. Evaluations of risks within IPCS are performed through a consultative process involving use of all available information according to standard guidelines, the results of which are published worldwide in several languages within the series "Environmental Health Criteria". Over 20 such documents have been prepared for a wide range of substances selected through a separate consultative process, based on adverse effect, exposure, targets and international concern. Other outputs of the IPCS programme are "Health and Safety Guides" which are short documents summarizing toxicity information for 50 substances, "International Chemical Safety Cards" summarizing essential data for 86 products, and "Poison Information Monographs" designed for use by poison information centres. Other major RA activities of IPCS involve chemicals in food, especially additives, toxic contaminants and pesticide residues. An international mechanism has been established that uses formal QRA procedures for evaluation and tolerance setting for these classes of compounds in food. Evaluations are made by two expert committees: the Joint FAO/WHO Expert Committee on Food Additives (JECFA) which evaluates food additives, contaminants and veterinary drug residues, and the Joint FAO/WHO Meeting on Pesticide Residues (JMPR), which result in recommendations for levels that are considered to be safe, e.g. acceptable daily intake (ADI) and maximum residual levels MRL). These recommendations are then used by governments and the 80 ...... 1 9 9 2 Codex Alimentarious Commission for establishing safe levels of substances in foodstuffs and for food standards. Over 30 volumes have been prepared on the methodology of QRA for chemical substances. At the United Nations Conference on Environment and Development (UNCED) held in Rio de Janeiro in June 1992, chapter 19 of agenda 21 on "environmentally sound management of toxic chemicals" was adopted. In the follow-up of this decision, activities have been started in order to establish an "Intergovernmental Mechanism for Chemical Risk Assessment and Management" with IPCS acting as the secretariat and serving as the nucleus for coordinating international activities in this field. Examples of other WHO/HQ activities involving RA are: the IAEA/UNEP/UNIDOIWHO Inter- Agency Project on the Assessment and Management of Health and Environmental Risks from Energy and Other Complex Industrial Development Projects which attempts to introduce the concept of uniform methodology for assessment and management of risks; Water Quality Guidelines and; Studies of the Health Effects of Emerging Environmental Hazards e.g. climate change including ozone layer, acid rain and the greenhouse effect. Within WHO, the International Agency for Research on Cancer (IARC), Lyon, has the task of dealing with various aspects of this group of diseases on a global basis. In addition to contributing to the basic literature concerning quantitative laboratory assessment of the effect of various putative carcinogenic agents, and organizing and contributing to the development of the methodology of laboratory testing and environmental and clinical cancer epidemiology, the Agency collects extensive information on the local variations in disease incidence and mortality which are published in updated versions of the Cancer Atlas. Within the ongoing IARC programme of the evaluation of carcinogenic risk to humans, monographs dealing with over 800 substances have been produced to date. The intent of this highly visible and authoritative programme is to provide a consensus of the views and expert opinions of a series of ad-hoe working groups convened periodically to consider - for related subgroups of compounds, substances, processes and their accompanying exposures - the evidence for carcinogenicity in the available, peer reviewed literature. The evaluations are limited to a determination of the degree to which available evidence supports conclusions concerning human carcinogenicity, and are not concerned with relative potency or risk per unit exposure. VMO and RA in Europe Within the WHO Regional Office for Europe (WHO/EURO), elements of health risk management are part of almost all programmes throughout the various service divisions, although formal elements of RA and QRA are not used systematically. One of the major themes running through the programme of work of WHO/EURO is the issue of equity, and a major goal is to reduce inequity between various national populations, and within various sub-populations within each member state. The formulation of Target I of HFA, "By the year 2000, the actual differences in health status between countries, and between groups within countries should be reduced by at least 25 %, by improving the level of health of disadvantaged nations and groups", implies the use of some form of QRA for understanding the underlying causes and interactions leading to heterogeniety of health status. 81 ...... 1 9 9 2 Reduction of inequity involves the measurement of a large number of indicators, but only mortality can be used with a reasonable degree of reliability for intercountry comparisons owing to deficiencies in health statistics in many countries. Life-expectancy at birth has increased from a range of 57-74 in the early 1970s to 65-78 at the end of the 1980s, although the improvements have mostly been in the countries of western Europe, with the countries of eastern Europe now exhibiting a distinct pattern of poorer health than in the West, due mostly to continued high mortality from cardiovascular disease. As mortality data in Europe becomes available at subriational level, a picture of extreme heterogeneity with respect to health status appears: for some diseases, e.g. tuberculosis, variations in standardized mortality rates vary by a factor of ten or more between various administrative districts. The question then arises as to the utility and appropriateness of developing QRA methodologies for use in determining the origin of these variations. From an epidemiological model of public health, four major factors will play important but unknown roles in determining local values of health status: quality of and ease of access to primary health care systems, life styles, the environment (in the broadest sense) and human biology. Developing a systematic approach to understanding the relative local contribution of each of these factors using QRA raises formidable questions concerning the quality and completeness of data, at low levels of aggregation, in a Region extending from Vladivostok to Portugal and from Israel to Norway. As for other HFA targets, monitoring and managing communicable diseases requires QRA to strengthen national programmes for immunization, and monitoring and reducing non- communicable diseases requires QRA to identify specific risk factors and explain their geographic variations. Understanding the way in which habitation, and especially urbanization affects public health requires the development of specific indicators which will permit risk assessment of the wide range of factors involved in producing urban-rural gradients in public health status. The development of the Healthy Cities network of more than 200 municipalities emphasizes the need for standardization and harmonization of RA methodologies if one is to determine how to interpret the effect of urbanization on public health: social class can be shown to account for more than 60% of the variation of health status among various sub populations, at least in those countries where the concept is well-defined. The coming challenge will be to generalize the methods used in countries of western Europe in assessing the sociological contribution to health status for use in countries where the usual definition of social class may not be applicable. RA/RM and Environment and Health Within WHO/EURO, the Environment and Health Service has perhaps the most systematic experience in the use of (Q)RA and RM methodologies in its programme of work. The most traditional area has been within the field of Occupational Health, since in general, most risks, especially for chemicals, have been identified through epidemiological studies of occupational cohorts who are routinely exposed for long periods to high concentrations of substances with potential biological activity. Furthermore, occupational physicians have been the major source of information, based on ad hoc observations, of health effects which appear among small, well- defined populations, which could not be explained by chance: it is these observations which then initiate full scale epidemiological studies with the intent to demonstrate causality, and establish dose- or exposure-response relationships, the final step in QRA. Recent issues have been the health risk of occupational exposure to man-made mineral fibres, and of exposure to fumes containing Nickel and Chromium among welders. 82 ...... 1 9 9 2 Other uses of QRA are in the development of environmental (health) impact assessments: the methodology and implementation of this type of assessment is dealt with by a WHO Collaborating Centre in Aberdeen, Scotland, as an example of the use, by WHO, of specialist institutions for dealing with technical issues for which manpower is not available directly within the organization. Other studies involving health risk of exposure to chemicals have been focussed on high risk populations, such as the effect of Cd exposure on the elderly, lead on children, and PCB's and PCDD's, etc. in human milk. A major continuing study is underway with respect to Spanish Toxic Oil syndrome. A recent major activity utilizing QRA has been the production in 1987 of Air Quality Guidelines for Europe (AQGE) by WHO/EURO at the request of the Netherlands Ministry of the Environment. This dealt with both carcinogenic and non-carcinogenic endpoints for 28 priority substances. Guidelines are considered as appropriate for safeguarding public health and providing guidance for national and local authorities in their risk management decisions. It is not considered appropriate to formulate standards since these are best left to governments and regulating agencies in local context. The WHO procedure is always based on the consultative process, using all available published and peer reviewed information. The use of QRA for carcinogens and genotoxic agents depends on factors of time and place, with practices in North America differing from those in Europe: within the past decade there has also been a shift away from conservatism towards more practical approaches. Because of Congressional demands for risk-benefit analysis, the U.S. Environmental Protection Agency (US EPA) routinely makes quantitative unit risk assessments of the carcinogenic potency of chemical substances. Since US EPA also sets standards, these standards are based on the concept of di minimis risk (less than I per million person years), which is a level low enough to be ignored on an absolute basis, not only when compared with other risks "of everyday life". The WHO approach has been to assume that there are no thresholds associated with public exposures to carcinogens. herefore, since any non-zero exposure is associated with non-zero risk, the definition of acceptable risk must be left to local authorities. As an aid to such exposure setting, carcinogens assessed in AQGE are associated with a potency, expressed as the "unit incremental risk estimate", similar to that used by US EPA. For an air pollutant, this is defined as "the additional lifetime risk occurring in a hypothetical population in which all individuals are exposed continuously from birth throughout their lifetime to a concentration of I/tg/m of the agent in the air that they breathe". Lifetime risks can either be established from human data based on accidental or long term occupational or other exposures, or based on extrapolation from animal experiments using various principles and models. Using the incremental risk model, it is possible to compare carcinogenic potencies for the same route for various substances in order to help set priorities for risk reduction. The concept avoids introducing the notion of "safe" or acceptable limits in the criteria document and leaves this task to the regulating agencies. A major feature of the WHO RA process is transparency: the consensus approach permits the documentation of all steps in the procedure, including the degree to which opinion and judgement play a role in the final assessment. When and if necessary, the procedure can be repeated including new information to permit a revision of guidelines, when appropriate. It is intended to revise AQGE in 1993 as-recent evidence suggests that they may not always provide adequate protection for certain criteria substances. 83 ...... 1 9 9 2 One difficulty encountered in Europe is the lack of uniformity with which regulating agencies apply QRA to their decision making process. In a recent national exercise, the Government of the Netherlands has made an overall assessment of the health impact of priority substances, as part of the development of an environmental management plan for the future. Of particular interest was the assessment that volatile organic compounds (VOC), i.e. polycyclic aromatic hydrocarbons from vehicular emissions, account for less than 612 cancer deaths per year in the country. A similar document produced in Sweden arrived at the conclusion that such emissions would account for 600 cancer deaths per year among their national population. Since the Netherlands population is about twice the Swedish population, and traffic and average population density approximately ten times that of Sweden, the discrepancy in the risk estimates requires some consideration: the origin is the fact that the Netherlands experts used the most probable value for the carcinogenic potency for VOC, while the Swedish report utilized the maximum possible value. Another characteristic of the European scene is the relative heterogeneity of the approaches of national and local governments, international agencies, and industrial bodies in respect to (Q)RA and RM practices. Although the concept of di minimis risk may be universally recognized, some stakeholders still consider zero risk for certain classes of situations as a realistic goal. Similarly the concepts of acceptable risk, unacceptable risk, action level, and especially the recently introduced notion of tolerable risk do not have universal agreement as to their appropriateness or with respect to the absolute levels of risk that these concepts should represent. International governmental and non-governmental organizations are also involved in (Q)RA to various degrees. The World Bank now utilizes QRA in environmental impact assessment, although it is not known to what degree of consistency issues of health and social impact, risk transfer, uncertainty etc. are dealt with. OECD is involved in developing a more concise view of the way in which (Q)RA practices are carried out in its Member States, especially with respect to prevention and control of accidents connected with industrial activities, most of which involve the accidental release of chemicals within and without the fence line. The reason for these activities is manifold: social demand for accident protection, awareness by industry to the economic and social cost of accidents, the availability of new and more sophisticated techniques for QRA, and the lack of uniformity in the application of QRA practices. The regulatory aspects of RA continue to develop, especially within the European Community, with the implementation of the Seveso Directive. A major activity should be undertaken within the European Region to develop a better understanding as to how (Q)RA and RM practices are applied within the Member States and within various sectors. It is clear, at least in the area of occupational health, that a wide range of exposure standards exist: in some cases low standards are established for substances for which there is no national exposure, leading to false comparisons in terms of actual protection provided by local legislation. Such issues are becoming of importance to WHO/EURO as interest develops in harmonizing efforts at assessing the health impact of the environment. Recently, WHO/EURO has started a joint activity with IPCS on "Guiding principles and methodology for quantitative risk assessment in setting exposure limits". In this context, a comparative analysis of current practices in chemical risk assessment (as carried out both at a national level and by intergovernmental organizations) will be conducted, providing a sound basis for the development of harmonized guiding principles for QRA of chemicals. 84 ...... 9 9 2 A major factor governing the introduction, application and harmonization of (Q)RA in the European Region is the question of manpower and manpower development. In North America, the International Society for Risk Analysis, which has a membership of several thousands, is considered by many to be their prime professional organization. In Europe, with a population base three times that of the USA, the Society has only a few hundred members, none of whom consider it as their prime professional organization: professional risk assessors belong either to the engineering societies or national technical bodies, e.g. the Royal Society of England, which have been dealing with the wide range of risk analysis issues for many years. It is estimated that there are only one hundred professional risk analysts in the countries of central and eastern Europe. This fact would suggest the need for a strong international effort in the field of manpower development. WHO/EURO has recently established, at the University of East Anglia, Norwich U.K., a Collaborating Cnter For Environment and Health Risk Assessment and Communication, which is in the process of creating a European Network for Risk Communication. It is felt by WHO that introducing and harmonizing the concept of risk communication will be a significant step in promulgating the principles of RA and RM, especially in countries of central and eastern Europe where there has been little public experience in participating in the decision making process. The European Charter on Environment and Health In December 1989, ministers and other senior-representatives from the Ministries of Environment and of Health from 29 (of the then 32) Member States of the Region met in Frankfurt to ratify the "European Charter On Environment and Health". The Charter is an extension of the European HFA policy and targets adopted by WHO in Europe in 1984. The charter outlines the entitlements and responsibilities of individuals and corporate and governmental entities with respect to informed participation in the process of managing the environment as a resource for health. It outlines a series of principles for public policy, including regulation at the source, the use of best available technology and the adherence to the pollutor pays policy. A series of strategic elements are proposed, including basic principles of comprehensive strategies and the need for research, information and epidemiological surveillance. The development of international programmes of interdisciplinary issues for the Region includes global disturbances, urban development, drinking water supplies, water quality, microbiological contamination of food, impacts of energy and ransport options, indoor and outdoor air quality, persistent chemicals, hazardous waste, biotechnologies, contingency planning for disasters and cleaner technologies. These priority elements are to be dealt with through intersectorial planning, health promotion and international cooperation. Two major themes of the Charter are the need for information and its dissemination and the need for better understanding of the local effects of environment on health. It is understood that knowledge of the true state of the environment is essential for rational decision making at all levels, from the individual to national governments and intergovernmental agencies. Risk assessment plays a role in the management of risks at all levels. This was made clear by the individual pleas from the majority of the ministers present at Frankfurt: they stated that they were faced with the evidence that environmental degradation was noticeably affecting the health of sensitive populations, and gave WHO/EURO the mandate to determine how, where, and to what extent, these effects were occurring. 85 ...... 1 9 9 2 The European Centre for Environment and Health As a direct result of the process that led to the Charter, the Governments of the Netherlands and Italy initially responded by providing the wherewithall. for creating within WHO/EURO the European Centre for Environment and Health (ECEH), with divisions at Bilthoven and Rome and a coordinating unit at Copenhagen. The initial task assigned to ECEH is the production of a report. "Concern for Europe's Tomorrow" (CET). The first edition of CET is intended to be an overall presentation of the state of the environment describing national conditions related to Water, Food Safety, Radiation, Multimedia Exposure, Air Pollution, Housing and Occupation and the state of Public Health, using harmonized methods of data gathering. It is expected that CET will make use of elementary models of QRA to estimate the potential for public health risks due to environmental burdens, and where possible, estimates of population exposure. For some areas it will be possible to make population-based estimates (e.g. for exposure to outdoor air pollution episodes and to environmental radiation) of risk which will allow a definition of expected areas of high impact (i.e. subriational hot-spots) and permit an estimation of the size of the potentially effected (i.e. exposed) populations. Since risk assessments will be used for national and local policy making, emphasis will be placed on the provision of alternate scenarios and of estimates of reliability and uncertainty. The exercise involved in CET will also serve to introduce questions of data availability and data quality in the area of environment and health in the European Region, with special emphasis on the countries of Central and Eastern Europe for which there is little public experience in either accessing or assessing relevant data. Furthermore, it is in these new Member States that technical cooperation and support for dealing with the legacy of the past decades of environmental neglect must be urgently provided. The magnitude of the task is illustrated by the fact that there are now, in mid 1992, 44 Member States in the Region, and 14 new nations are prospective members, to be compared to the 32 Member States in 1989. This is also an opportunity for the introduction of harmonized and reliable methods of QRA and RM in the many sectors involved in issues of environment and health. The challenge facing WHO/EURO and ECEH in this respect is the development of a catalog of appropriate methods for (Q)RA and RM, and for guidelines for priority setting and their implementation. Emphasis will be given to the frequently neglected areas of estimation of uncertainty, societal (e.g. social and economic) costs, and risk transfer. Of equal importance is the opportunity to introduce methods of assessing the effectiveness of the techniques used, since the ultimate goal of these efforts should be the identification of priority issues on a local basis, and provision of appropriate decision support to national and local policy makers charged with the responsibility for reducing the impact of environmental conditions on public health. In this respect, two issues will deserve attention: harmonization of the basis for QRA for estimates for public health impact and o the methodologies for dealing with competing risks. In the first case it is necessary to reconcile, in the case of chemicals, the extreme positions of either using toxicology and cross species extrapolations and accepting the inherent uncertainties in the chain of estimates, of relying on analytical epidemiology to estimate dose- (or exposure-) response relationships with their inherent uncertainties, or of considering the significance of biological indicators of exposure and other developmental or unvalidated methods such as molecular epidemiology". In the second case, the issue will revolve around risk trade-off, the 86 ...... 1 9 9 2 need for substainable development of e.g. the industrial sector, and the general question of public risk perception and "how safe is safe enough". Conclusions Risk assessment has not been a methodology which has been applied systematically to the overall problem of the management of public health. Although the techniques of (Q)RA have been developed and applied with success in preventing loss in the Chemical Industry for many years, it is only within the last few decades that the economic benefit from their systematic use has been appreciated. Public health costs have rarely been included in traditional uses, although QRA techniques are commonly used in environmental impact assessment and urban and industrial development and planning in various parts of the world. The goal of WHO is the introduction of harmonized reliable methodologies of (Q)RA, characterized by transparency and respect for human values, in the decision making process in Europe. As time progresses, it is becoming evident that there is less and less freedom of choice in the management of the environment, and especially in the management of the environment as a resource for human health. This freedom becomes restricted because more and more issues become visibly critical, and compete for priority for resource allocation. The real challenge becomes that of distinguishing real risks from perceived risks, since frequently it is the perceived risks, regardless of their magnitude, which receive resources at the expense of the actual threats to health and wellbeing. The issue, especially for WHO, is made complicated by the recent changes in the political, economic and social structure of the Central and Eastern part of the European Region. These changes have had a profound effect on the urgency for which priority issues must be resolved, not only in that part of the Region, but also in highly industrialized Western Europe which must now share its limited resources in sectors for which there was little concern just tree years ago. Where starvation is a threat, the protection of the global environment appears less important than survival. Where unemployment is a threat, issues of occupational health and safety become unimportant compared to the need to maintain employment. Where poverty and disfunctions in social structure lead to direct health risks, the local state of the environment is not a direct concern. The burden of proof that priority issues represent a real threat to human health rests with WHO in its capacity of advisor to national policy makers and provider of decision support tools. We have recently been made aware that we inhabit a unique and fragile planet, and that we in Europe share a Common European House. The preservation of the health of this "house" will depend to a great extent upon our success in further developing and applying appropriate and effective tools of risk assessment and risk management in the near future. References Our Planet, Our Health: Report of the WHO Commission on Health and the Environment. WHO Geneva, 1992. 282 pp WHO/EHE/92. 1). Targets for Health For All. WHO Regional Office for Europe. Copenhagen 1985, 201 pp. Environment and Health: The European Charter and Commentary. WHO Regional Publications, European Series No. 35, WHO Regional Office for Europe. Copenhagen, 1990. 154 pp. Air Quality Guidelines for Europe, WHO Regional Publications, European Series No. 23, WHO Regional Office for Europe, Copenhagen, 198Z 426 pp. 87 ...... 9 9 XA04NO310 An OECD Perspective of the role of Risk Assessment in Policy Development Dr. Jim Brydon Head, Environmental Health Safety Division, Organization for Economic Co-operation and Development 1. Introduction OECD is an intergovernmental rganisation bringing together 24 industrialised countries from North America, Western Europe, and the Pacific. Its basic aims include the following: • to achieve high sustainable development, economic growth and employment; • to achieve high economic and social welfare and a high standard of living throughout the OECD area and in non-Member countries: The specialised Agencies and Directorates of OECD cover the full breadth of economic and social activities of concern to the Conference. Under their programmes, there are a variety of activities which involve various elements of qualitative and quantitative risk assessment. Risk assessment methodology, policies options regarding the use of risk assessment, the role of risk assessment in policy and decision-making are all routine in OECD work. This work ranges from, for example, work on the economics of investment policies, through work on food safety, to the analysis of nuclear safety technology. 2. The Role of Risk Assessment in Policy and Decision Making The basic aims of te OECD, which I outlined above, reflect the, fact that OECD and its Member countries are committed to strengthen current work, as well as undertake new initiatives, with the aim of enhancing economic development. It has become clear, however, that global economic and development decisions should not, and can no longer be made in isolation from environmental, health and social welfare considerations. This fact is still fresh in our minds particularly after the recent United Nations Conference on Environment and Development. Consequently, decision-makers in government, industry and public interest groups, must increasingly consider questions concerning sustainability and environment quality, including human welfare in their policy work. But without adequate information to assess the potential risks associated with alternative development possibilities, reliable policy and decision-making is not possible. In addition, without adequate information the degree to which various options might be sustainable into the future cannot be evaluated. In any case, we can no longer assume that we can always cure problems in either the world economy, or the environment, simply by reacting to them as they occur. These decision-making trends have exacerbated a perennial problem, that is, scarce resources. The resources of international organisations such as OECD, as well as those of Member countries, are limited. We need, therefore, to be able to focus our attention on those issues 88 ...... 1 9 9 2 which are of highest priority. It is my belief that an increased knowledge of risk assessment, and its associated disciplines, have become increasingly important in the work of the OECD and its Member countries, partly because they enable us to set priorities in an increasingly meaningful manner. Personally, I would like to see this process go further, and look forward to these disciplines playing an increasing role in the business of setting priorities. In fact, I expect that we each have a number of expectations from tis Conference. Certainly, it sould provide a "snap-shot" of risk assessment as it is currently practised in a variety of sectors. But beyond that, I hope, as I just implied, that we will see, the start of an analysis of the future needs with regard to the development of appropriate risk assessment methodologies, in relation to policy needs. As an illustration, I would like to pose a number of questions and make some comments about each. These questions are not intended as a comprehensive "wish list" of future needs. On the other hand, if we had answers to some of these questions, I feel we would have come a long way. Certainly, these are the types of questions which must be addressed in the future: (i) My first question is: how can we better cope with the situations we usually face, namely those in which we have limited information and much uncertainty? As I understand it, Dr. David Fisk, in particular, will be dealing with this question at the Conference. In 1991, the Group of Economic Experts of the OECD's Environment Policy Committee began dealing with the role of uncertainty in decision-making in the area of development and implementation of policy for environmental protection; such work is attempting to mix objective scientific assessment and the need for subjective judgement. (ii) My second question refers to a special case of uncertainty: how can risk assessments be linked to, and supportive of, the precautionary principle? One of the implications of the precautionary principle is that one might wish to take action before one has exhausted the research effort; in fact, risk assessment can sometimes lead us int6 a trap where its results will lead to a conflict of precautionary action versus further research; so one goal might be to try to develop some guidance for determining wen prudent action should be taken immediately and when it should be deferred; (iii) My third is, do we de-emphasize or delay action on low probability/high consequence events; and focus on the high probability/low consequence events? (iv) If so, what do we do about high probability/low consequence events? Do we develop, for example, alternatives for only the most "dangerous" situations or do we attempt to promote a general shift to cleaner technologies to cover broad categories of events? Following from these questions, I draw your attention to the Annex of this paper concerning the OECD's Nuclear Energy Agency and its programme on probabilistic safety assessment. 89 ...... 1 9 9 2 3. Participation in the Process I would also like to say something about participation in the risk assessment process. For risks to be effectively managed, I believe that it is a prerequisite that government, industry and the public are able to understand the sources of risk as well as the priorities in managing them. Many risk assessment methodologies are sophisticated and also generate large quantities of data which can be difficult to present and comprehend. New computerised systems which improve access to information offer only part of the solution. But a key aspect is to increase participation in the risk assessment/risk management process. There are a number of key groups who must be involved, including those from government, academia, science, industry and the public. Recent experience in the OECD Chemicals Programme on the co-operative investigation and risk reduction of existing chemicals has shown that the participation of industry is providing valuable input into the work. In the commercial and product life-cycle analysis being conducted as part of the development of risk reduction strategies on lead, cadmium, mercury, methylene chloride and brominated flame retardants, the knowledge of industry concerning the products and their end uses is essential. On the other hand, the initiation of these activities by the governments of OECD Member countries has increased the awareness of several industrial sectors of the need to improve their technologies and their products. Switching to cleaner technologies or developing new cleaner or more recyclable products, and switching to alternative materials are examples (based on relative risks). Of course, the ultimate aim of risk assessment is the identification and reduction or elimination of those risks which are inconsistent with sustainable economic growth and development, and with environmental health and safety. There will be a high-payoff for work that improves, streamlines, and increases the transparency of approaches to risk assessment and links their results to innovative, priority-based and effective methods for risk reduction and management. Following some of the dialogues opened by this and other international conferences and the many new initiatives of government, industry and international organisations a new era in risk assessment and management will encourage increased participation in the risk assessment process. 4. Concluding Remarks It is my impression that there has been in recent years, a renaissance in the field of environmental risk assessment. Of course, we are all keenly aware of the many analytical shortcomings, as well as a lack of information which has often given ambiguous even contradictory results. On the other hand, we know that many new concepts, methods and approaches are emerging. I hope that this conference will help us in the continuing evolution of the approach, so that ultimately, risk assessment practices will lead to a more rigorous policy development and decision-making. 90 ...... 1 9 9 2 ANNEX Risk Assessment Programmes at the OECD's Nuclear Energy Agency For the Safety of Nuclear Power Plants PSA techniques were originally developed in order to assess the risks to the health of populations from the operation of power plants. They are now becoming more important in supplementing the traditional deterministic approach to determine potential weaknesses and to assess the safety level of the plants. Presently, about two hundred probabilistic studies have either been completed or are in progress. PSA experts of the Principal Working Group No. of the NEA's Committee on the Safety of Nuclear Installations are deeply involved in the process of reviewing the methodology, application and maturity of this emerging safety tool. Methodology This is the main direction of the group's activities. Most recently, a workshop was held to provide a forum for exchanging information on the experience, development and needs in the most problematic areas which include analysis of dependencies, time-dependent phenomena, human errors, uncertainties and external events. Some conclusions from this workshop are given below: In the area of analysis of dependencies, Common Cause Failures (CCF) are still a concern due to scarce data in this field and a procedure of framework for acquiring common cause data as well as allowing to exercise engineering judgement more consistently should be achieved. For time-dependent phenomena, components may be noticeably affected by "ageing" or "learning". This effect varies with the type of component, its function at the plant as well as the maintenance programme, therefore plant specific evaluations are required. In the field of human errors there is a need of reliability data to analyse knowledged-based actions especially those related to non-full power and shut-down situations and accident management. A very important aspect is the evaluation of operating experience and simulator experiments for obtaining human error probabilities. The problem of transferring simulator experience to real situations was believed to be solved to a large extent by using experience gathered in experiment situations. Expert judgement is also believed to give good results in this context. For those reasons, human error is still considered as an area which requires research in order to arrive at more reliable PSA results. Concerning the task of quantifying uncertainties, there is a general agreement that the uncertainties affecting reliability data could be treated satisfactorily. However, some weaknesses in the handling of modelling uncertainties have been recognised and an exploration of a possible theory for treating uncertainties is deemed desirable. For external events, it also appeared that uncertainties of results are still substantial and that for analysing plant responses to seismic loads considerable expert judgement is still needed. Application The development of PSAs has resulted not only in an increase in the number of analyses performed but also and more importantly, in expansion of their scope of application. A study, published by the OECD Nuclear Energy Agency in 1989, entitled Probabilistic Safety Assessment in Nuclear Power Plant Management demonstrated the benefits afforded by PSA in the management of safety in NPPs. According to the authors, the implementation of PSA will reduce the frequency of significant incidents and accidents. A wider use of PSA in the 91 ...... 9 9 2 process of accident management for the prevention of severe accidents as well as the mitigation of their consequences can be expected in the near future. A new report published by the NEA entitled Living Probabilistic Safety Assessment for NPP Management describes recent international development in the use of PSA. Recent applications of PSA techniques have demonstrated their unique ability to assess proposed modifications and engineering configurations to operating plants. This ability to monitor the impact of design and procedural changes that improve the overall safety makes a "living" PSA programme a powerful tool for supporting decisions that affect plant safety and foster understanding between the utility and the safety authorities. A workshop was held in May 1992 where it was possible to observe the progress in the development of living PSA programmes in most of the OECD countries which should fairly soon lead to many applications. The Future Experts agree that establishment of generalised criteria for the acquisition of reliability data is needed. Supposing those criteria would be adopted internationally, an intercomparison of data would then become possible that would give a better guarantee that generic data would really apply to components under consideration. This approach is needed not only for active mechanical and electrical components, but also for passive components and human actions. Despite the current limitations, the practical experience gained from the application of PSA methods and the confirmation of PSA results by comparison with operating experience data, has led to the emergence of PSA as an essential role in the field of nuclear safety, in combination with the deterministic methods and defence in depth. Some countries are already using PSAs in the licensing process for their latest plants and several others are planning to use them for future plants. It is only a matter of time before PSA becomes a universal fixture in the licensing and regulatory process. For Environmental Protection and Radioactive Waste Management Risk assessment is also an essential element in other areas of nuclear activities, such as the evaluation of the potential impact of nuclear accidents and the evaluation of their consequences and the safe disposal of radioactive waste. Radiological Risk With regard to the evaluation of the radiological consequences of nuclear accidents, NEA has prompted an intercomparison exercise on probabilistic consequence assessment codes (PCA) with the objective to: • contribute to PCA code quality programmes; • guide future developments in the PCA field by identifying the merits and appropriate use of different methods; • enhance the general appreciation of the applicability of PCA codes by those who develop and use them, particularly in decision-making and regulatory contexts; 92 ...... 9 9 2 • provide a forum for discussion on various international approaches to PCA model and code development, and to encourage harmonization of codes; • produce a report on the exercise which will act as a basic PCA code comparison reference. Another area where the activities are rapidly developing in the field of radiation protection is the application of the new ICRP recommendations to the control of the so-called potential or probabilistic exposures, namely those exposures to radiation that can result from accidental events. In this field, the NEA is beginning to study the possible approaches and criteria for the management of potential exposures in the operation and the regulation of nuclear facilities. In addition to the above, it is also expected that the exercise will generally increase the understanding of uncertainties in PCA codes and thereby assist in the identification of priorities for future research. Risk Due to Radioactive Waste In the field of radioactive waste management, the safety of any proposed disposal system must be convincingly shown prior to its implementation and risk assessment studies are therefore the most important part of licence applications. In particular, safety assessments provide the principal means to investigate, quantify and explain the long-term safety of a selected disposal concept and site. Safety assessments consist usually of the following interrelated elements: • identification of the possible future evolution of the system or scenario development; • development and application of appropriate odels; 40 evaluation of potential radiological consequences in an integrated assessment; • uncertainty and assessment analysis; • validation and review of all components of the assessment; and • comparison of the results with safety criteria. Over the last decade, many safety assessments have been conducted for various purposes, such as assisting in the understanding of the long-term behaviour of a disposal system, quantifying this understanding and making predictions about the future, preparing licence applications or orienting research activities in the areas of significance. Risk assessment methodologies in the nuclear waste disposal field have now reached a certain degree of maturity and sophistication. There is a consensus at international level that reliable safety assessment methods are available today to evaluate the potential long-term radiological impact of nuclear waste repositories (Ref. 1 2 3 and 4 The improvement of these methodologies is still one of the main priorities of the programme of NEA, notably the modelling of the transport of radioactive substances through the geosphere and the biosphere, the use of probabilistic techniques, the quantification of uncertainties in risk assessment and the identification of potentially disruptive scenarios, including human intrusion at disposal sites. 93 ...... 9 9 2 References 1. "Can Long-Term Safety be Evaluated? - An International Collective Opinion", NEA/IAEA/CEC, Paris 1991. 2. "Review of Safety Assessment Methods", NEA, Paris, 1991. 3. "Disposal of Radioactive Waste - The International Probabilistic System Assessment Group: Background and Results", NEA, Paris 1991. 4. "Safety Assessment of Radioactive Waste Repositories - Systematic Approaches to Scenario Development", NEA, Paris 1992. 94 9 '9 XA04NO311 Risk Assessment A European Community Perspective R. Haigh Head of Industrial Medicine and Hygiene Unit Health and Safety Directorate Directorate-General Employment, Industrial Relations and Social Affairs Commission of the European Communities, Luxembourg Introduction The world is a risky place in which to live! The world tolerates that 750,000 deaths occur on the roads each year. Society has not yet come to terms with the added burden that urbanisation brings to developing countries. Pollution from the use of fossil fuels creates incalculable loss to the world's environment and to the health of its inhabitants. The misuse of chemicals provokes suffering and deformity. In the European Community alone, over 21 million tonnes of toxic waste have to be treated each year. Of course, there are different types of risks: individual and societal. Individuals continue to travel by air in defiance of terrorists or faulty machinery. Whilst society urges caution in diet and nutrition, the individual is probably more worried about food additives that he is about eating too much or making a rigorous appraisal of the value of his diet! As the Conference progresses many people will die from the causes of malnutrition, from war or societal neglect while we, individually, will be more at risk from overeating. In other words, we perceive risks in a multitude of ways. We tolerate these risks according to our perception of what we feel is acceptable without carrying out scientific assessment of the relative severity of those risks. If applied at a governmental level, this subjective tolerance can lead to unnecessary burdens or constraints that are disproportionate to the risk. Clearly, this is not acceptable for policy makers. We have just seen te closure of the UNCED World Conference on the Environment in Rio de Janeiro, where the absolute need for more effective cooperation in the protection of the environment and the world inhabitants was convincingly demonstrated. The European Communities already coordinate risk assessment with its twelve Member States in a large number of areas and is increasing its international cooperation. We have recognised that it is no longer possible to carry out effective risk assessment in one country alone or to have good risk management by forgetting the next door neighbour! The role and application of risk assessment The Report of the WHO Health and Environment Commission's Panel on Industry records that 6risk assessment" is a developing science and art with the goal of estimating the probability of occurrence of unwanted events. The Panel continues that risk assessment may provide a wide range of estimators and that uncertainty is the predominant feature of risk assessment procedures as they are described today. Risk management is a societal judgement process which 95 ...... 1 9 9 2 complements risk assessment. Despite its limitations in the current state of the art, for policy makers in government, risk assessment has a very substantial contribution to make in providing a more balanced and open basis for decision-making. Risk assessment is a useful tool for setting priorities, for underpinning an effective risk management programme and for evaluating that programme, and also for achieving a better and more accurate public perception of the risk by communicating information about that risk. Risk assessment does not imply that risk exists, but simply that it is necessary to establish, by some means or another, whether a risk exists. Risk assessment can range from a simple determination done in a few minutes to an extensive review using one of the various methodologies for risk assessment, that have been developed. Risk assessment is an integral and important part of a three-stage process, involving: • hazard identification; • risk assessment; • risk management; which is at the centre of decision-making. The EC, which is in its concept supranational, applies risk assessment widely throughout its programmes and initiatives. A few examples from Community action will suffice to indicate the scope for its use and the differences in the techniques of risk assessment as they are applied in practice. - Community legislation for the Internal Market places obligations on the manufacturers and suppliers of products. For example, with the completion of the Internal Market at the end of this year, manufacturers of machinery will be obliged to make a risk assessment of their product in foreseeable uses and to eliminate risk at te design stage as far as this is possible. As part of the process of obtaining an "EC" mark for supply within the Internal Market, it will be necessary to show in the assessment process that essential safety requirements have been met. - For chemicals, extensive rules exist in the Community for the protection of man, whether as a worker or as a member of the public, and for the environment. Thus, in the framework of the rules on classification, packaging and labelling based on te hazards of chemicals placed on the market and in the proposed notification scheme for new chemicals to be placed on the market, requirements are being introduced on the submission of toxicity data for the assessment process so that the risks can be more properly assessed in a scientific and quantitative way. The requirements of this legislation are complemented for groups of politically or technically sensitive chemicals by specific legislation englobing azard identification, and risk assessment and management. This more specific control is valid for pharmaceuticals or food additives or pesticides where extensive scientific data, specifically designed 96 ...... 9 9 2 to reflect the particular use of the substance, are assessed by specialists as a prerequisite to its authorisation, and as will be explained later, further in-depth analyses may be required for setting Community workplace limit values. - Protection of the public, for example, from chemicals, from other agents or from defective products, is also a subject that is addressed at Community level and which draws on the results of risk assessment. The ultimate, and often controversial, final step may be prohibition where the risks are deemed unacceptable. In other cases, the risk assessment may lead to specific restrictions or safeguards. Within its public health actions, the Community is addressing several risks of major health scourges in Europe, such as heart disease, cancer and the problems of alcohol, tobacco smoke, drugs and AIDS. - For the environment, a subject of increasing societal concern, risk assessment is again a helpful tool in evaluating the various options for action to safeguard the quality and integrity of air, soil, water and of the flora and fauna. However, in many of these areas, te scientific knowledge and data available are often inadequate to provide a basis for more than a qualitative approach to risk assessment. - For the workplace, in contrast to rules developed in many areas of Community activity, EC legislation has put the obligation on the employer to assess risks at work and in the working environment, and to take the necessary action. In order to meet these requirements, he has to identify te hazards and decide on the measures to take which will provide the necessary protection of the health and safety of his workers. These measures may result in the replacement of a hazardous chemical or biological agent by one which is less hazardous; by improving process control; or by providing protective equipment to prevent exposure. Although a machine might have the "EC" safety mark and fulfil the appropriate standards, the employer is still obliged to address the risks relating to the actual use of the machine in his workplace. This assessment may result in the replacement of the machinery in question. The legislation also requires medical surveillance in certain situations for exposed persons where the risks cannot be managed without it. In the area of biotechnology, as for chemicals, these workplace constraints are complemented by detailed specific programmes and the Commission has been centrally involved in the European debate about the conjectural risks of biotechnology and their management. Scientific opinion has been critical of the focus on the supposed risks inherent in the techniques and effects of DNA recombination per se; arguing that the risks of chemicals are assessed, not of "chemistry". But public and political attention has been alerted by publicity surrounding the dramatic progress in the life sciences and technologies, and the assessment of any risks as might be affected in consequence. There is apprehension about the possible effects of applying such powerful tools to the familiar and value-laden processes associated with all forms of life, health and growth. 97 ...... 1 9 9 2 Environmental and health risk assessment of biotechnology is complex and uncertain and still in the process of development. It is not easily comparable with more well-established methodologies such as those for chemical risk assessment, and the Commission, through its various concerned departments, has initiated a number of measures to ensure that the risks can be assessed whilst maintaining a competitive environment for industry. The European Community has progressively developed an adaptive regulatory framework for biotechnology and, in parallel, efforts in risk assessment research have built up. These efforts are planned in consultation with the Member States, some of whom also have active programmes. Results of this research are regularly published. In this area, the Commission also participates in international fora such as the OECD expert group on safety in biotechnology, and has regular bilateral contacts with scientists in other areas of the world with comparable interests and programmes (e.g. USA). The benefits of risk assessment and a perspective for the future The challenges which governments have to face are many. Inevitably, choices have to be made between competing priorities. The same can be said at the level of the individual employer or enterprise. What risk assessment offers is a tool to help in making these difficult choices and a basis on which to explain to others the decisions taken. Inevitably resources are limited. Inescapably, therefore, to be effective, international cooperation is essential. The Community, through the expertise of its component Member States or indeed through the European Commission is forefront in cooperating at international level in assessing the hazards and risks of a multitude of offensive situations. It contributes to the management of those risks. One needs only to recall the Community's contribution to OECD, to the WHO, UNEP and ILO programmes on chemical safety, with FAO on agricultural matters, with UNIDO, and with Council of Europe to rebut any criticism of that commitment. The Community is assisting the emerging nations in Central Europe. EFTA countries are already enjoying privileged relations with the Community. Risk assessment is not a replacement for the societal decision which leads to the management of risk. In itself it does not solve the problems and, in many cases, social, political, economic or other imperatives will result in final decisions, policies or choices, which may, in risk assessment terms, seem illogical or incoherent. But that is not to decry its contribution to the assessment process which has permitted a more open and structured debate on risk acceptance and cost benefit of common actions in the EC or in international fora. Better identifying areas of need helps to promote a more proactive approach to prevention. In the future, risk assessment and especially quantitative risk assessment will become increasingly important wen establishing priorities and legislative provisions. It follows that legislative actions must be subject to adaptation to technical progress and the advance of scientific knowledge, and as risk assessment techniques are refined. Already a number of needs to be addressed later in the Conference can be identified. 98 ...... 1 9 9 2 As a first priority, there is always a need and a call for more and better data. The evaluation can only be as good as the data on which it is based. Within the Community, we are addressing this need in a number of areas such as improved toxicity data on existing chemicals, better accident and ill-health data, more uniformity in data systems so that data is compatible from different sources, and so on. Techniques of risk assessment are frequently still of a rudimentary nature, not only because the data is lacking but because the understanding of how to go about assessment is not well understood, or because its pitfalls and limitations are not well appreciated. We need to learn from successes in one area and apply them in others, and we must avoid over-elaboration and overemphasis where the benefit does not justify the means. This Conference will focus attention on the need for more realism about the acceptance and limitations of risk assessment; it will catalyse progress by showing the way to obtain better data and by directing experts to more refined techniques which reduce uncertainty. Perhaps more importantly, it will, by its existence, demonstrate to the world the recognition by experts, by governments, by te European Community in this European Year of Safety, Hygiene and Health Protection at Work that there is international concern that decisions are not taken on an ad hoe basis but on the basis of a proper scientific assessment of the real risks. I would like to close with a quotation from Lord Rothschild in a recent BMA publication on risk assessment: "There is no point getting into a panic about the risks of life until you have compared the risks which worry you with those which don't." Thank you. 99 9 19 XA04NO312 Risk Assessment and the Environment D. J. Fisk Chief Scientist Department of the Environment UK Introduction This paper reviews the use of risk assessment techniques in the field of environment protection. I will argue that in some important instances the development of environment policy has been a source of fruitful development of a risk based methodologies. In other cases the importation of risk assessment techniques has proved much more problematic. As the scope of environmental regulation increases so does the possibility of inconsistent and arbitrary solutions to problems. The need for a more systematic approach to the development of environmental regulation has never been stronger, so it is important to understand the reasons for the mixed success of risk assessment. This applies equally to those nations with long traditions of the regulation of private sector industry and those just beginning on this course. The way ahead may be to extend our ideas of how to express risk and uncertainty. Some of the recent cause celebres of environment policy show this challenge very clearly. As an example, this paper will look at the problem of assessing the risk of man-made climate change. Defining "Environment" It is not surprising that environmental regulation severely tests many of the well tried decision making approaches, because it embraces such a wide field of activity. Einstein is reputed to have defined the environment as "everything that is not me". This certainly reflects the breadth of the problem - can any single decision making methodology hope to embrace the whole of environment policy? However this definition misses a further important ingredient - that the resolution of environment policy is an intensely social and political process. One distinguished commentator has paraphrased the Einstein definition as "the environment is everything that is not my fault". Tongue in cheek or not, this definition reflects tat people may have very different views as to damage and risk of damage, dependent on their relationship to the environmental issue. This will emerge as I look at some examples. Classes of Regulator Issues It is convenient, when analysing the success and limitations of risk assessment in the environment, to break down the regulatory problem into sub-categories. Four possible categories of regulatory context suggest themselves: - process safety e.g. regulation of the process of safe disposal of waste to land - incidental exposure e.g. controls on marketing and use of chemicals to limit general exposure of the environment to dangerous chemicals - once off events e.g. assessment of the impact of a civil engineering project on a natural habitat 100 ...... 1 9 9 2 - "continuous accidents" e.g. permitted discharges to rivers I will look at risk assessment in each of these categories in turn. I will follow the convention on terminology hazard to mean the property of a substance or process that has the capability to do harm; hazard analysis to mean the identification and quantification of those hazards; consequence analysis to mean the identification of the consequence of the harm being realised; risk analysis to mean the identification of the probabilities of such consequences occurring, and risk assessment to cover the whole assessment process. The natural place to begin is the risk assessment of environmental process safety since this is the origin of risk assessment methodology. Use of risk assessment techniques - Process Safety An industrial accident can lead not only to harm to people but also damage to the environment. The Seveso, Sandoz and Chernobyl accidents led to widespread environmental contamination. However it is only recently that environmental as well as human health and safety considerations have been widely factored into the regulation of hazardous installations. Before this recent development the main issue of process safety to merit the special concern of the environment regulator has been the process of disposing of waste. Under most regulatory systems of waste disposal, the process is intended to be safe. Damage occurs when an unintended event, such as the failure of a lining or other method of containment occurs. By far the most developed process risk assessment in waste disposal policy is the assessment of the disposal of radioactive waste. The development of risk assessment techniques for radioactive waste disposal undoubtedly owes much of its origin to close proximity in the public eye to te highly developed assessment techniques used elsewhere in regulating the nuclear industry.(i) As typical studies show(2) assessing radioactive waste disposal does represent its own problems in risk assessment. One of the most challenging aspects is the long timescale over which the risk is to be borne by society, perhaps 100,000 years. Thus while a typical industrial risk analysis might reflect a relatively high risk over a short period to an individual, the regulator of radioactive waste disposal has to assess exposure not only to society, but to many future generations. What is more, different engineering solutions to radioactive waste disposal may give the same time average risk but distribute it quite unevenly in time. Despite this difficulty some choice has to be made. Even if no more nuclear power stations were to be built, the world has still to find a home for the spent fuel and radioactive waste generated by the operation of some 400 installations, once the potential for the recovery of further nuclear fuel from radioactive waste is exhausted. The hazard 101 ...... 1 9 9 2 is realised and the regulator has to find the approach which will reduce levels as low as reasonably achievable, economic and social factors being taken into account. A similar impetus applies for toxic wastes. Once a toxic waste is produced a process has to be found for disposal. Assessment techniques are less developed in this field. This difference in part reflects the very high value added of radioactive waste disposal which can justify extraordinary efforts to assess geological and other factors at a site. Studies of this nature are seldom feasible for more ordinary waste disposal processes. However hazard assessment of classes of waste is a common regulatory approach which is used to control the final waste stream to which a waste may be directed. Unlike radioactive wastes, most toxic wastes are by their nature reactive and transformed by the disposal process. A well engineered site is effectively a slow bio-reactor. The greatest risks are then either through improper disposal, or from the products of the transformation for example atmospheric micro-pollutants from incinerator stacks or toxic leachates from landfills. Many of the current problems of land contamination now emerging as a particular difficulty in Central and Eastern Europe, are a direct result of failure to identify and deal with the risks of toxic waste disposal. Waste regulation has become an increasingly scientific and tightly controlled enterprise since the 1970's in much of the developed world. As it develops further there will be a need for more structured hazard and risk assessment methodologies. Some materials may however be particularly difficult in the waste disposal process, and the regulatory response may choose to prohibit the manufacture of the chemical in the first place. Recent examples are PCB's and asbestos. This leads naturally to my second class of regulatory function, the control of entry into the market place of substances that could be hazardous through incidental exposure to the environment. Use of Risk Assessment - Incidental Exposure There are supposedly some 100,000 man-made chemicals on the European market. New chemicals arriving on the market are required to be notified to the competent authority, and have associated with them a number of tests which indicate their possible impact on the environment. Notification is not strictly a control procedure. It simply establishes the prima facie case for the hazard assessment. In judging whether further controls are necessary regulators employ a number of different approaches. For some substances there is a clear dose response relationship, with an identifiable threshold concentration at which effects occur in the indicator tests. The regulator might then employ a safety factor to judge the significance of the chemical. This can provide a rough and ready means of comparing the scale of realised hazard. For example, the Royal Commission on Environmental Pollution, in making its recommendation on the withdrawal of lead from the environment, noted that the safety factor between realised blood lead levels and those producing frank symptoms of lead poisoning was smaller than for many other chemicals.(3) It is also not uncommon to use an analysis of the possible pathways by which a chemical once released into the environment could reach its target. Environmental lead to which I have already referred, can reach the bloodstream through contamination of water, contamination of food or inhalation. Dioxins can reach a human target by many different routes.(4) The total dose will be the sum of the contributions from these different pathways. 102 ...... 1 9 9 2 Pathway analysis is therefore very important for the control of persistent chemicals which can accumulate in the environment. A top predator such as a bird can accumulate dangerous chemicals accumulated by its prey across its whole habitat. In some ways a pathway analysis is analogous to the fault-tree analysis in traditional risk assessment. Since we are considering continual dosing of the environment, "failure rates" are now replaced by expected average dose rates. The main common element between fault-tree analysis and pathway analysis is the need for a rigorous investigation of the possibilities that could constitute pathways. A special case of the control of the environmental effects of chemicals is the regulation of pesticides. By definition pesticides are both actively distributed into the natural environment, and are designed to be harmful to at least one living creature. A commercially successful pesticide can be exposed to the environment in very large quantities and over many years. Persistence in the environment, and accumulation in the food web are therefore real hazards. For this reason consideration of environmental safety has been added to the existing regulation of pesticide use for food safety. It is clear that risk assessment of chemicals has a firm grounding in hazard identification and analysis. In principle it could include a greater element of probabilistic analysis in its risk analysis of the consequences. However reducing the average consequence is not always the thrust of the relevant legislation. Thus is might be possible to calculate the probability that a fisherman liked to fish in a river that was contaminated with mercury, that he was successful with his catch, that he was married, that his wife was pregnant, and that she shared his enthusiasm for a fish diet. However legislation more frequently focuses on the protection of a sometimes hypothetical "target" (or critical group in the case of radiological terminology) thus short circuiting, for good or ill, much of the meat of a risk analysis. It is in effect the allocation of a right to the target group concerned. In my example to pregnant fisherman's wives. Releasing a new chemical also poses a different type of hazard. The regulatory test procedures may simply have not anticipated something. It is often remarked how difficult it would have been to have anticipated the effects of CFC's on the stratospheric ozone layer at the time that they were engineered. Of course it is possible to look at the performance of the regulatory regime retrospectively, and fortunately, as we would expect, the likelihood of error is decreasing. A novel form of hazard assessment is associated with the regulation of the deliberate release of genetically modified organisms.(S) Since by definition the phenotype, or outward expression, of a transgenic organism is not finally characterised until the organism is released, the regulatory system has to judge the appropriate step by step approach by which to manage any risk. This risk can have a probabilistic content, for example, a containment process might fail. However, much of the difficulty lies in judging the uncertainty in the full implications of the release. Incidental contamination of the gene pool through horizontal transfer of genetically material might occur. As with introductions of alien species, freak competitive advantage may enable the organism to spread rapidly to the disadvantage of natural competitors. Various proposals have been made to address hazard assessment in this context.(6) This technology is likely to become a major part of world traded products, and it is therefore essential that international agreement is reached on the general approach to biotechnology safety. The difficulty facing risk assessment of chemicals is a very diffuse understanding at the time of placing a substance on the market of its likely fate in the environment. This limits the use of quantitative risk analysis. For a new chemical there would simply be no data on which to quantify 103 ...... 9 9 2 the risk it posed let loose in the world. However the overall risk of the release can be reduced by monitoring the subsequent changes in concentration in the environment. It has often been through monitoring (for example of DDT in species in parts of the world remote from its use) that has led to a reassessment of a chemical. Risk management therefore complements risk assessment, and in part reduces the problem of determining the mechanisms and pathways of the chemical. This is especially true for the major reassessment of existing chemicals now underway in the OECD. Risk management with continuous monitoring and response. is seldom likely to be an option for a one-off impact on the environment, say through the construction of a dam or a new inotorway, yet there are difficulties at least as great in assessing the secondary and tertiary impacts on an ecosystem of such projects. Use of Risk Assessment - Once Off Events Many States now require a full environmental impact of major projects. These impact assessments can include the possibility of accidental damage, such as a tanker accident on a motorway. This would be handled very much in the way that I described in my first category of assessment. If risk management plays a part it would be in determining the accident response strategy, as for example with marine oil spills. The environmental impact assessment could also include the impact of continuous emissions, although this is usually assessed in the context of local pollution control systems. I will deal with these in the next category. However, the impact assessment can involve assessing the wide impact of a project through its disruption of the local natural environment. Possibly the most common problem is to have to assess the long term impact on an ecosystem of the loss of some proportion of the habitat. Using our traditional risk assessment terminology hazard analysis is straightforward, in the sense that the assessment is on the basis that the hazard - say the loss of 25 % of a nature reserve - is realised. It is the consequence analysis that is the major challenge. First there is seldom enough data. Most models of an ecosystem require not only large quantities of cross-sectional data, but also significant longitudinal time series data. Nevertheless since many of the theoretical ecosystem models of the 1970's were probabilistic in structure, it might have been thought that there was a good prospect of representing data gaps as sampling errors and so lead to a quantitative risk assessment. However while text book models derived for heuristic purposes have simple structural relationships, nature contrives to make real world cases much more complex. Real world efforts to model ecosystems thus have problems of structural uncertainty as well as scaling errors. A particularly interesting recent development in theoretical ecology which has yet to make its mark in risk assessment generally has been the introduction of models based on chaotic sequences rather than probability distributions. Some of the best known chaotic sequences are those that produce the pseudo random numbers in computer languages. Each next step is determined by previous steps, but over time the change in values take on random-like qualities. The ecological population growth equation can be cast in just such a chaotic form. Chaotic models of the natural environment are not that much less hungry for data than stochastic models, but they do offer the prospect of better qualitative descriptions of the total effect of a project on an ecosystem. Chaos related models could have more prospect of determining the answer to questions such as "will the species survive?" and thus becoming a 104 ...... 1 9 9 2 tool of a hazard analysis approach. In contrast probabilistic models tend to provide an over differentiated output of probabilities and population densities which may not be so useful as a basis for decision. Use of Risk Assessment - "Continuous Accidents" The bread and butter of environmental pollution policy is the control of polluting processes. There is no question in this case of a hazard analysis. Te hazard, if it exists, is being realised continuously. Once again the focus of attention is on the consequence analysis, and once again it proves to be uncertainty in the consequence analysis itself which introduces the possibility of unwanted outcomes. The classic textbook model of a pollutant is a substance with a very short lifetime in the environment and a threshold concentration below which no harm occurs, and above which harm gradually increases. With a short environmental lifetime, realised concentrations are immediately proportional to rates of emission. The aim of policy would be to get emissions low enough to keep the concentrations below the harm threshold. The aim may be too expensive to achieve at first and a trade-off between abatement cost and emissions has to be settled in the form of a regulatory requirement on the abatement technology to be employed. This is a problem which is now being addressed in an acute form in several Central and East European Countries. Even this simple world view offers substantial difficulties for consequence analysis. It would seem a reasonable pollution control policy to locate polluting plant as far as possible from environments with low thresholds of harm. However to be sure that no harm is taking place requires detailed analysis of the transport properties of pollutants. The cause celebre of transport models has been those modelling the transport of acid oxides. It is not only the transport of the pollutant that is important, but its transformation either to strong acid oxides or secondary pollutants while it is transit. Models to incorporate all the known science are very complex and difficult to validate in their entirety. Models of this form seldom contain stochastic analysis of their internal errors. In Europe confidence in the output is expressed by comparing the outcome of models of the same process by different modelling centres. The natural risk management strategy is to monitor the environment at the remote location and at the first sight of damage abate emissions. However recent patterns of environmental damage suggest that this approach may be inadequate as a risk management strategy. An environment may be initially resistant to acid rain damage because as the acidity of the environment is raised, alkaline compounds present in the soil begin to dissolve into surface waters restoring the neutrality. The alkaline compounds "buffer" the acidity and obscure its effect. When the rate of deposition exceeds the rate at which the buffer is replenished by natural processes, the buffer becomes depleted. When the buffer is exhausted the environmental chemistry changes suddenly. Even if the deposition was to stop altogether recovery would not be immediate, but determined by the rate at which the natural buffer was replenished. This can be very slow. One of the most dramatic examples of chemical buffer depletion has been the depletion of autumn Antarctic stratospheric ozone. Although the depletion was first observed in 1988, and although world emissions are dropping rapidly, it is not expected that the ozone hole will be repaired until at least 2000. If consequence analysis can be difficult to validate, and effects monitoring may be misled by buffering effects and unanticipated accumulation, a risk management strategy as to give some 105 ...... 1 9 9 2 bias towards prevention of damage ahead of direct evidence when the hazard shows some particular properties. These add dynamic elements to the assessment. First the hazard has to show the property of being irreversible, or quasi-irreversible. This includes the property of bio- accumulation, long environmental lifetimes, and interference with the large scale geo-chemical cycles which could produce buffering. This bias is sometimes called the precautionary principle.(7) This approach is to be found in much of modern environmental legislation. The legislation may require that attention be paid to not exceeding specified environmental quality standards. These are set with a degree of safety margin, will require environmental monitoring to ensure compliance, and correspond to conventional risk management strategy of monitoring and response. Within this outer framework of regulation, the emissions of certain substances are required to be reduced further to the level that best available technology could reasonably be thought to achieve. These emissions are in effect subject to a precautionary approach. Examples of such substances are the Red List substances in UK legislation for discharges to water. Decision Making in Environment Regulation After this review of environmental regulation and risk assessment it is worth taking stock. First while truly random processes do occur in environmental issues, they are not as dominant a part of the story as in safety policy. The main issue addressing risk assessment in environmental regulation is one of describing the degree of confidence that we have about one man's effect on another man's natural environment. In almost every case this uncertainty increases as we move from the physics to the chemistry and from the chemistry to the biology. Scientific enquiry has a key role to play by providing a proxy for the state of the environment, because the natural environment cannot speak for itself. However science reduces but does not eliminate uncertainty. Indeed it would be classic risk management philosophy to assess at each stage the balance of collecting more information but otherwise doing nothing and taking some action with the risk that it may not later prove to have been necessary. This management decision is no less social and political in context than any other regulatory process. Risk assessment is probably more exposed to criticism in environmental issues because there are few opportunities for addressing through 'some second stage any shortcomings or approximations in the assessment. Workers will continue their negotiation of their pay rate after a risk assessment of their task has been completed. The local populace, who are about to have the risk they bear for the benefit of society at large to be judged "acceptable", are seldom in that position. Unlike workers whose job is re-classified as high risk, who will probably make some economic gain the wage bargain, the neighbourhood may actually lose money. It may not then be surprising that the appetite for quantified risk assessments exhibited by some does not always appear amongst those bearing environmental risks. Environment law is about who owns or has the stewardship for what. Ignoring distributional effects - including the redistributions of risks and gains - was a major defect of 1970's technical evaluations of many environmental issues. Distribution effects are environment policy. In fact every new element of environment policy (biotech regs, transboundary air pollution, CFC's) is a renegotiation of societal property rights and liabilities. It is no wonder that environmental risk analysis when generalised from the risk to a specific individual to risk to society as a whole gets itself into such deep water. 106 ...... 1 9 9 2 The Need for Better Methodology To argue that some of the earlier rule-based "rational" approaches to environmental decision making proposed in the 1970's had serious shortcomings, is not to argue against continuing the search for a generally more systematic approach. There are a number of reasons why the need for a better systematic approach is more urgent now than ever before. - Most countries have a serious problem of "vintage" in their environmental legislation. The priority for a piece of legislation may have reflected a particular event in a particular media, and frozen handling in that area ever after. - The legislation may specify a "no harm" condition, before the stochastic hazard of carcinogens was identified. - legislation may have been ambiguous about whether maximum concentrations referred to the limit of measurement at that time or to the threshold of effect. - legislation may reflect different bargains on precaution at different times favouring either the polluter or the protected to different degrees. - some countries e.g. those in Central and Eastern Europe, cannot tackle all their environmental problems at once and need tools to help them set priorities and direct limited resources most effectively. When environmental regulations represented only a few legal instruments, these difficulties could perhaps be tolerated. Now that they represent such a major stock of legislation, the difficulties of fitting all these pieces together is becoming a major challenge. Major reviews are under way in a number of regulatory fora. The US EPA has begun a major review across its fields of operation.(8) The UK took its own first steps with the 1990 Environmental Protection Act which brought together waste, air pollution, integrated pollution control, dangerous substances and genetically modified organisms within one act. The Government's White Paper on the Environment in the same year laid out a plan for some 300 targets for the environment and was one of the most comprehensive approaches that has ever been attempted. Integrated pollution control in particular requires one regulatory agency to assess the impact of a scheduled process on all media at one pass and therefore establishes the basis for a more consistent approach. We hope to see a parallel integrated permitting Directive within the European Community soon. These revisions require appropriate decision making tools. As a first step, and as one of the 1990 White Paper commitments, the DoE has produced a guide to Policy Appraisal and the Environment, which you will find discussed elsewhere in this conference.(9) Some areas of environment regulation are technology based, and the decision making tools must be closely related to other risk based approaches to process control. Some examples from this area, like radwaste disposal, already make their contribution to risk management theory. How far 'can risk assessment help provide this missing synthesis of approach in other areas of regulatory process? 107 ...... 1 9 9 2 I believe there are a number of themes running through my review of risk assessment in environment policy tat shows that it could have a useful role to play in unifying our treatment of environment policy. At the same time there are areas which will still need much development. It is not clear despite courageous efforts that this methodology is fully portable across all environmental issues, yet there is a clear need for a better systematic approach. The challenge is to import the best of existing approaches, and to unify terminology, but to develop extensions where a conventional risk approach would be artificial. It would be a backward step to discredit an established approach like risk assessment, by trying to apply it to an area where it is not completely applicable. To escape from the abstract I will try to demonstrate my thesis by reference to an example drawn from a current cause celebre. At the same time I hope to suggest areas for the development of decision making tools. An Example - man-made climate change There is no point choosing anything simple. Rather I will try a 4-minute risk analysis of man- made climate change.(10) It neatly brings all the new issues to the fore. In risk assessment terms the physical hazard is well established. Certain gases - the so called greenhouse gases - transmit short-wave radiation but absorb long wave radiation. When the hazard is realised through an increase in the atmospheric concentrations of these gases, they will change the earth's energy balance. When the Intergovernmental Panel on Climate Change reviewed the science in 1990, it was, characteristic of all risk assessments, this hazard which they identified as their most certain conclusion. In one sense the hazard analysis is straight forward since these greenhouse gases are being released into the atmosphere all the time. However transferring emissions into realised greenhouse gas concentrations proves not to be that easy as increases in some gases is thought to change te removal mechanisms and ence concentrations of others. We saw in the case of incidental exposure that the next step in the assessment was to try and scale the azard against some benchmark. Our nearest scale is fluctuations in the earth's radiation budget caused by natural fluctuations. These are largely caused by fluctuations in aerosols, from volcanic eruptions and fluctuations in the solar output. We note at this point, as with other hazard analysis, that we do not need to complete a consequence analysis to compare absolute hazards. As the IPPC found the expected change in the radiation imbalance in the next decade due to greenhouse gases was likely to be comparable with natural fluctuations. This establishes the prime facie case for risk assessment. The next step we saw for risk management was to identify the dynamics of the hazard to establish whether monitoring and corrective action was sufficient. For some gases such as methane with short atmospheric lifetimes this might be possible. However two important gases, carbon dioxide and nitrous oxides are very stable with long atmospheric lifetimes. Thus even if emissions of these gases were stopped tomorrow their atmospheric concentration would drop only slowly. This quasi-irreversibility suggests that the risk management should be precautionary. This conclusion is reinforced when we begin the consequence analysis. The change in radiation balance changes the average temperature of the earth's surface. However thermal inertia of the 108 ...... 1 9 9 2 surface, principally the oceans, slows this warming so that it lags behind the radiation imbalance. Thus even if the emissions were to stop, not only would the greenhouse gas radiation imbalance drop only slowly, but the earth's temperature would continue to rise for say a decade or more. If this were not enough the hazard is not strictly the emission as much as the operation of the fossil fuel plant and the changes in land use practices. These take a long time to change. The IEA quote a figure of 25-50 years based on the historical analysis of Marchetti.(11) If, following the international climate change negotiations the aim is to avoid a dangerous interference with climate, then the risk assessment shows that we cannot simply monitor its arrival and then take action. We need to take action well ahead of its onset. "Well ahead" must be defined by the models we use to undertake the consequence analysis. However these models are not only enormous, taking over whole Cray XMP's but have limited validation in the real world. The basic problem is that the world has not had this imbalance in greenhouse gases forcing for over 100,000 years and we are therefore extrapolating beyond empirical experience. This raises two issues for the consequence analysis. First while the models are invalidated we cannot even eliminate with confidence the prospect that we are approaching a new climate transition. This is a characteristic of chaotic systems, and for all we know might occur rapidly. The second issue is that if our models are unreliable but the climate record is beginning to be contaminated by anthropogenic emissions then the historic records on which general climate risk analysis is based become of less and less use. A risk analysis does show up one surprising short coming of other analyses. Much is often made of the wide uncertainty in the outcome of different models of the consequence of climate. These differ over a factor of three. However risk analysis would naturally ask not only how wide the range of outcomes would be, but also whether they embrace a dividing line below which the consequences were of no concern. If even the lower range of climate change was in some sense "unacceptable" this would influence the risk management strategy significantly. It is not surprising that the international community has agreed that some form of action to contain emissions is called for now while the modelling effort is improved. It is not too surprising that they have found it difficult to reach full agreement on what premium to pay for this "insurance" and who should "pay". The first difficulty is to express what we know at present. The Intergovernmental Panel on Climate Change has resisted expressing its conclusions as probabilities on possible outcomes, despite the fact that many of its distinguished contributors come from meteorology where probabilistic descriptions are common place. There seems good reason for this. We are not trying to work out the average fate of the planet. Nor should we suppose that the average of the output of different models is a better judge of the future than the "best" model. IPPC has consistently used ranges to express the state of knowledge of key variables a method also advocated(12) elsewhere. However as each new fragment of knowledge is acquired it is difficult to judge the effect on the overall picture. Advocates for more or less action may find it tempting to play up the significance of each finding, and public understanding of the issue dissolves in confusion. There is a need to explore new ways of expressing our developing state of knowledge in non-probabilistic situations in a value free way. The more complex our environmental models become, the more pressing is this need. In national debates on the environment, the regulator will often take the role of Solomon and decide the final balance between winners and losers. This is not happy country as we have seen 109 ...... 9 9 2 for risk analysis. In negotiations between sovereign states the true nature of the bargaining process is explicit. In climate change case even the largest emitter accounts for only a quarter of the climate forcing that is changing its own climate. The best it can do is inspire others to action. The bargains that have to be made are much closer to the conflict resolutions exercises used in some countries as complemenatry to regulatory intervention. We find that some of the tricks of conventional analysis, particularly aggregation unhelpful. There is not a great deal of point calculating the change in "global welfare" consequent on climate change. Individual countries wish to know what happens to them, and how those who may believe they are winners intend to help those who are losers. What this debate reveals is that we have limited tools for presenting the substance of the issue, particularly when there is considerable uncertainty. It is clearly an area for development. Overall the general pattern of risk assessment and risk management fits well even in this unique and difficult problem. Traditional approaches do provide insights and rapidly evolve strategies. There are limits, which reflect how environmental bargains are struck, and how knowledge about environmental problems is to be presented. These are the challenges. Conclusions There has been a significant growth in environmental legislation over the last two decades. The EC alone has almost 300 environment related directives. This growth leads to strong pressures for better consistency and smoother machinery. There may not be one single "risk of mortality" in all this legislation but it would be reasonable to ask for an audit trail that leads from one legislative framework to another in a consistent way. The climate is ripe for better use of risk management techniques suitably adapted. This review has shown that the framework of risk assessment can be structured to cover a wole range of environmental problems. It is sometimes only the habit of generating a new technical dialect for each area that obscures the similarities. We can even see how applying this framework to the unique and pressing problem in environment policy adds some transparency and order to the reasoning. However, there are dangers. Risk management is a well established technique in some areas, and there is a real possibility that it may become tarnished if it is transferred uncritically I would offer three areas which are in need of attention: Expressing Uncertainty Uncertainty about the consequences of our actions, is the hall mark of an environmental risk assessment. This contrasts with traditional safety and reliability assessments where the issue is one of probable outcomes of random processes with well defined probability distributions. If probability is the natural way to describe stochastic processes, is it necessarily the best way to formulate uncertainty? Many of our concepts in risk analysis come from systems engineering. There we find a number of more general techniques, of which our familiar probability axioms turn out to be a special case. This has been an active area of research in Artificial intelligence work, where the aim is to codify expert knowledge and design computer programmes that can update that codification as new data arrives. Some areas of risk assessment have begun to explore the possible uses of such systems. One AI approach is to use so-called fuzzy logic.(13) This logic relaxes the 110 ...... 1 9 9 2 additive and multiplicative properties of probability theory. Another approach is to employ intervals rather than single probabilities to describe our degree of belief.(14,15) Perhaps these new artificial intelligence techniques could help us improve the neutrality with which we describe uncertainty. Risk and Conflict Resolution A lack of opportunity to address distributional effects outside the assessment was another hallmark of the environmental risk assessment. This raises the question as to whether it is necessary to take the computation as far as estimating societal aggregates, such as "social welfare" at all? Is not transparency in calculation and public debate a better way to use risk related data than to subsume tese in some grand weighted sum called "the social welfare function"? Would this not open the use of risk techniques in conflict resolution and a better national discussion on environmental issues? In which case bow should the assessment be modified? Chaos and Probability In some important areas of environment policy the systems of concern may show chaotic rather than random dynamics. What is the risk assessment and risk management methodology of such systems? Might they actually be easier to analyse because of the opportunity to use non-parametric techniques? In embarking on risk assessment we have to be clear as to its purpose. It may be to provide a rule based system so that a regulator, or the regulated, can be given a clear prescribed method to carry out required duties. In such a case it may be necessary to accept some compromises between accurately describing the real world and obtaining a workable system. On the other hand, in using risk based approaches to develop policy, we may need a more generalised approach that is not open to criticism of technical compromise. However there are common risk assessment themes across the wide range of environmental issues. There are some major areas that are in need of further methodological development. There is also an expensive appetite for data in environmental risk assessment that needs to be contained. Nevertheless there should be optimism that refinements in traditional risk assessment techniques will be a major contribution to improving the consensus in environmental issues, and the identification of priorities...... 1 9 9 2 The author would wish to express his gratitude for the helpful suggestions and comments by colleagues in the various Divisions of DoE Environmental Protection Command during the drafting of this paper. References 1. See for example the Health and Safety Executive 1990) Tolerability of Nuclear Power HMSO. 2. Department of the Environment 1986) Assessment of best practicable environmental options for management of low and intermediate level solid radioactive wastes HMSO London. 3. Royal Commission on Environmental Pollution 1983) 9th Report HMSO. 4. Department of the Environment 1989) Dioxins in the Environment Pollution Paper 27 HMSO. 5. Royal Commission on Environmental Pollution 1989) The Release of Genetically Engineered Organisms to the Environment HMSO. 6. Royal Commission on Environmental Pollution 1991) GENHAZ, HMSO. 7. Royal Commission on Environmental Pollution 1988) 13th Report pp. 10-12 HMSO. 8. USEPA 1990) Future Risk SAB EC 88 040. 9. Department of the Environment 1991) Policy Appraisal and the Environment DOE HMSO. 10. For a full account Intergovernmental Panel Climate Change 1990) Scientific Assessment CUP; Intergovernmental Panel on Climate Change 1992) Supplementary Report CUP. 11. IEA 1990) Energy and the Environment Paris. 12. op cit. 13. Zadeh L. A. 1965) Fuzzy Sets. Information and Control 338-53. 14. Shafer G. A. 1976 A mathematical theory of Evidence Princeton University Press. 15. Gordon J, Shortcliffe E. H., 1985) A method of managing evidential reasoning in a hierarchial hypothesis space Artificial Intelligence 26 323-57 112 9 XA04NO313 Occupational and Environmental Risk Assessment Problems in Central European Countries Needs and What has been Achieved - Poland as an Example Dr. Tadeusz Sulkowski Chief Labour Inspector, Poland Summary The externally imposed system of planified economy inhibited the development of methods of risk assessment in the countries of Central Europe. The trends in using the experience of other countries and the help of foreign experts in determining realistic hygienic norms at the workplace, limiting the hazards for the natural environment, increasing the safety of industrial installations and transporting dangerous substances will be discussed. Introduction For several decades in Poland, similarly as in other countries of Central Europe, an inflexible system of directed planning imposed from outside reigned supreme. In such a system there was no place for chance and risk. Such an approach proved to be particularly unrealistic in respect to matters of protecting life and health, and inhibited the search for new methods of evaluating hazards, in particularly the methods of risk assessment. So far, with the exception of a few examples, there is no practical application of risk assessment methods on a significant scale in the countries of Central Europe. 1. Risk for peoples life and health at the workplace In Poland almost one thousand fatal accidents at work occur per year. Each year 8-10 thousand new cases of occupational diseases are noted. Over one million people work under conditions, in which the concentration of agents harmful to health exceeds the established hygienic norms. The comparison of the number of accidents at work shows significant differences(l) Number of Index of fatal fatal Total number accidents per Country accidents of accidents 1,000 employees Czechoslovakia 607 209,067 0.076 Poland 842 108,300 0.109 Hungary 428 88,684 0.212 Sweden 96 92,242 0.016 Denmark 80 50,482 0.030 Great Britain 367 185,256 0.016 These differences would be even more unfavourable in respect to the countries of Central Europe when the total number of accidents at work is compared with the number of fatal accidents. 13 ...... 1 9 9 2 In Poland, similarly as in other countries of Central Europe, when the permitted exposure limits of harmful agents at the workplace are determined, the real possibility of ensuring certain conditions is not analyzed and only the criteria of absolute biological safety are taken into consideration. The definition of the permitted exposure limits are taken as such values of harmful agents whose action at this level on a worker for the whole life should have no adverse effects on his health and health of his progeny. As a result for about 240 chemical substances for which in Poland the permissible exposure limits (PEL) have been established, almost two-thirds of these values were often severalfold lower than those determined in the more industrial developed countries. For instance, the threshold xylene concentration in the air was fixed at 0 mg/M3 in the data for 1989-1990, per year in the Soviet Union, 200 Mg/M3 in Czechoslovakia, and in Great Britain, Sweden and the United States at 434 nig/m In Poland this value was 100 mg/ni At the same time it is worthwhile to point out that established standards were mainly based on experimental data and publications derived from Western countries. During the discussion on establishing the PEL for xylene in Poland about 120 publications were taken into consideration, whereas, e.g. in Great Britain 350 publications were used for this purpose including one by the Polish scientist Choroszewska. The values of the permissible exposure limits established in Poland are verified every few years, and their values are often set even more rigorously, and the hygienic norms are decreased even severalfold. This is sometimes completely cut off from the industrial reality. Thus for example decreasing the hygienic norm for carbon disulfide from 25 mg/m3 to 10 mg/M3 met the limit of technical possibilities in the technologies used for production of viscose silk in Poland. For the sake of comparison, this concentration was set at 30 mg/m3 in Great Britain, as well as in Germany and France. The setting up of severe hygienic norms greatly increases the number of people working in conditions which are hazardous - understood as working in the presence of harmful substances, whose concentrations exceed the actual standards. A nontrivial aspect of this problem is the compensation due to certain groups of workers employed under harmful conditions, which are legally guaranteed or obtained in group contracts. These include longer holidays, treatment in sanatoria, pay supplements, free meals and earlier retirement. This aspect of setting up the. severe hygienic norms would found general approval. There are approximately 800 harmful chemical substances used on industrial scale in Poland but hygienic norms have been established for only about a quarter of them. Thus a situation has been created in which for some chosen substances very severe norms are established, whereas for others, possibly equally harmful, there are no legal bases for demanding improved working conditions. Work on establishing the allowed concentrations of harmful agents is proceeding very slowly, in general 10 new proposals are prepared each year. In spite of the tremendous amount of work involved, these norms give a warranty of chemical safety only with a certain probability. It is impossible to completely eliminate the uncertainty and variation in the individual reactions of the human organism as well as the precision of the measurement of the concentration of the harmful substance at the workplace. The harmful substance can be a mixture of isomers in various proportions of which each is metabolized differently, and may contain its own impurities, depending for instance on whether it is a petrochemical or carbochemical product. At the 1 14 ...... 1 9 9 2 present rate determining hygienic norms for a substances would take about 50 years. Various possibilities of speeding up the work are being considered. One of these would be the adaptation of the values of allowed concentrations from the regulations in other countries. It seems that such a solution may even be necessary. Above all, however, a general change of the approach in evaluating the highest allowed concentrations of harmful agents is required - the values of hygienic norms must be more realistic, this is true both for those which have been elaborated and established as well as those newly elaborated. It is necessary to create a uniform system based on the idea of accepted risk, whose level should be determined in negotiations between organizations of employers and the trade unions. Such realistic values should fulfil the conditions set forth in directives prepared by the European Community (EEC). In this work the help of Western nations, especially of the EEC, is necessary. The help of experts on risk assessment should have a methodological character, but it also seems that an efficient way to help our countries would be to have foreign experts conducting some basic investigations and applying these methods, similarly as in the Kazincbarcik Works in Hungary (investigations of mercury contaminations). The practical results obtained and the trained group of local specialists convinced of the benefits of the technique of risk assessment might have begun the creation in our countries of consulting companies which would serve small enterprises, which are often helpless in the jungle of regulations and incapable of transforming them into everyday practice of labour protection. The help of Western experts should be coordinated by international organizations which offer such help but not always are aware of our immediate needs. 2. Risk for peoples life and health in te natural environment The general conditions of life, the state of the environment, the living conditions, the quality of food, the possibility of resting, the mode of transport, the working conditions form a linked system which determines the level of civilization. They are also sources of danger for health or life. The average life span in Poland is 66.8 years for men and 75.5 years for women, in Hungary these figures are 67.8 and 75.3 whereas in Great Britain are 71.9 and 77.8, in Sweden 74.0 and 80.0, respectively. The index of cancer diseases has increased in Poland from 172.4 per 100,000 in 1970 to 213.5 in 1987, and causes 20% of the deaths. Deaths as a result of illness of the blood circulatory system increase by 3 each year; there were 525 deaths per 100,000 in 1989. Multifacted interactions of many ecological factors on a human organism makes the tragical effect appearing not always where the real cause has acted. Sometimes the observed risk is te proverbial drop of water which makes the restricted human endurance break. The action of acute and subacute harmful chemical, physical, biological and psychophysical agents often occurring together in surprising combinations, leads to a decrease in somatic and psychic resistance, and the immunity to sudden hazards decreases while the susceptibility to chronic illnesses increases. Perhaps from this point of view the setting of more rigorous hygienic norms in Poland was justified, because an individual living in a poisoned environment, psychically tired and not always properly fed is more sensitive to extra hazards also at the workplace. In Poland 11.3% of the country is ecologically menaced, and these areas are inhabited by 35.5% of our population, in Hungary about 44% of the population live in ecologically endangered areas, and in Czechoslovakia over 60%. 115 ...... 9 9 2 Contamination of the atmosphere is mainly derived from energy producing facilities and metallurgic works found in the direct vicinity, but some of the contaminants come from very distant areas. Comparatively, the number of contaminants calculated as pure sulphur is as follows: Tons of sulphur/kM2/ year Country total own sources migration Czechoslovakia 6.15 3.23 2.92 Poland 4.55 2.63 1.92 Hungary 4.05 2.33 1.73 France 1.46 0.66 0.80 Sweden 0.68 0.10 0.58 The area of Central Europe also emits a considerable amount of pollutants outside its boundaries. Besides gaseous contaminations in the atmosphere, dust emission is also very harmful. Basing energy supplies on hard coal and lignite causes the formation of troublesome pollutants, both in the dumping grounds of the mines and the mounds of ashes which cause dustiness of the air. Among the energy producing plants operating in Poland only 10% have equipment for removing gaseous impurities, and 20% have equipment for effective removal of dust impurities. In Poland a list of the 80 factories most harmful to the environment has been published. It would be reasonable to obtain the opinion of Western experts as to the risk which their exploitation carries for the health of the people living in the immediate vicinity and a greater distance. This would create a basis for counteracting these hazards taking risk management methods into consideration. Our activity should be concentrated in places where they truly limit the risk of loss of health or of life and not where they can affect some hypothetical, unspecified hazards. Our current economic situation makes a careful choice of these actions necessary. Another serious matter in Poland is the seeping of many impurities into surface and underground waters. These impurities are absorbed by the human organisms from drinking water. Methods of risk assessment appropriately applied to the choice of places for storage of communal, industrial and other kinds of waste may help in the solution of these problems. Similar problems are caused by the destruction of the out of date pesticides and the disposal of empty containers. When they are buried in the earth by farmers they contaminate wells with drinking water. Similar contamination is caused by inappropriate disposal of manure from industrialized breeding farms. The risk assessment for drinking water is essential, as often contaminants can reach distant places and are harmful in concentration below taste threshold. The degree of contamination often is determined by epidemiologic observations, which is too late. These matters are directly linked to the problem of healthy food. An undesirable consequence of industrialization and of development of road communications is the pollution of soil and next, of food by heavy metals, including lead. The methods of risk assessment could also be helpful in exclusion of certain areas from food production. The use of chemicals (glue, varnish, binding agents, insulation) and asbestos in the construction industry may also pose a risk to the health of people who are constantly in their presence. 16 ...... 1 9 9 2 A particular problem is the use of asbestos, which is generally considered carcinogen. The allowed asbestos concentration (0.5 fiber/cm3 of air) as been normalized for working conditions. Meanwhile, many insulation elements which have been present for a long time and made with asbestos erode and emit asbestos dust into living quarters. The values of threshold concentrations of harmful substances have been established in Poland for 12 substances, but this list doesn't include asbestos dust. Proposals to stop processing of asbestos in our country are being considered, but the problem of the elements wich have already been used remains unsolved. Before appropriate decisions are taken the method of risk assessment with the help of experienced foreign experts should be applied. The application of the method of risk assessment in the field of protection of the natural environment might not be limited by country boundaries, as the contaminants of air and water do not respect these boundaries. The help of foreign experts will be useful if it encompasses several countries simultaneously. In our countries joint action has been taken to improve the situation. In Poland a programme of the reduction of air pollution at the so-called sulphur triangle covering junction of the boundaries of Poland, Germany and Czechoslovakia has been undertaken. A practical result of this collaboration is the abandonment of the construction of a Czech coke factory in Stonawa, on the Polish boundary, which would greatly contaminate that part of Silesia. This is a good start for a closer collaboration. In March 1992 a conference under the auspices of the European Economic Commission of the United Nations took place in Warsaw. It was devoted to the problems of cooperation and of balancing the development of the chemical industry in Central Europe. One of the subjects of the sessions was the idea of chemical safety. As a result of the conference it has been established that the countries of Central Europe should receive technical and financial aid in their effects to decrease chemical pollution. The will was expressed to comply fully with the Convention on Environmental Impact Assessment in Transboundary Context and on Transboundary Effects of Industrial Accidents. An agreement was reached on the need to create in Warsaw an international center for the problems of environmental protection resulting from the activity of the chemical industry. 3. Danger for life and health due to the sudden release of harmful substances So far in our country no events whose results could be described as major hazard accidents have taken place. To a certain degree this is due to the care in maintaining technical safety. As an example the analysis of the technical safety in the first nuclear energy plant in Poland can be discussed. The subject of the analysis was the application of PSA methods (probability risk assessment) for a simulated failure of the nuclear reactor initiated by a small leak in the primary circuit. The investigated model was the Soviet reactor type Vodo-Vodianyj, Energeticzeskij Reactor VVER-440. The investigation was carried out by scientists from Poland, Czechoslovakia and Hungary and sponsored by the International Atomic Energy Agency, IAEA ensured expert help and access to computer codes. The risk related to the failure of particular safety subsystems and the probability of particular accidents scenarios were estimated. In the first stage some simplifications were made, which excluded mistakes made by te staff, and te possibility of having to rectify incorrect action of the personnel was not taken into consideration. 117 ...... 1 9 9 2 Such an uncorrected error was the source of the Tschernobyl disaster. The failure of reactor nr 4 type RBMK in the nuclear energy plant in Tschernobyl on April 27, 1986 led to the greatest release of radioactive substances in the history of mankind. The main hazards were isotopes of iodine 131 transported in masses of contaminated air, causing both external irradiation and also internal irradiation due to respiration and consumption of contaminated milk. In North-Eastern regions of our country the concentration of J31 in milk reached 3 kBq/liter. In this situation the action was undertaken in Poland, for the first time in the world on a massive scale, to block the thyroid with natural iodine, prescribing iodine preparations to infants and children under 18. The fact that overdosage of natural iodine could cause malfunctioning of the organism was well known. It is estimated that as a result of the Tschernobyl disaster 6 thousand people will die in Poland due to cancer, including about 2 hundred suffering from thyroid cancer. The same estimates foresee the occurrence of heritable mutations in two subsequent generations in about 2 hundred people, including twenty mutations causing mental retardation. As a result of the Tschernobyl shock the Polish government decided to abandon the already started construction of a nuclear energy plant in 2arnowiec on September 4 1990. Both the public opinion in Central Europe and international opinion began to regard this region more critically. It turned out that over a dozen atomic energy plants with low technical safety are present, as well as several industrial clusters which are harmful and destructive to the environment and emit considerable amounts of harmful gases and dust to more distant countries, in effect endangering the natural environment in all Europe. Public opinion is less sensitive to individual accidents separated in time and space. Nevertheless, these accidents occur frequently and are the main source of hazards. In Poland they include road accidents related to car circulation. Unfortunately Poland has the highest number of accidents - the index of deaths per 10,000 cars is 60 per year, compared to 37 in Denmark and 25 in Great Britain. This is due to the poor state of the roads, lack of motorways, and to the technical state of the cars. In the last two years about 350 thousand second hand cars with high achievements, but with a technical state not always easy to ascertain have been imported into Poland. Detailed investigations of this great number of accidents have made it possible to identify over 500 particularly dangerous road sections where 4 or more accidents take place every year. Applying the methods of risk assessment, this knowledge may be used to plan the routes for dangerous cargo. The carrying of dangerous loads is a real problem in spite of the acceptance by Poland of the resolutions of the ADR. Examples are given by movements of the Soviet army going back to Russia. On February 29, 1992 an accident occurred to the vehicle belonging to Sovtransavto who declared to carry piece-goods. The rescue crew arriving on the scene detected a leak of a component used for explosive production which could cause a powerful explosion. Therefore in determining the transport routes the distribution throughout Poland of chemical safety stations, their preparation for combatting various types of hazards (chemical, explosive, toxic) and their state of technical readiness should be taken into consideration. As a result of this work a list of the safest transport routes for carrying specified materials might be made, comprising the directions of transport, seasons, dates and even the names of trustworthy carriers. 118 ...... 9 9 2 Concluding remarks The countries of Central Europe are currently undergoing fundamental changes on the road to a market economy. There is a rapprochement to Western Europe, initiated by the promising act of Association with the European Community. The amount of changes which already have taken place makes it possible to state the significant progress in this field in Czechoslovakia, Hungary and Poland. The model of the protection of human life and health inherited in our countries lags behind our current needs, the international conventions and the requirements of the EEC. Attempts are made to modify this model by broadening the list of hygienic standards in force, the development of the system of measuring the concentrations of dangerous substances at the workplace and increasing the range of medical epidemiological investigations. In Poland a system of differentiated compulsory insurance payments was elaborated. Payment depends on the accidents rate and occupational diseases rate occurring in an enterprise, and on the part of workers exposed to dangerous substances at concentrations exceeding the threshold values. These proposals are still burdened by the tendency of penalization and prefer administrative solutions. The methods of risk assessment are an interesting alternative, which if used with the help of Western experts, and with joint participation of Central European countries in a more or less institutionalized structure, will speed up adaptive actions. Enrichment of the existing system by the additional dimension of risk assessment will make the problems appear more plastic and will make the choice of goals more easy. It will enable individual enterprises to autonomously create their own concepts of labour safety and to include insurance companies in the process of improving living and working conditions to greater avail. ...... I...... References 1. W. H. Hallenbeck, K. H. Cunningham - Quantitative Risk Assessment for environmental and occupational health - Levis Publ. 1988. 2. Quantified risk assessment: its input to decision making - HSE report of working group. London, 1991. 3. Seiji Machida - The ILO program on chemical safety - The UNECE Conference, Warsaw, March 1992. 4. J. Indulski - Chemical safety - Ochrona pracy nr 12/1990. 5. H. Kirschner - Relation of work protection to ecology - Ochrona pracy nr 12/1990. 6. K. Szymczykiewicz - Hazard and professional risk at the workplace - Medycyna pracy nr 21985. 7. T. Dawydzik - Hazards in the work environment - the Assessment of current and foreseen difficulties in their determination, Udi, Institute of Occupational Medicine, 1992. 8. Launce Grainger - The role of UNECE in the safe transport of chemicals in the region - The UNECE Conference, Warsaw, March 1992. 9. Report of Environmental Consultant Team regarding the mercury contamination problem from BVK facility, Kazinebarika, Hungary, May 1991. 10. H. Paul, A. Illing - Possible risk considerations for toxic risk assessment - Human and Experimental Toxicology no 10/1991. 11. Toxicity review - Xylenes - HSA 1992, issue no 26. 12. Documentation for the proposed permitted values of occupational risk; Xylene. Bulletin of the Inter- departmental Commission for the actualization of the list of the permitted exposure limits for agents harmful to health in the work environment. Issue No 7 p. 109, Warszawa, 1991. 13. H. Borysiewicz - Development of risk criteria for the Nuclear Fuel Cycle - VVER Risk Analysis and Assessment IAEA TECDOC-505 Vienna 1989. 1 9 9 XA04NO314 Risk Assessment: A Regional Approach M. PaWek Occupational Safety Research Institute, Prague, Czechoslovakia A danger to human health, to the environment and to material values is present at any place on our planet. However, the extent of the danger and the severity of consequences widely differ. In Europe, we can hardly find a region presenting greater risk to a living organisms including man, that the region of North Bohemia. The black triangle of North Bohemia suffers not only from the consequences of the forty-year socialist economy, but also from heavy emissions from German and Polish factories and power stations. Air pollution in this region reaches average concentrations of S2 and airborne dust of approximately 150 /Ag/m3 per year. Czechoslovak emissions participate in the deposition of sulphur by 53% in a long-term average. The highest proportion coming from abroad is 14% from the lands of the former German Deniocractic Republic and 11 % from Poland. This region is also a source of emissions for neighbour states. These are largely sulphur oxides and nitrogen oxides. Total emissions of S2 in this region are approximately I million tons per year, and as for solid emissions, the number is 280 thousand tons per year. Emissions of CO are not negligible either. The high energy consumption of our industries and the use of fossil fuels for heating, which is still widespread, as well as the density of transport in this region result in average emissions of C2 exceeding 12.7 tons per km 2 The situation is not better in the water economy. The quality of surface waters is affected not only by the emissions, but also by the inefficient agricultural industry applying excessive quantities of industrial fertilizers and pesticides. Monitoring of residual pesticides is at the same time very scarce due to the sophistication of analytical methods and technical equipment required. The low quality of surface water results also in the deterioration of ground water. In addition to overall contamination caused by agricultural activities, the worsened quality of ground water is affected also by leakages from sites, storages and pipings as well as by infiltration of polluted water from irrigations, water flows and, last but not least, by industrial accidents. In the area of the river Labe, for instance, the concentration of nitrates in ground water is 700-800 nig/ - , the permissible content of NO, in standard drinking water being 50 mg/1 Also considerably high are contaminations caused by chlorinated compounds. The situation is adversely influenced by the fact that in a number of localities there are inadequate, if any at all, sewage disposal plants. And where they exist, there are problems with the treatment of sludges. The sludges containing heavy metals cannot be used in agriculture and their processing is not possible due to the lack of the appropriate processing plants. In a number of cases, the sludges are let out back to water flows. The impact of industrial waste sites and sewage waste sites is not negligible either. In most cases, the waste sites are not adequately and safely protected against a release of contaminants to the environment. In most cases, the data about properties and quantities of deposited wastes are not available. 120 ...... 1 9 9 2 In the North Bohemian region, there are 112 waste sites of the total area of 1423 ha, containing 150 million tons of deposited waste. The capacity of the waste sites is 70% filled at present. The North Bohemian region is also a very dangerous area with respect to potential hazards of industrial accidents or emergencies with possible consequences to health and environment. The developed industry, the density of both railway and road transport, river transport, storage and transport of manufactured or processed materials - all this creates serious hazards to the environment. For instance, leakages of oil products present 50% of all accidents causing water pollution. Leakages of chemicals represent 15% of this number and in 12% of the cases, these are accidental leakages of waste water. The accidents endangering ground and surface water are mostly caused by human factor failures - due to negligence, carelessness, wrong handling or ignorance of workers. Technical causes often lie in obsolete and non-satisfactory conditions of plants and equipment. Extraordinary atmospheric effects in this region, the effects of many sources of air pollution as well as the emissions from neighbour states result in frequent temperature inversions associated with a slow dispersion of pollutants. At the time of the inversion, daily concentrations of sulphur dioxide reach over 500 Aglm However the industrial accidents connected with the release of harmful substances into the air are not as frequent as the leakages to soil and water, e.g. in 1990 two accidental releases of generator gas occurred in the plants of the firm Glass Union and an explosion took place in the industrial waste site of the Company for Chemical and Metallurgical Manufacturing. The adverse conditions and development of environmental factors together with bad economic, industrial and agricultural policies as well as bad planning and management are also reflected in the state of the soil in the region. Vast areas of the soil are endangered by water and wind erosion. Chemical degradation of the soil appears as a result of an increased content of acid substances present in rains, fog and ice accretions, as well as of a direct absorption of S2, NO,, and NH4. The content of heavy metals and toxic organic compounds is increasing. A part of these contaminants comes from the air, a part from preparations for plants protection and from phosphate fertilizers of a poor quality. It can be summarized that the North Bohemian region has an undisputed primacy in the extent and intensity of the environmental impairment. The percentage of territories with satisfactory conditions does not reach even 20% of the region's area. However, only 4% of the population of the region are living in this territory. 96% of the population are living in the environment adverse to human health and half of them even in the extremely impaired environment. This enormous environmental impairment has resulted in the greatest damages to forests and soils in the European scale, as well as in great losses of agricultural production. The most threatening is, however, the fact that the population of the region lives in the extremely impaired environment with all the implications, known only in part, to their health and their life-span. 121 ...... 1 9 9 2 The mean life expectancy in the North Bohemian region, compared to the average of the whole population of the Czech Republic, is by more than 1.5 years shorter. Compared to the Czech region with the relatively highest level of environmental factors, it is shorter even by more than two years. For a comparison the average mean life expectancy in the North Bohemian region is 69.5 years, in the whole Czechoslovakia 71.0 years, in Denmark or Germany 74.5 years while in Sweden and Switzerland it is 76.5 years. In Europe, the Czech Republic ranks among the countries with the highest mortality rate of both men and women. The adverse influence of the worsened environment is manifested mostly in the health of children. In the region considered, the prevalence of allergic diseases is considerably higher - more than 40 cases per 10,000 children in comparison with 15-20 cases in other regions of the Czech Republic. The high number of endangered pregnancies and the high infant mortality rate can be also attributed to the worsened environmental factors. Results of investigations also show a significant correlation between the extent of the environmental impairment and the incidence of some negative social phenomena, such as a high suicide incidence rate, alcoholism, drug addiction, divorce rate and abortion rate. The North Bohemian region as, above all, the highest drug addiction incidence rate 16 and more identified cases per 10,000 inhabitants older tan 15 years). The suicide incidence rate sows te highest frequency in this region in men, it reaches 50 times higher than the European standard per 10,000 inhabitants. In women, the suicide incidence rate is lower, approximately half of the number of men. The extreme number of divorces in this region together with other incidences are evidence of the instability of the social environment. The relatively high abortion index reaching in some localities of the region even 111 abortions for every 100 live-born child is further evidence of the unsatisfactory state of the environment. The long-term and deep impairment of all components of the environment produces parallels in the sphere of social processes. The negative development of these processes is consequently endangering the function and future of the whole region. The devastation of the environment as a consequence of the inadequate and inconsiderate management at all levels cannot be simply stopped by changing the economic structure, increasing environmental investments and introducing more perfect technologies. The principal change must be that at the social level. It is evident that the problem of the region is a complex problem requiring an overall solution. In our opinion, the solution should be based on an integrated risk assessment stemming from the idea that all the health and environmental risks in the region should be identified, analysed and evaluated so that a reasonable decision about the risk reduction could be made. At the same time, the social and economic consequences of such risks should be considered, as well as the benefits of their reduction, cost of the reduction or possible elimination of risks. We think it necessary to formulate principle approaches to integrated health and environmental risk control. 122 ...... 1 9 9 2 In Czechoslovakia, the comprehensive integrated approach to risks has not been applied in any region or locality so far. Therefore we should greatly appreciate possible participation of national and international organizations in the project, their technical aid and professional advice. The objective of the project, a part of which dealing with the prevention of major accidents is coordinated by the Occupational Safety Research Institute in Prague, is to develop a system of prevention and disposal of accidents corresponding both to the requirements of the European Community and the needs of the regions and industries of Czechoslovakia. Considering the economic demands of the project and the shortage of experts in this field, we were forced to work out such a procedure that would enable us at least temporarily, to get over these shortcomings. The procedure chosen can be summarized in the following steps: 1. One model locality in the North Bohemian region was defined for the development and verification of basic methods and approaches and the second, comparative locality in the Central Bohemian region. 2. R-Lsk identification and risk analyses in both localities are being carried out and the impact of risk within and outside the localities is being considered. The risk identification is based on a whole range of supporting data. These are, above all, data on the quality of the environment in the region concerned as well as geographical data. The data collected include also data about basic activities carried out in the region considered (industrial and agricultural production, storage, transport etc.), as well as related information on technical plants and equipment in which the activities take place, about hazardous materials, the transport of such materials etc. The initial risk identification will include two areas: - hazards presented by accidents and other abnormal events, - hazards presented by normal operation. 3. At the same time as the risk identification, the analysis of existing safety measures, both technical and organizational, is being carried out. 4. Based on the information obtained, appropriate measures will be proposed, implemented and verified in the model localities, with the view that these measures could be adopted not only in the North Bohemian region, but also in other regions in Czechoslovakia. The measures proposed will be aimed at the fields of - legislation - inspection - education 123 ...... 1 9 9 2 In the process of creating the legislation, it will be necessary - to work out a method and to establish a uniform way of identifying and recording risks connected with the construction, operation and disposal of technical installations, - to propose suitable and effective means and measures aimed at the reduction of operational accidents hazards, and - to develop, verify and introduce adequate analytical methods enabling us to identify causes of failures both of the technology and human factors, - to propose effective preventive measures. The correctness and completeness of risk identification and recording as well as the suitability and adequacy of the measures adopted must be a subject of independent inspection. For the inspection purposes, it is necessary - to propose and work out a system of inspection and suitable inspection methods and procedures, - to ensure a sufficiently high professional level and qualifications of the persons carrying out inspections. It will be necessary that the measures include a system of education and training of workers in organizations with major accidents hazards, education of inspectors and persons responsible for auditing and, last but not least, education of the public, particularly in the localities that can be affected by the consequences of these events. The educational system has to ensure, among other things a continual training of professionals and a training aimed at emergency control. With regard to the extent and demands of the project and the shortage of experts in this field, the participation of foreign experts would be very helpful particularly in the training of our professionals in methods of risk identification and quantification, prevention of failures etc., complemented by visits of selected professionals to foreign firms and organizations so that it will be possible to fix the knowledge learned in the training and to gain practical experience. Another field of cooperation, which we think to be of great importance, is the research and development of required methods, particularly - methods and systems for identifying and monitoring risks, - methods of prevention of the human factor failures, as well as technical equipment failures, - methods and ways of increasing the preparedness and informedness of the public in emergency situations, - methods of causal analyses of accidents and their causes, - methods of quantifying major accidents hazards and potential consequences of accidents, - systems of rating, classification and recording risks, accidents and their causes, - means of reduction of the consequences of major accidents. We should also highly appreciate help in the training of the staff of undertakings and in the education of the public to desirable and adequate behaviour in emergency situations. 124 ...... 1 9 9 2 These issues have been neglected for years. It will be therefore necessary not only to work out own specific procedures, but also to search for analogies and to learn from foreign experience and approaches. One of the ways could be the training of Czechoslovak experts aimed at improving the informedness and preparedness of the public, improving the cooperation of industrial enterprises and the public, representative bodies and local authorities, as well as at creating suitable attitudes and behaviour of the public. As can be seen from the characterization of the problem, an improvement of environmental conditions in the black triangle of North Bohemia is a very demanding challenge. Its successful solution depends on a thorough and comprehensive analysis and assessment of health and environmental risks. Considering the nature and severity of the problem, we think that there is not only considerable space for the cooperation here, but the international cooperation is necessary for its solution. 125 ...... 1 9 9 2 z < 94 ' I L A% CID ki-L J-1 Yu 70 -0 CO 0)(1) coE M -0 cm co E 0 CZ 0 x CZ 7C3 > C) cc: w 126 ...... 9 9 2 tell I CD CZ CZ _0 >1 rr CZ LLI CL (Z _0 E L>, 70M >0 0x 127 ...... 9 2 ;g. Z cr I J Iq '40. -LLJ CZ CZ CZ (z W 0 o o C) -t CD co o o C) C\j -I- co MNOEM 128 ...... 1 9 9 2 Years F gure 4 Mean life expectancy 78 76.5 76 74.5 74 72 71 70 ------69.5 101111,gn 68 WIMMMIZZ, North CSFR Germany Sweden Bohemia Denmark Switzerland 129 ...... 9 9 2 Figure 5 1. Choice of model locality 2. Risk assessment 3. Analysis of existing safety measures 4. Design, implementation and verification appropriate measures - legislation - inspection - education Figure 6 Fields of cooperation 1. Training professionals in risk assessment 2. Development and implementation of required methods 3. Training and education of the public 130 ...... 1 9 9 2 Panel Discussion lZ30-18.00 ProfessorLeszek Pacholski (Chairman, Session 2) The starting point of our discussion of the all-day proceedings are two introductory reports prepared by our official rapporteurs Mr Jim Hammer, President of the International Association of Labour Inspection, and Professor Jeyaratnam, Head of the Occupational Medicine Division at the University of Singapore. After the report of the Conference participants we have the possibility to give questions, comments or remarks concerned with two general problems: What is Risk Assessment and What do we Want from it? And the second: The Role of Risk Assessment in International Policy. Let's start with the first short report prepared by Mr Jim Hammer. Jim Hammer (Rapporteur, Session ) Thank you very much Mr Chairman. To summarise a morning of considered papers of this quality is no small task and I am conscious that I will not do justice to all the contributors. We covered a vast range of issues and we've been asked to consider industrial health and safety, nuclear safety, environmental safety, public health, carcinogens, food hygiene, biological hazards, genetic hazards, genetic manipulation and transport; and we were reminded that it was not only a matter of the hardware but that also, increasingly, suppliers were selling on the basis of performance and the disenchanted consumer therefore had an additional cause of concern if he or she did not get what they expected. We ranged from individual sites to whole swathes of land and whole industries. We ranged from children's toys to national policies in cancer prevention and we were reminded that crucial to this was the distributional effect, the risk to whom, borne by whom, of benefit to whom. Having considered the definitions, we were reminded of the differences between the perceptions of professionals on the one hand and the public or consumers on the other; in the case of the public perhaps motivated by fear, but looking very much at events and effects. On the other hand the professional looking at the probability (also of course looking at the probability of the effects) but primarily starting with the probability of the initial event. On the other band we were also reminded of the difference between acceptability and tolerability. We were reminded that acceptability is really a definition of below the level of concern whereas tolerability is one which involves political and social judgement. These are by no means the same thing. And we were reminded that what is tolerable depends too on the extent of personal benefit and the extent to which it is a voluntary assumption of risk or an imposed risk over which the individual has no control. John Rimington suggested that risk assessment was about -giving structure and order and rationality to coping with a harm or detriment whether or not correctly perceived and above all he focused our attention on the difference between individual, or in Dr Silbergeld's terms unit risk, and societal or population risk and both will need to be addressed at this Conference but particularly the latter. Risk assessment was then perhaps about empowering society to make a choice: a choice on investment, a choice on, as John put it, self-denial or abstention and a choice on how to put things right and in what order after the event. A number of people spoke about principles. Should we go back and reassess the philosophy? Risk assessment had to be coherent, consistent and credible. Risk assessment was about prediction not hindsight but with the 131 ...... 1 9 9 2 cautionary words that uncertainties were icreasingly involved and certainty was not certain. John Rimington suggested we had to inject objectivity and both Ellen Silbergeld and Dame Rachel Waterhouse looked for greater numerical precision, whereas we were also reminded by Mr Aro of the importance of the consultation with the work people, the people who had double commitment not only from being within the workplace but living around it; and we were reminded that we had to quantify not only the risks but we had to quantify them in terms of human concern, in terms of economic considerations and in terms of political accountability. Ultimately however it was the public to whom we were accountable and they wanted the risk kept to the minimum while at the same time benefiting from what industry and processes provided. It was observed that consumers tended to see risk as a relatively minor part of their decision- making process but at least they did expect information. Information about issues thought to be commercially confidential was not always as crucial as was sometimes suggested and sometimes the problems of the public with information were the methodology and the poor presentation rather than its absence. There was perhaps a need therefore for conventions for comparing different risks, their probability and their social burden and several speakers spoke about the importance of using risk assessment to establish priorities to ensure the coherence of precautions (the strong lock on the weak enclosure!) and to rationalise that use of resources. Ultimately however, as was said, the judgement on risk tolerability is a political choice and in dealing with that we had to remember not simply the hardware but the human factor, a crucial component in our future discussions. Control was spoken about in the industrial context: the need to manage risk in the way set out by Mr Williams, the positive management of risk, including training and the monitoring of results, and including planning for mitigation. But in the end we were presented with a number of crucial issues on which to come to a conclusion. To get a clearer view of societal risk, acknowledging that in fact there was not precise certainty: as Dr Giarini said, "the vulnerability of technology". To agree about quantifying and weighting risk to avoid the shock reaction, to enable people to understand te cost benefit approach and to put it in numerical form, although as Dame Rachel graciously admitted it was no small task. And thirdly, to agree principles about communication of risk. The public are not as stupid as they are sometimes presumed to be, they can understand racing odds and they want to understand both the standards and the criteria. And the other point is that publicity, as Ellen Silbergeld says, itself creates a motivation through making the facts available. Fourthly it was suggested that we had to accept uncertainty and that in spite of these problems we had to lead, to take action. So if we say that it is ultimately for society, for government to decide, it is for those here this week to provide the guidance on how government, how employers, how trade unions and how the public can respond in an informed fashion. Thank you. ProfessorJerry Jeyaratnam (Rapporteur, Session 2) Thank you. I cannot match Jim in voice nor eloquence but what I'll try and do is try and summarise the afternoon's three hours in the next three minutes. To kick off we had Dr Pinnagoda from the International Labour Office which is also called the International Labour Organisation, if you can unravel the two differences, and he suggested that risk assessment is a component of ILO's chemical safety programme and he, set out the structure of that programme. He also suggested that, for the future, risk assessment be considered as a component part of 132 ...... 1 9 9 2 ILO's occupational safety and health programme which has several common features with risk assessment itself, and that people use these ILO activities to enhance risk assessment in the future. This presentation was followed by Dr Van der Heijden from the WHO who suggested that a large amount of the work undertaken by the WHO, particularly the global programme as well as the European programme, impinges on risk assessment itself and in fact in 1989 a European charter on environment and health resulted in the setting up of a European Centre for Environmental Health. And Dr Van der Heijden showed us for the first time several maps showing the pattern of population densities and urban rural settings in the European Continent. Subsequent to Dr Van der Heijden's presentation Jim Brydon, on behalf of Bill Long, presented his paper for the OECD Group and he suggested that economic development cannot be divorced from environmental and health issues as such. He suggested again that increasing importance of risk assessment is a major consideration of the OECD rganisation. He also developed a number of questions and comments relating to this issue. Ron Haigh from the Commission of the European Community told us that the world is a risky place and he also emphasised the issue that risk perception varies considerably among te perceivers. He also spotlighted the issue of the uncertainty surrounding the concept of risk assessment and this uncertainty must be recognised in any decision-making process, particularly the political implications. Finally, he had the government balancing risk assessment on the one hand and risk management on the other hand as an important consideration in the final analysis. David Fisk underscored the need for a systematic approach to environmental policies and took on the task of defining environmental terminology and identifying some of the issues. He examined the pros and cons of extending risk assessment techniques in government policies and, for the future development, identified three areas which we should keep in mind, notwithstanding the fact that risk assessment is indeed a very useful too]. The features which he thought you should keep in mind in making the final decision-making process should be: the uncertainty associated with risk assessment; the risk and conflict resolution; and the chaos and probability associated with it. These were all factors which we should keep in mind before we think of a perfect solution. For the post-tea session the speakers were from Central Europe. Dr Sulkowski presented from Poland and the point he would like to make is that there are rapid political changes going on in the Central European countries, particularly in Poland. He took on the risk assessment as relating to the fields of occupational health, natural environment and sudden release of harmful substances and suggested that there were enormous problems resulting from the past political practices and ideologies. There was extensive need for collaboration between these countries and the industrialised market economy orientated Western European nations and he was looking forward to this collaborative effort. Similarly, Dr Palecek from Czechoslovakia presented the situation from Northern Bohemia and suggested that Northern Bohemia suffers from trans-boundary as well as internal problems of pollution and once again suggested that there should be immediate and intensive collaborative activities between the countries of Central Europe and the Western market orientated European countries. Thank you Mr Chairman. 133 ...... 1 9 9 2 Question and Answer Session DrJim Pikaar(Shell International Petroleum, The Hague) Q: My question is to Mr Rimington, and I'm sure to others. At a conference like this one sees uncertainties becoming a little clearer and that is one of -the objectives of attending them. What has interested me and I think challenges most of us is what Mr Rimington has come up with and he's asked us to look at societal risk and develop the concept. I come from the group that has dealt with major hazards and for many of us societal risk has been something that we thought we knew the definition of, it is something like this F/N curve. Though hard to grasp and fathom it is at least defined but I think by posing this question, what is societal risk, Mr Rimington does indeed force us to think about something which is very important because what society considers a risk is obviously very much more than that which is encapsulated in this F/N curve and if there is some way of getting a broader definition of societal risk, which clearly is necessary in such a broad conference as this, then my question is this. Is it really the intention of Mr Rimington that we should abandon the F/N curve and find something more meaningful for societal risk? Dr Sam Harbison (Director, Nuclear Safety Division, HSE) A: There is no suggestion in any way, certainly from HSE's point of view, that we would abandon the F/N curves that represent the quantified outcome of harms to society. The extension that we're looking for is a method or agreement on how, first of all, one might integrate together in some method or join together the different F/N curves that arise from a particular expression of a particular hazard. The one which it's perhaps most developed for is nuclear power where the risk to society would consist of a number of F/N curves such as the likelihood of fatal cancers; the likelihood of early effects; the amount of land which might be contaminated by releases; the number of people who might be evacuated; all of these are harms or costs to society but they've got different values to society. How does one weight these in some sense? But even more difficult is: how does one take account of the largely unquantifiable damage to society in terms of political damage and loss of confidence in the industry? Those are two particular ones. The question really for us is to decide if there is a structured way by which we can first of all go through the quantified parts and compare one with the other and secondly, how do we get into the more difficult part of taking account of these largely unquantifiable factors which, if one is cruel about it, were far more important post-Three Mile Island in America than any of the quantified effects? So that's what I think John Rimington was meaning. Lt Cdr Jack Rose (London Underground Limited) We are doing extensive risk assessment within London Underground and using it to quantify our risk. We are also using it to justify safety expenditure which, in our case, represents a fair proportion of government expenditure. One problem we have is trying to quantify the risk along with the various other benefits; for which purpose we've had to look at items like value of life to try and cost them against other benefits - operational benefits, economical benefits and so on. The other aspect is using that as well to put our risks in order to see which ones we must do first; but we've hit a small problem particularly with the Treasury Department of our country. They are very worried about the whole principle of value of life because it implies that there is no lower 134 0- ...... 1 9 9 2 limit to safety. As long as one can show a cost benefit we must do it and that technique has no bottom limit. There must surely be a point at which we say enough is enough and the HSE have already admitted there is a point, what they call acceptable risk, tolerable risk, the point which we either accept it or tolerate it, we don't do any more. What is the HSE's view on this in terms of what guidance can they give us (and we know they give quite extensive guidance for instance in the nuclear industry)? But for any of the rest of us - and the Treasury I know will press on this as well - where do we stop, where do we look to say this is the limit? Perhaps also to Dame Rachel Waterhouse, would we look to her sort of rganisation to get a public view on where do we stop, what is the limit? Where do we go in this particular area? Dr Sam Harbison Well perhaps I should venture some answer to Jack Rose before this goes into the next question. First of all I would take slight issue with the way he phrased it, probably he was just in a hurry, but there is definitely in HSE's terms quite a considerable difference between a tolerable risk and an acceptable risk; and tolerable risks are at the level where they are just tolerated and we very much expect, under HSE's regulatory approach, that something will be done at the tolerable end. However there comes a point, as he has quite rightly said, where we reckon that the risks are made so low that they are acceptable in comparison with other risks that people quite apparently accept. But there is the difficulty, and that's one of the purposes of this Conference I think, that we can make, as regulators, some educated guesses as to what are the risks that the "public" is prepared to accept for different industries. But it's not at all clear that we're always right and we are quite possibly not the right judges of this; and people, such as the Organisation which Dame Rachel Waterhouse represents, probably have a far better handle on judging what is the level of acceptability than we have. So we're looking for assistance in telling us what is the level of risk which is generally acceptable in different phases of life. Bob Clare (Smith Kline Beecham, UK) Q: Following on from those two comments perhaps I can say two things. The first is that we have already entered, and should take the opportunity to benefit from, an enormous opportunity for public debate in this and a number of other crucial moral and social issues; and may I make a plea for transparency throughout the discussion because one of the problems that we face right across the profession of risk management is our inability, our unwillingness to be transparent. Biotechnology is a prime example of where transparency would have solved a number of problems of misconception that the public now have. So I would return to the reasonable man on the Clapham omnibus and I apologise to our overseas visitors - he is the basis of common law in the United Kingdom. A test of what a reasonable man might consider to be appropriate in the circumstances is usually - providing there are enough reasonable men who have been asked - a good way of judging what is socially acceptable. The problem we have with risk is that we've never actually done that exercise. Lord Robens tried to launch it, I think it must have been at least 10 years ago on a Dinibleby Lecture on the television I remember, but it never got anywhere. So my plea is that we actually ask the public and put the issues before the public; and whilst this Conference is wonderful, until we've done that latter exercise we're really going to be struggling or the legislators are. 135 ...... 1 9 9 2 Jim Hammer (Rapporteur, Session 2) A: Perhaps if I could just say ... it is actually quite difficult to provoke debate. We published, while I was still with the HSE "The Tolerability of risk from nuclear power stations" which John Rimington referred to this morning. We awaited the explosion of indignation; we expected the Secretary of State to be hauled before Parliament; we expected debate; we expected TV programmes; and there was a deafening silence. There were some 15 or 20 extremely considered comments from specialists, academics, participants in the field and one more publicly orientated one, perhaps from the Friends of the Earth; but apart from that the silence was impenetrable. We published all the comments and once again drafted a Press Notice and again there was (I think I'm right in saying) silence. So to stir up a debate on whether nuclear power stations are tolerable ... it is extremely difficult to get a public reaction. I can only say that's our experience. Dr David Ewing (Health and Safety Executive): Q: I would like to ask my question to Dr Sulkowski. It's very presumptuous of me, sir, to react to your comments but it would seem that in Poland you have an extremely well educated population, some highly trained engineers; you know the techniques of risk assessment; you have already identified the hazards. Can I suggest, sir, that perhaps what you need is confidence and that then you can do it yourself Do you really need these Western experts to tell you, at great expense, what you know already? Dr Tadeusz Sulkowski (Chief Labour Inspector, Poland) A: I have tried to present the problem of hygiene limits, that about two-thirds of the substances in use have not still the hygiene limits elaborated and that elaborating this standard lasts very long. And at least in this field I'm quite sure that the collaboration and the assistance of Western partners would be very useful and even necessary simply to avoid what we call 44opening of the open door". Of course you are right and I agree entirely that there are some fields where we have the specialists qite well prepared even in some particular aspects in the field of risk assessment, especially nuclear problems, but there still are a lot of areas where problems of risk assessment are entirely unknown. I understand of course your remark about Western aid and Western experts. Of course the countries of Central or Eastern Europe should ask and use this assistance in a reasonable way. Thank you. Dr Maria Tasheva (National Centre of Hygiene, Bulgaria) I am not from Central Europe, I am from East Europe, I a from Bulgaria, Maria Tasheva, and I would like to join with the statement of the last speaker, Dr Sulkowski, and to talk not about our needs but about what has been done in these two years of social revision in our country. While we need experts for these standards we collected standards from countries such as Canada, Australia, Germany, France and during this time we evaluated all our threshold limit values for about six months and I could suggest our friends from Central Europe to come and sample our experience and maybe in the future have a joint meeting because our technologies and conditions are very similar. And also for risk assessment we have a proposal from EPA, not from European International Organisation, but EPA organised courses in Central and East 136 ...... 1 9 9 2 European countries. I think these courses will start this October and they will be on risk assessment to train the trainer and in the future these courses may be regular in our countries. Thank you. ProfessorDr H J Seidel (University of Ulm, Germany) Q: I have a question to Dr Palecek. In your talk you touched a very complex area and I mean the correlation between environmental pollution or destruction and human health. I think we are still in te position tat we have a need for measurement of human health and indicators of human health. You mentioned life expectancy and in your talk you mentioned endangered pregnancies and infant mortality although you have not really correlated it to the severe environmental conditions in that area. But you said that, and I just am asking because later in your talk you mentioned the severe social consequences of this destroyed environment. Do you think that this area is a possibility to study indicators of human health which are dependent on environmental destruction and not so much on social destruction? This is a deficiency, I think, in health science that we have so few indicators of human health and environment. Thank you. Dr Milos Pale6ek (Director, Occupational Safety Research Institute, Czechoslovakia A: Thank you for this remark. I think that there is a need to prepare a complex project dealing with the whole problem of the influence of environment on social conditions and health in this area. This project is prepared now, it is an inter agency programme, and I think that there is a need to co-operate with foreign experts to find criteria, to find methods and to invite methods so that it would be possible to compare results in this area with other areas. Dr Van der Heijden (European Centre for Environment and Health, WHO, The Netherlands) A: I would like to give some comments on the last two points. It's related to my presentation as mentioned in the Polish situation. You have all this expertise and you do know how to deal with it and make your own standards, you have the best experts and engineers, I believe that's true. The problem is - and in some countries it's more exaggerated than in, for example, the Western countries - the problem is the implementation, the management of the implementation; that science and risk assessment tends to be a goal on its own, and I believe that's a worse situation. We should always see it in a continuum, let's say it's action directed and it's just one part of the whole system. In the last question on the health aspects - let's say for the European Centre - it's one of the main targets, main objectives to find the health-related environmental issues. Also from WHO there are several very national programmes to assist countries in setting up more epidemiology and networking and data processing to establish their own national standards and evaluations. But on the other hand we have to be very careful, it's not only in Eastern Europe, it's not only in the United States, but it's also in Western Europe. Health aspects and health threats are very often misused as is regard for environmental issues and very often the public and the media make believe that they will die or it's very unhealthy but it's not true at all. So I think the basis of our work will be to find what's true and what's not true. 137 ...... 1 9 9 2 ProfessorPacholski's closing renwrks Thank you very much. I hope that today's discussion will be a good starting point for the next discussion ... tomorrow and after tomorrow. 138 ...... 9 9 Tuesday, 6 October 1992 Risk assessinent iin praetitee 139 9 XA04NO315 General Approaches to the Risk Assessment of Chemicals Patrick Murphy Commission of the European Communities Directorate-General XI Environment, Nuclear Safety and Civil Protection Introduction In the context of the UNCED 92 "Earth Summit" in Rio, the following definition of chemical risk assessment has been developed: "Chemical risk assessment is a scientific process that identifies and quantifies the potential adverse effects on human health or ecosystems of defined exposures to chemical substances, to mixtures that include chemicals, or to chemically hazardous processes or situations. Risk itself is the probability of the occurrence of a defined adverse effect in a defined group and in defined circumstances." I would not be so impertinent as to try and improve upon a definition that has the tacit endorsement of the majority of world-leaders. Furthermore, I consider that too many man-years have been spent discussing this topic. Thankfully the UNCED definition recognises chemical risk assessment as being a process and not some immutable physical law. In this presentation I will attempt to explain some of the details and mechanisms of that process but first of all it is worthwhile to spend a few moments putting chemical risk assessment in its proper context and asking the simple question: why do we want/need to assess the potential risk of chemicals? In general terms, chemicals risk assessment is carried out in order to ensure that neither man (consumer/worker/general public) nor the environment are exposed to unacceptable risks arising from the production, use and disposal of chemicals. At a national and/or international level, risk assessments are performed by the regulatory authorities before they accept notification dossiers (e.g. new industrial chemicals) or grant authorizations (e.g. pharmaceuticals, pesticides, cosmetics, food additives). At the local level, plant-operators must carry out risk assessments to ensure that in the particular circumstances of their factory the workers are adequately protected and that satisfactory accident prevention and contingency plans are prepared. Similarly, local authorities must carry out risk assessments before deciding upon the granting of permits for landfill sites or the discharge of toxic chemicals to water or air and in doing so they must take into account the hydrology, geology and climate of the specific locality. While the basic approach to chemical risk assessment will be the same, irrespective of the specific objective for which the assessment is carried out, the details will vary as a function of- the product type (pharmaceutical, pesticide, industrial chemical, etc), the target population of interest (patient, environment, consumer, worker, etc) and the exposure scenario (global, international, national, local). 140 ...... 1 9 9 2 Components of chemical risk assessment While risk assessment is in practice an iterative process it is nevertheless possible to identify several, more or less, discrete components within that process (fig. 1). Data input All chemical risk assessment requires the collection of two types of data: those relating to the potential of the chemical to cause undesirable effects (effects data), and those relating to the environmental fate and behaviour of the chemical and which aow an assessment of the concentration to which the target population will be exposed (exposure data). The quality of any risk assessment is dependent upon the quality and quantity of the input data driving the process. However, the extent and nature of the data to be collected will be dependent upon the precise objective of the assessment. Fig. 2 details the minimum pre-marketing data set (MDP) recommended by the OECD for the evaluation of new industrial chemicals before they are placed on the market. The data requirements specified are relatively modest but, nevertheless, this data package costs some 150,000 dollars to generate. In contrast to industrial chemicals, registration procedures for new pharmaceuticals require far more extensive toxicological data although exposure, being regulated by controlled dose levels, is less problematic: furthermore, environmental concerns are generally of secondary importance. For pesticides, on the other hand, while extensive mammalian toxicological data is still required, information relating to effects on fish, birds and algae are equally important as are data allowing the estimation of exposure concentrations, for example, in the tractor cab, in food and in the environment. If the extent of the data package is important, then so-is the quality of that data. Given problems with inter-operator and inter-laboratory differences compounded, in tests on living organisms, by natural biological variability, the best way to ensure quality and comparability is to use standardised test methods. Methods, or guidelines, developed by organisations such as the OECD(1) constitute one of the essential pillars for developing harmonised international approaches to chemicals risk assessment. One further adjunct to the use of standardised test methods is the application of quality assurance control by means of good laboratory practice (GLP). Again, OECD(2) principles of GLP applied throughout the world are the assurance that tests allegedly carried out according to OECD test guidelines are in fact carried out according to best practice. Hazard identification Hazard identification was defined in the UNCED context as follows: "Identifying the adverse effect which a specified chemical or mixture or process has an inherent capacity to cause. This is developed from determination of chemical and physical properties, epidemiological observations, animal experimentation, in vitro testing or structure activity relationships." From the above definition the essential point to notice is that hazard-identification concentrates on the intrinsic properties of the chemical and ignores the likelihood or extent of exposure to the 141 ...... 1 9 9 2 chemical. At first sight, hazard identification may therefore appear to be of limited interest in the risk assessment process because risk assessment requires that exposure is also taken into account. However, while effects data frequently allow an objective assessment of whether a chemical is explosive, flammable, very toxic, mutagenic, etc. exposure will frequently vary from person to person, ecosystem to ecosystem and country to country. It is therefore often easier to achieve agreement on hazard identification than on risk assessment and many countries and regional organisations use hazard identification as the starting point for their chemical control programmes. One example of such an international/regional programme is the system for the classification and labelling of dangerous substances and preparations established in the European Community. In this system the intrinsic properties of a chemical lead to a classification, a term synonymous with, hazard identification, which in turn results in the requirement to label the chemical in a certain way (see Fig. 3. Classification in certain categories may also lead to additional risk management measures being applied to the marketing of a chemical and or its use in the workplace.(3,4) As indicated above, hazard identification (classification) is frequently the starting point in the risk assessment process. This has been recognised in Chapter 19 of Agenda 21 at the UNCED 92 meeting where it was agreed to develop an international system for the classification and labelling of dangerous substances. Dose response assessment UNCED defined this component in the process as: "Estimating the relationship between dose, or level of exposure, and the incidence and or severity of any effect." For ease of presentation it is convenient to differentiate between dose response assessment in relation to human health and dose response assessment related to the environment. a) Human health - dose response assessment The objective here is to define, on the basis of the available toxicological data, the highest acceptable exposure level in relation to the most sensitive end point. Toxicity data are therefore reviewed to identify the No Observed Adverse Effect Level (NOAEL). Then, in order to take account of the uncertainty in extrapolating across species, various assessment factors (usually in the range I X 1 -I - I X 10-4) are applied to this experimentally determined NOAEL in order to generate a) Maximum Acceptable Concentrations M.A.C. values) for the workplace or b) Estimated Doses of Concern (EDQ or, with adjustments for normal dietary patterns, c)Acceptable Daily Intakes (ADI). The above approach to dose response analysis is probably acceptable when dealing with chemicals and end points for which it is possible to establish a threshold level below which a chemical is expected to have no adverse effect e.g. general systemic toxicity, reproductive toxicity. This approach is based on the assumption that below a certain concentration, the organism's natural defence and de-toxification mechanisms will respond to, and cope with, toxicant challenge but that above this threshold these mechanisms are overwhelmed. However, when dealing with chemicals which are known, for example, to be genotoxic carcinogens, the concept of a threshold is probably meaningless and one has to assume that the chemical may exert an adverse effect at any concentration. 142 ...... 1 9 9 2 b) Environmental effects Here again the objective is the definition of a no concern level, but in contrast to human health effects assessment, the no concern concentration normally applies to the entire ecosystem not to one target species. Natural ecosystems are highly complex and sometimes extremely fragile. Nevertheless, it is frequently the case that environmental effects data is limited to acute toxicity studies on one or two species of fish or aquatic invertebrates (frequently the water flea Daphnia magna). Chronic or long term toxicity studies are not usually available and tests on terrestrial organisms are extremely rare. Furthermore, studies involving multispecies test systems (mesocosms) or artificial ecosystems (e.g. artificial streams) are usually reserved for pesticides. Given the paucity of data the obvious problem is how to determine a realistic no-concern level. In practice, scientists have responded to the high degree of uncertainty by building-in very large safety or assessment factors when extrapolating from restricted data sets to the real environment. The general principle is: the smaller and weaker the data set, the greater the assessment factor. Fig. 4 shows the assessment factors agreed recently by the OECD Hazard Assessment Advisory Body and which have achieved a high degree of concensus in the international Community. Padiway and exposure assessment To this point I have concentrated on the evaluation of probable effects and the objective of defining a safe concentration of a chemical with respect to a given population or ecosystem. We now need to consider an evaluation of the concentration of the chemical. to which the population or environment will be, or is, exposed. This is known as Pathway and Exposure Assessment. In the UNCED context this element in the risk assessment jigsaw was defined as: "Determining the pathways and rates of movement of a chemical in the environment, its transformation or degradation and its concentration, when possible, at critical points thereby estimating the doses to which various populations or ecosystems are actually exposed." Obviously, the most precise method for evaluating exposure concentrations is to measure them in situ. However, for chemicals about to be placed on the market this is clearly impossible and furthermore, for many existing chemicals such monitoring data do not exist. Therefore, in the absence of such information, models must be used to predict or estimate the likely exposure concentrations. There are a variety of such models, some addressing ecosystems others concerned with occupational or consumer exposure. a) Consumer exposure In our own homes we are exposed to a variety of chemicals: household cleaning products - detergents; emissions from paints and varnishes applied to walls and furnishings; pesticides applied to carpets and fabrics; flame retardants applied to fabrics, plastics and upholstery; dyes applied to furnishings and clothing. With the exception of the accidental ingestion of such products by children, the main routes of exposure to these chemicals are by inhalation and through the skin. Our exposure to these chemicals will be a function of their physical properties, their use pattern, whether we are passively or actively exposed, the amounts used, the amount of time we spend in the house, the time we spend in particular rooms and the pattern of our movements inside the house. 143 ...... 1 9 9 2 Obtaining accurate estimates of exposure to chemicals within the home is extremely difficult and can only be established on the basis of detailed market research, analysing how people spend their time within the house as well as how they use certain products. These sociological data are then combined with data on the physico-chemical properties of the substance e.g. physical form, vapour pressure and information on the quantities used, in order to generate estimates of exposure which are usually expressed as the annual dose or the lifetime dose. b) Occupational exposure Occupational exposure is determined as a function of the properties of the chemical, the quantities used, the industrial process used and the personal protective equipment which is worn by te operator. Again, as with consumers, exposure is usually by inhalation or through the skin rather than by ingestion. In the absence of monitoring data, occupational exposure is usually determined on the basis of anology with similar chemicals used in similar industrial processes. Many agencies charged with ensuring occupational safety have, on the basis of experience, developed generic models relating to specific industrial processes and product types e.g. paint spraying in the automobile industry. c) nvironmental exposure assessment In a few, rare, cases monitoring data will be available with real, measured, concentration in the environment but for the most part, environmental exposure concentrations must be predicted using models. Such models will be dependent, among other factors, upon: 1) the stage in the life-cycle of the chemical which is of interest (production/use/disposal); 2) the tonnages of the chemical which are produced or placed on the market; 3) the type of production process or the type of product concerned; 4) the properties of the substance including: solubility, volatility, potential to degrade (physical/ chemical/biological) absorptive capacity, etc. The end product of such models will be a predicted environmental concentration or PEC. Risk characterisation The following definition is to be found in the UNCED documentation: "Estimating the incidence and severity of the adverse effects which are liable to occur in a population or an ecosystem due to actual or predicted exposures. It brings together the results of the dose-response and exposure assessments. The term risk estimation is used when the estimated probability of an effect is precisely stated: risk characterisation is also used to cover less precisely quantified assessments." As made clear in the above definition this last step in the process involves a comparison of the (predicted) exposure concentration and the maximum acceptable concentration. Where the margin of safety between these two values is sufficiently wide then risk reduction or risk management actions will not be required. Where the two values overlap or where the margin of safety is too narrow, risk management initiatives should be considered. 144 ...... 1 9 9 2 For human health protection, this means in practice, that MAC values, or EDC values or ADIs are compared to the concentrations or doses to which consumers/workers will be exposed in the workplace, in the home or through the diet. If the margins of safety are too narrow then restrictions on marketing and use may be required. An example of such a risk characterisation process is given for the chemical methylene chloride in Fig. . For substances suspected of being genotoxic, carcinogens and for which a threshold or cut off value does not exist the situation is less clear. If it is assumed that any exposure to such substances will carry some risk the objective must be to set an acceptable dose. Some authorities use a value of between -5 and 10-7 for such a safe-dose; meaning that the daily exposure of a population to that dose for a lifetime would be expected to increase the incidence of cancer by between I person in 100,000 to one person in 10,000,000. In Western Europe the risk of dying of cancer is approximately I in 4 i.e. 25%, therefore an additional risk of 10-5 would equate to an additional 0.01 % i.e. the total risk of dying from cancer would increase to 25.01 %. The use of the concept of a safe dose is an essentially political decision which gives rise to heated debate. With regard to the environment, a comparison of the predicted environmental concentration with the environmental concern level will yield a hazard quotient. If this quotient exceeds unity then there will be a need to refine the hazard assessment either by further testing or by generating further exposure data. If further refinement still indicates a high level of concern then restrictive measures may be appropriate. Fig. 6 gives an environmental risk characterisation for the substance, Linear Alkyl Benzene Sulphonate which has been used as a component in household washing powders for many years. Discussion Chemicals risk assessment is a relatively new discipline and there are many improvements which need to be made. In particular, the degree of uncertainty regarding exposure concentrations must be reduced. For the most part, regulators will, in the absence of better data, base their exposure predictions on worst case scenarios, which may, or may not be realistic. This is quite understandable, their responsibility is to protect the population and the environment from unacceptable risks and they will always err on the side of safety. Nevertheless, industry, as producers of the chemicals in question, must have more detailed knowledge on likely exposure scenarios. Without such information, it would not be possible for them to place chemicals on the market. Certainly, concepts such as Product Stewardship and Responsible Care, if they are to be more than rhetoric, must require companies to carry out their own detailed risk assessments before marketing a chemical. Therefore, if industry has such data, it should share it with the regulatory community and consumers alike. If industry cannot respond to this challenge then regulators will continue to be "overly" protective. Our knowledge concerning the possible environmental effects of chemicals is deplorable. For many end-points of concern, (e.g. soil functioning, soil chemistry, photodegradation in water and air and anaerobic degradation), we have no agreed standard test methods. For other end points, our methodologies are crude, and unreliable: and our knowledge about inter-species differences and the ability to extrapolate from the laboratory to the real world is totally insufficient. We make a considerable mistake if we continue to invest significant resources in improving the sophistication of testing methods designed to pick out possible human toxicants, but do not make similar efforts in relation to potential effects on ecosystems: are we content that one acute 145 ...... 1 9 9 2 toxicity test on a water flea is sufficient to satisfy environmental concerns? The link between general environmental quality and human health is universally accepted and if we allow our environment to deteriorate then the quality of our own lives will also diminish. Far more research effort should be devoted to understanding te mechanisms by which chemicals impact upon our environment and in developing test methods which will detect and predict such effects. One issue which I have not addressed so far is the communication of information on chemicals risk assessment to the general public and their involvement in the risk assessment debate. With such a complex subject there is a real danger of Regulators and Regulated forming a charmed circle from which the uninitiated are excluded by means of impenetrable jargon. Other speakers in this Conference will discuss this issue in more detail; my personal view is that resolution of this problem would be facilitated if we could identify and separate those aspects which are of an essentially political nature e.g. the desired level of protection (acceptable level of risk); the economic and social costs of achieving or not achieving a given level of protection, from those aspects which are essentially scientific/technical e.g. how to calculate/determine a predicted environmental concentration or a no effect concentration. References 1.Organisation for Economic Co-operation and Development: Guidelines for the Testing of Chemicals. 2. Organisation for Economic Co-operation and Development, Council Decision of 12th May 1981 on the mutual acceptance of data in the assessment of chemicals, Annex 2 Principles of Good Laboratory Practice. 3. Directive 90/394/EEC on the protection of workers from the risks related to exposure to carcinogens at work (Sixth Individual Directive within the meaning of Article 16 (1) of Directive 89/391/EEC) Official journal of the European Communities, Series L, Number 196, pp 17, 26th July 1990. 4. Directive 76/769/EEC on the approximation of the laws, regulations and administrative provisions of the Member States relating to restrictions on the marketing and use of certain dangerous substances and preparations. Official Journal of the European Communities, Series L, Number 262, pp 201-203, 27th September 1976. 5. Organisation for Economic Co-operation and Development, Draft Report on the OECD Workshop on the extrapolation of laboratory aquatic toxicity data to the real environment held in Washington (USA) on 11-12 December 1990 - OECD 1991. 6. Directive 67/548/EEC as amended for the sixth time by Directive 791831/EEC on the approximation of the laws, regulations and administrative provisions relating to the classification, packaging and labelling of dangerous substances. Official Journal of the European Community, Series L, number 259, pp 10-28 of 15th October 1979. 146 ...... 9 9 2 flgure I General Overview of Risk Assessment Data Input Hazard Identification Pathway and Exposure Assessment -Consumers Dose Responce Assessment -Workers a) Human Health -Environment b) Environment Risk Characterisation Risk Management 147 ...... 1 9 9 2 Figure 2 Data Components for the OECD Minimum Pre-marketing Set of Data Chemical identification data Production/use/disposal data Name according to agreed international Estimated production, tons/year nomenclature, e.g. lUPAC Intended uses Other names Suggested disposal methods Structural formula Expected mode of transportation CAS-number Spectra ("finger-print spectra" from Recommended precautions and purified and technical grade product) emergency measures Known impurities, and their percentage Analytical methods by weight Essential (for the purposes of marketing) Physical/chemical data additives and stabilisers and their percentage by weight Melting point Boiling point Density Ecotoxicity data Vapour pressure Fish LC50 - at least 96 hours exposure Water solubility Daphnia - reproduction 14 days Partition coefficient Alga - growth inhibition 4 days Hydrolysis* Spectra Adsorption-Desorption Degradation/Accumulation data Dissociation constant Particle size* Biodegradation: Screening phase biodegradability data Acute toxicity data Bioaccumulation: Acute oral toxicity Screening-phase bioaccumulation data Acute dermal toxicity (partitioning coef., n-octanol/water, fat Acute inhalation toxicity solubility, water solubility, biodegradability Skin irritation Skin sensitisation Eye irritation Repeated dose toxicity data 14-28 days, repeated dose Mutagenicity data *only the screening part to be done for base set 148 ...... 1 9 9 2 Figure 3 Classification and labelling of dangerous substances on the basis of their acute toxicity according to Annex VI of Directive 67/548/EEC on the classification, packaging and labelling of dangerous substances(6) Category Very Toxic Toxic Harmful LD50 Oral 25 25-200 200-2000 rat mgkg- LD50 Dermal 50 50-400 400-2000 rat mgkg- LC50 Inhalation < 025 0.25-1 1-5 mg/L/4 H Very Toxic Toxic Harmful I T + T Xn Symbols to be used on packaging Very Toxic Toxic Harmful 149 ...... 1 9 9 2 Figure 4 Assessment factors for Aquatic Toxicity Data to derive Environmental Concern Levels (OECD, MI) Assessment factor Available information applied to the lowest value NOEC-valuec or QSARd estimate for chronic toxicity derived from a set of data at least consisting of algae, crustaceans and fish l0a acute UE)C50e or QSAR estimate drived from a set of data at least consisting of algae, crustaceans and fish 100b acute 14E)C50 or QSAR estimate for acute toxicity 1000 a), b): the lowest value of the two (a) or (b) may be preferred. c): No observed effect concentration d): Qualitative structure activity relationship - estimate of toxic effect based upon knowledge of the chemical structure e): Median lethal or effective concentration usually determined by experimentation. If NOECs are not available for each of the three taxonomic groups mentioned in the table (algae, crustaceans and fish), an assessment factor of 10 may be applied to the lowest NOEC. This calculated ECL is compared to the concern level calculated from the L(E)C50 values and the lowest value should be selected as the Environmental Concern Level. 150 ...... 1 9 9 2 Figure5 Outline Risk-Assessment for Methylene Chloride Substance: Methylene Chloride Synonyms: Dichloromethane Methane dichloride Methylene dichloride Formula: CH2CI2 World Production (estimate) 570,000 tonnes per annum Types of Use % Neurotoxicity Aerosols 20-25 200-300 ppm, psychomotor and Paint Remover 25 audiovigilance impaired, mild headaches and mild nausea Process Solvent 35-40 500 ppm alterations in visual evoked Misc 10-15 response Above 500 ppm, neuro-depressant Producttypes effects become more pronounced headaches, nausea, fatigue, reduced Paint strippers ability to concentrate Household cleaners Lubricants Carcinogenicity Degreasers Classified by IARC as a carcinogen category 2B (Animal data but human data lacking) Exposure Concentrations IARC also considers there to be (Measured, ppm) evidence of genotoxicity Manufacturing Industry 50-485 Occupational Paint Stripping up to 500 Physiological effects (Exceptionally 1000-2000) Non occupational Paint Stripping up to 500 Exposure to 100 ppm for hours (Occasionally 1000s) results in blood carboxy-haemoglobin I levels in non smokers, of % Assessment indicates the need to reduce exposure. This can be done by local control measures. Wider restrictions on use constrained by doubts as to the safety and efficacy of possible substitutes. 151 ...... 9 9 2 Figure 6 Outline Risk Assessment for a Detergent Substance:LinearAlkylBenzeneSulphonate(L.A.S) C16-H25-S03Na-C19-H31-S03Na Mixture of alykl hornologues C1 0-C1 3 average carbon chain length = 12 Household Washing Machine TOXICITY DATA (All values in tg per litre) 0.9 Kg per capita per annum Acute toxicity for Fish(selection) 0 Average per capita water L. machrochirus1700-10300 consumption per day = 250 Litres 0. mykiss 1700-5500 Dilution of household waste water with industrial waste water = 110 B. rerio 4600-8700 S. alpinus 4600-8700 S. alpinus 4600-8700 S. alpinus 4600-8700 Degree of treatment in sewage works, determines the concentration in the effluent Chronic No Effect Concentrations B. rerio 2000 D. magna 870 No Primary Secondary/ C. virginica 50 Treatment Treatment Tertiary R integra. 4400 Treatment P. integra 4400 P. integra 4400 Concentration in pg per litre Field Studies Outdoor stream 45 day No Observed Effect Predicted concentration in Concentration=350 river assuming 1: 10 dilution - redicted No ect Concentration 50-350 1 5 2 9 9 XA04NO316 Risk Assessment of Chemicals A Central European Perspective Vladimir Bencko(i) Qybrgy Ungmiry(2) (1) Institute of Hygiene, The 1st Medical Faculty, Charles University,Prague, CSFR (2) National Institute of OccupationalHealth, Budapest, Hungary Introduction During the last four decades in the Czech and Slovak federal Republic (CSFR) and Hungary, similar to the previous Soviet Union and all the Central and East European countries the prevention of the adverse health effects of chemicals in occupational and environmental settings, drinking water and food of population was intended t be achieved by determination and compulsory observance of hygienic limit values (MAC, TV, ADI). From recent political developments it can be expected that due to the decay of the former Warsaw Pact and the related Council for Mutual Economic Assistance (CMEA) the next development of risk assessment will reflect that new situation. The OECD principles of toxicity testing and risk assessment of exposure to chemicals will be sooner or later accepted in principle in the framework of a general economic integration of Central and Eastern European countries with EEC. During the 80s there was a growing feeling of the necessity to harmonize the activities of OECD and CMEA countries in the field of toxicological methodology and approaches to risk assessment of chemicals because of the growing production and mutual trade including transport of industrial and agricultural chemicals. As a contributory factor supporting this effort there was growing cross boundary air and river pollution in Europe. The former Council for Mutual Economic Assistance was an intergovernmental organization of 10 countries with a number of other countries and agencies cooperating with it in its work. CMEA was, and OECD still is concerned primarily with economic development but both had stated commitments to protecting human health and the environment. To help to mutual understanding and to promote harmonization of above mentioned obligations and commitments of the both organizations United Nations Environment Programme's International Registry of Potentially Toxic Chemicals and The International Programme on Chemical Safety (UNEP/IRPTC/(IPCS) convened an International Consultation on Toxicometric Methodology held in Moscow in November 1985 and PCS organized a Technical Review Meeting to compare CMEA and OECD Approaches to Toxicity Testing and Risk Assessment in Geneva, May 1987 (UNEP/IRPTS/IPCS 1985; UNEP/lLO/WHO/lPCS 1987) A part of this effort was the publication of a Collection of Training Materials in Preventive Toxicology in the USSR within the UNEP/IRPTC project, Control of Hazards Posed by Chemicals to Human Health and the Environment (UNEP/IRPTC 1984). Calculation of hygienic standard values The methodological pattern of this approach in our countries has been based partially on the results yielded by animal experiments and partially on epidemiological data gathered in industry (Sanotsky, 1974; Ulanova, 1984; CMEA, 1986). As to the animal experiments the spectrum of parameters examined was wider, or at least more differentiated compared with the original the OECD set due to the inclusion of such parameters as e.g. conditioned reflexes and 153 ...... 9 9 2 immunological parameters. This is against the background of previous substantial differences between some of MACs or TWAs values in CMEA and OECD countries. This difference was not of a constant nature and there are some examples of opposing situations e.g. occupational MAC for vinyl chloride VQ which is from four to ten times lower in OECD countries - mg per m' or I ppm (which equals 26 mg per m) compared with the standard of 10 mg VC per M3 in a work room air valid in CSFR, Hungary and Poland. The "limit value" of vinyl chloride was calculated on the basis of the following principles (Ungvdry, 1981), taking into account the results of the studies documenting its carcinogenic effect (Maltoni et al, 1974; Janysheva and Balenko, 1966; Shabad, 1979), as well as some physiological parameters. Table 1: Tumour frequency in Sprague-Dawley rats exposed to VC (exposure: 4 hours/day, days/week, for 52 weeks) VC Initial No. No. of Tumours of Nephro- Hepatic Other Brain Other All concentration of animals zymbal-gland blastomas angio- angio- neuro- tumours animals surviving sarcomas sarcomas blastomas Control (air) 68 58 0 0 0 0 0 10 10 50 64 59 0 1 1 1 0 10 13 250 67 59 0 6 4 2 0 7 19 500 67 59 4 4 7 2 0 8 25 2 500 74 59 2 6 13 3 5 7 36 6 000 72 60 7 4 13 3 7 10 44 10 000 69 61 16 5 9 3 7 11 51 Maltoni et al 1974 - observation up to the 120 week. Based on the data shown in Table I we have established, that the malignant tumour frequently induced by VC redoubles with the increase of the dose (y = 10.7 times XO.3). Frequencies of VC and spontaneous tumours occurring in the control group (0/58, 10/59, respectively) do not differ significantly from the frequencies of VC and spontaneous tumours in Sprague-Dawley rats exposed to 50 ppm VC concentration 3/59, 10/59, respectively); ppin = 26 mg m3 VC. The Sprague-Dawley rats weighed 600-900 g at the age of I year. Taking into account the 52' weeks of exposure, the mean body weight can be tken as 600 g. On this basis, the respiratory volume of the animals is: 21 600 075 ml/min = 945 ml/min I I/min. The total volume of the air inhaled by one animal during the whole exposure time equals: 60 4 x 5 x 52 = 62.400 litres = 62.4 m3 in the case of 50 ppm VC concentration this volume contains 8 11 0 mg of V in case of complete absorption 8112 g of VC gets into each rat weighing 600 g, which is not carcinogenic. (Only 40% of VC is absorbed, the rest is exhaled - Bolt, 1980.) Considering a man weighing 70 kg: 70 X8.110 g VC=946.4 g VC (which is not considered carcinogenic for Sprague-Dawley rats) 0.6 If the respiratory volume of a worker during light physical work is 15 litre/min, during 40 years of employment with 50 weeks of 40 working hours each, the total volume of inhaled air will be: 15 x 60 x 40 x 50 x 40 litres = 72 000 000 litres = 72 000 n3 154 ...... 1 9 9 2 946.4 g VC divided by 72 000 m3 gives the concentration which is not carcinogenic. This is the following: 946 400 mg - 13.14 MgJM3 72 000 m 3 Considering that 1. Sprague-Dawley rats are more sensitive to the tumourigenic effect of VC than the other kinds of rats, as well as hamsters and rabbits; 2. According to human epidemiological data, VC hepatic angiosarcorna had occurred in places where several hundred ppm of VC exposure lasting for decades was proven, the value of 0 mg/m 3 seems to be well founded, as it contains a safety factor of at least 30-100%. Based on these, the workplace limit value (MAC) of VC recommended by us is 10 rng/m 3. This means that the concentration of VC in the work area must not exceed 10 mg/m 3. The MAC value of the vinyl chloride monomer is determined by the concentrations, doses which do not cause tumours, mutations, and immune defects on the basis of the data presently available (Bencko et al. 1988). The other harmful effects appear only at significantly higher exposure values. In the CSFR, Hungarian and Polish workplaces the maximal allowable concentration (MAC value) of vinyl chloride determined as a standard is 10 mg/m .3; the MAC value is marked by the small letter k, which indicates the carcinogenic property of the substance and at the same time that the substance has no threshold value which could guarantee the protection against tumourous diseases ("carcinogens have no threshold value"). Thus, this, MAC value represents an acceptable risk. The Hungarian ambient air hygienic standard for VC immission gives 5 /Ag/M3 limit value for the priority protection areas both for 24 hours and 30 minutes, and 100 tk&3 limit value for areas of 11 category protection both for 24 ours and 30 minutes. In Hungary there is no VC monitoring either in the residential areas or in the natural environment. A limit value is used for emission calculations when planning and operating factories and sanctions against factories are based on calculation of emissions. In the CSFR the regional ambient air standard for VC MAC day mean 30 Ag/rn3, and 80 Ag/M3 that cannot be exceeded during 30 minutes sampling period has been accepted. In contradistinction to Hungary, in the CSFR the concentration of VC is regularly monitored in the vicinity of chemical plant polluting the environment with VC emissions in Upper Nitra Valley, Middle Slovakia region. In the majority of our countries, all chemicals listed in the standard ("Air purity requirements in the workplace") are controlled (measured and registered) by environmental monitoring. Most significant air contaminants such as Sx, NO,, and particulate matter are also controlled by continuous environmental monitoring in residential, ambient air of industrial agglomerations. Chemical contamination of drinking water is controlled systematically, while the chemicals in food (in vegetables, fruits) are controlled only randomly. Unfortunately - possibly with the exception of drinking water - the environmental monitoring is not yet satisfactory. 155 ...... 1 9 9 2 In contrast with environmental monitoring activity one of the most important accomplishments of occupational hygiene and health in some Central and East-European countries is that, for about 20 chemicals, biological exposure indices are being regularly monitored, and if their levels exceed the biological limit values, tey must be reported. A report is followed by the workplace physician and/or the hygienist of the locally responsible control body to identify te source responsible for te increased exposure and to control it. For example, in Hungary, between 1982 and 1991, 20 chemicals were responsible for about 20004000 cases of reported increased exposure annually, all detected by this monitoring. This represents at least 15%-2.0% of those occupationally exposed to the monitored chemicals. Just for illustration the total number of chemicals used in Hungarian industry is estimated to be between 20,000 and 40,000 (Ungvdry, 1990, 1991). Generally special attention is paid to both environmental and biological monitoring of mutagenic and carcinogenic chemicals. The maximum allowable concentration (MAC) of nickel valid in the workplaces is 0.0005 mg/m3. The letter "k" with the MAC value in the standard indicates that the substance is carcinogenic. (All chemicals listed in 1, 2A or 2B category by the International Agency for Research on Cancer (IARC) are marked by letter "k" in the standard: "Air purity requirements in the workplace"). Limit values guaranteeing safety for carcinogens generally are not recognized today. This principle, however, cannot be observed in case of chemicals, which are at the same time essential elements. Nickel is well known to play an important role in the regulation of the cardiovascular system (Anke et al., 1984; Mancinella, 1991). Thus it can be included in the list of substances having two limit values - one minimal and one maximal. Above the maximal limit value the substance increases the frequency of tumours, while below the minimal limit value it causes deficiency. Based on this principle, nickel has a real biological and environmental limit value (excluding the risk of tumours). The biological limit value of nickel is 100 jtg NO blood 1.70 Itmol/din3). The number of workers with nickel exposure in Hungary can be estimated as 500-600. Biological monitoring of nickel has been compulsory since 1986. The number of cases of increased nickel exposure reported varied between and 20 in the years 1986-1991. Due to the small number of cases of increased exposure reported, changes of the risk of tumours caused by nickel could be hardly determined even if the time since the first year of reporting - 1986 - was longer than 10 years which is necessary to the manifestation of tumours (Ungvdry, 1992). In the hair of workers exposed to nickel the concentrations of nickel increased (up to a mean value of 216.75 1),g1g). At the same time immunological parameters (IgG, IgA, IgM) were elevated. A significant rise in the concentration was also recorded in the case of AIAT, A2M, CPL and LYS (Bencko, et al., 1986). It can be concluded that biological monitoring and immunological parameters increase the validity of risk assessment of adverse health effects of nickel. Biological monitoring of environmental pollution by toxic metals The data provided by biological monitoring can constitute a basis for identifying areas excessively contaminated with industrial emissions, including their geographical delimitation. This aspect is of growing importance from the point of view of public health authorities. Against the background of growing interest is the simple fact tat the total extent of environmental pollution is often difficult to assess, both qualitatively and quantitatively. Analyses of non-systematically collected air and surface water samples yield virtually worthless data in this respect, for the actual degree of environmental contamination may vary across a relatively wide range. 156 ...... 1 I...... 9 9 2 Ideally from the technical point of view, a continuous measurement of the environmental pollution can be effected through the use of a network of automated monitoring systems preferably those capable of automatic sampling, analysis, registration and evaluation of data. These automated monitoring systems are not easily accessible at present, both technically and economically, and their use in the near future is in Central Europe expected to remain limited to localities having the greatest degree of environmental pollution (Bencko, 1991). As an alternative to the technical approach to this problem, biological indicators could be used to monitor pollution of the environment. What we understand by the term biological indicators could be used to monitor pollution of the environment. What we understand by the term biological monitoring is a systematic collecting of human material or other biological specimens with a consequent determination of the substance concentration, its metabolites or biotransformation products in these specimens (IAEA, 1976). This approach appears to be particularly well suited to demonstrating environmental pollution by potentially toxic trace elements, including toxic metals. Already in the first half of the thirties Teisinger 1936) used biological methods for monitoring exposure in persons occupationally exposed to lead. With respect to the use of this approach for monitoring of environmental pollution mention should be made of the pioneer work of Svoboda (1936) who described approximately at the same time the so called Tesin's disease in honey bees caused by arsenic contamination of the environment. Both these examples demonstrate that at least half a century old Czechoslovak tradition of using biological methods to monitor the environmental pollution and the exposure of man to toxic trace elements, including toxic metals. It is now common knowledge that e.g. peat moss plants, due to the special anatomic structure of their respiratory system, have the ability to accumulate toxic metals which are present in the air in the form of aerosols. As an example, impacts on animal kingdom species such as the above mentioned count losses or even extinction of honey bee populations may serve to indicate environmental pollution in areas affected by emissions containing arsenic. Of animal species, game animals such as hares have repeatedly proved to be useful as indicators of environmental pollution (Novdkovd and Paukert, 1973). Considering the depression which affected the hare population in Central Europe and also the fact that the use of hare is limited to a relatively short time period of the hunts in the autumn, some authors decided to try the use of a murid rodent, the common vole. The advantage of the common vole lies in the fact that it can be caught all the year round and that its radius of activity amounts to about 10 m only, so that this species may provide more detailed information (Paukert and Obrusnik, 1986). In our field studies we preferred to test biological material from domestic rabbits (Bencko et al., 1981). The examination of plant and animal samples is able to complete the information obtained by the examination of humans. It may be even assumed that the changes of body burdens of toxic metals in animals start earlier than those in man, because the animals are exposed to the impact of contamination more directly, by all routes, including the local food chain. Thus, the free-living animals might signal in advance the danger threatening the human inhabitants. Interestingly, hematological changes found in hares living in an area polluted from a known source of industrial emissions were comparable wit tose encountered in local children. Similarly, we found virtually identical concentrations of arsenic in the hair of children and hair of rabbits living in the same locality (Bencko, 1990). 157 ...... 1 9 9 2 The health risk assessment of environmental pollution is becoming increasingly a major public health concern. In this context a primary concern is the assessment of human exposure. For this the examination of a suitable human tissue appears to be more appropriate than the analysis of plant or animal materials currently used in ecological studies to demonstrate environmental pollution. The human biological materials that are accessible for sampling include blood and urine, but also hair and nails. Successful attempts have also been made to measure e.g. lead, and of non metal elements fluorine accumulation in deciduous teeth to demonstrate over exposure to these elements in children (Cikrt, Bencko, 1990). 6 Values of arsenic in 6) 5 the hair of children. Source of emissions 4 marked "O" C: Cn 3 - 0 0 CZ 2 - A B C D E F G I J L -- M N 0 ------I------0 0- :10 0 --- 10 3 Distance of dwelling place from the source of air pollution (km) Figur I Our study conducted in the mid-60s demonstrated that determination of arsenic in human hair has potential utility as an indicator of the environmental contamination with this noxious agent (Bencko, 1966). Later studies Figure revealed that a correlation can be established between arsenic content of hair and the expected degree of arsenic contamination of air (Bencko and Symon, 1977). These studies show that the tests for arsenic content of hair, used for years to demonstrate the existence of arsenic exposure for the purpose of forensic or industrial toxicology, are equally applicable to environmental pollution monitoring if the method of group examination is applied. Today, extensive data are available in the literature, including monographs which document the advantages of this approach. The usefulness of hair analysis for the demonstration of exposure to toxic metals in man has also been repeatedly confirmed in a series of our studies concerned with toxic trace elements e.g. manganese, lead, nickel, cobalt, antimony and mercury (Bencko et al., 1986). Sampling, transport and storage of hair samples is incomparably easier than sampling, transportation and processing requirements of blood or urine samples, the most common specimens used today to demonstrate human exposure to a variety of noxious environmental agents. Human hair samples are also very easy to preserve for later control re-analyses. 158 ...... 9 9 2 In contrast to blood and urine specimens, hair should first be cleansed of external contaminants prior to element analysis. Despite the fact that there is no general concensus on how the washing procedure should be done the procedure recommended by IAEA 1977) should be preferred to other techniques that are not standardized. In this context it should be noted, however, that excessive external contamination of hair may reflect massive exposure to a given environmental pollution, which is actually what is to be demonstrated by these tests in most instances (Bencko, 1990). Naturally, the use of human hair analysis technique is far from being the universal tool for monitoring exposure to environmental pollutants and considering the broad spectrum of pollutants encountered in the general environment one can hardly expect that such a screening tool would ever exist. However, for a great majority of toxic trace metals this technique has proved well suited for the monitoring of over exposure of man originating either from man-made or natural sources. Conclusion On basis of the examples given we have tried to demonstrate some specific features of risk assessment of exposure to chemicals in environmental and occupational settings. The objective of our presentation is not an exhaustive description of existing differences between previous CMEA and present OECD approaches to risk assessment and management. Although the approach to risk assessment and management was similar in many respects in the CMEA countries, implementation and hygienic practice was different in the individual countries in terms of many details and effectiveness. Due to long lasting experience with environmental pollution including health impact on humans e.g. in the "Dirty Triangle of Europe" and other heavily contaminated areas a considerable knowledge has been gained. In some instances there are quite unique examples of exposure of humans to specific pollutants in environmental and occupational settings. Bearing in mind that in our part of the world: - workplaces are frequently in poor condition from both technical and technological point of view; therefore the workplace contamination, and the risk of adverse effect of 4 4workplace-chemicals" (among them carcinogenic and mutagenic ones) is relatively high - emissions arising from industry are not adequately controlled, the environment in industrialized areas is highly contaminated - the residential air in those areas and capitals of our countries is frequently extremely contaminated by industry, local heating and traffic - SO., CO, NO, 03, PAH, Pb and particulates - rivers draining industrial areas are heavily contaminated and some areas have severe problems with drinking water quality - the food basket of our population is contaminated by some agrochemicals due to their local over use e.g. by Cd, nitrates, PCBs 159 ...... 1 9 9 2 - a decisive part of industrial as well as municipal waste is not properly treated and disposed - at least part of the gathered knowledge has been published in local languages only, another part has been "confidential" and not allowed to be published. For that we would like to propose: - to collect existing knowledge concerning both methodology and local epidemiological studies including those in Central and East European countries - to analyse critically and evaluate the knowledge and experience and present it to the international scientific community and international institutions and organizations such as UNEP, ILO, IPCS, IRPTC and last but not least OECD - these institutions/organizations could take part in harmonizing further activities. To achieve this we would recommend the organization of - post graduate education or seminars for experts from Central and East Europe dealing with risk assessment and management to harmonize approaches to the problem - to introduce regular interlaboratory quality control practice procedures for environmental and biological monitoring - to establish an international system of regular exchange of information concerning environmental quality control and management. References Anke, M., Groppel, B., Kronemann, H., Grun, M. 1984) Nickel - an essential element. IARC Si. Publ. 53: 339-65. Belitsky, A., Borzsonvi, M., Ilnitsky, A. P., Pdrkdny, M., Pint6r, A., Shabad, L.M. 1979) Hygienic standards of carcinogenic chemicals. In: Carcinogenic Chemicals in our Environment. Ed.: Borzsonyi, M., Szabvdnvkiad6. Budapest. Chap. III., p. 319-332. Bencko, V. 1966) Arsenic in hair of non-occupationally exposed population. Cs. Hyg. 11: 948-957 Reprinted in A Coll. Stud. Hlth. Effects of Air Poll. on Children. U.S. Publ. Hlth. Service 3 pp. 948-957 Bencko, V., Symon, K. 1977) Health aspects of buring coal with a high arsenic content. Ist Part: Arsenic in hair, urine and blood in children residing in a polluted area. Environ. Res. 13: 378-385. Bencko, V., ArbetovA, D., SkupenovA, V. 1981) Use of domesticated rabbit tissues for monitoring of environmental pollution by toxic metals (Mn, Pb, Cr, Cd, Ni). J. Hyg. Epidemiol. Microbiol. Inimunol. 25: 113-120. Bencko, V., Wager, V., WagnerovA, M., ZavAzal, V.: 1986) Human exposure to nickel and cobalt: Biological monitoring and immunobiochemical response. Environ. Research 40: 399-410. Bencko, V., Geist, , ArbetovA, D., Dharmadikari, D. M., Svandovd, E. (1986a) Biological monitoring of environmental pollution and human exposure to some trace elements. J. Hyg. Epiderniol. Microbiol. Immunol. 30: 1-11. Bencko, V., Wagner, V., Wagnerovi, M., Bdtora, J., Hrebacka, J.: 1988) Immunochemical profiles of workers differing in the degree of occupational exposure to vinyl chloride. J. Hyg. Epidemiol. Microbiol. Immunol. 32: 375-384. 160 ...... 1 9 9 2 Bencko, V. 1990) Biological monitoring of human exposure to toxic metals by hair analysis. In: Biological monitoring of exposure to chemicals: Metals. K. Dillon, M.H.Ho(Eds.) J. Wiley Intersci. Publ. New York, pp. 245-256. Bencko, V. 1991) Ambien air pollution in Czechoslovakia. In: Proc. Symp. Air Pollution in Central and Eastern Europe: Health and Public Policy. B. S. Levy (Ed) Manag. Sci for Health, USA (Boston) Fr dek-Mistek (CSFR) P. 0- 18. Bolt, H. M. 1980) Metabolic activation and pharmacokinetics in hazard assessment of halogenated ethylenes. Int. Conf. on Industrial and Environmental Xenobiotics: Biotransformation and Kinetics. Prague. Abstracts 9. Cikrt, M., Bencko, V 1990) Biological monitoring of human exposure to metals. J. Hyg. Epidem. Microbiol. Immunol. 34: 233-241. CMEA 1986) Problems of industrial toxicology Eds.: Kaloyanova-Simeonova, F., Korbakova, A. I. (in Russian) Soviet Ekonomiceskoi Vzaimopomosci (CMEA), Moscow, pp. 101. IAEA 1976) Advisory group meeting report: Application of nuclear methods in environmental research. Attachment 2 22-26 March. Janysheva, N. Ja., Balenko, N. V. 1966) On experimental lung cancer caused by the introduction of varions doses of 1 2 5, 6-dibenzantracen (in Russian). Gig. 4 Sanit. 6 2-15. Maltoni, C., Lefernine, G., Chicco, P., Carretti, D. 1974) Vinyl chloride careinogenesis: current results and perspectives. Med. Lav., No 65 11-12): 421444. Mancinella, A. 1991) Nickel, an essential trace element. Metabolic, clinical and therapeutic considerations. Clin. Ther.: 138 3-4): 159-65. Noviikovd, E., Paukert, J. 1973) Influence of industrial missions on cation contents in the hairs of common hare (Lepus Europaeus Pall). In: Proc. XIth Int. Cong. of Game Biologists. Stockholm, pp. 423-438. Paukert, J., Obrusnik, 1. 1986) The hair of the common hare (Lepus Europaeus Pall.) and of the common vole (Microtus Arvalis Pall.) as indicator of the environmental pollution. J. Hyg. Epidemiol. Microbiol. Immunol. 30,1: 29-35. Sanotsky, 1.V. 1974) Prevention of harmful chemical exposure of man - an integrated problem of medicine, ecology, chemistry and engineering (in Russian) Zh. Vses. Khim. Obsch. in Mendeleyeva, 19, No 2 (quoted from Ulanova 1984). Svoboda, J. 1936) So called "Tesin disease" of bees and its treatment (in Czech) Sb. Cs. Akad. Zemed. 12: 589-594. Teisinger, J. 1936) Eine rasch mikropolarographische Metode zur kvantitativen Bestimmung des Bleies im Blut. Z. ges. exp. Med. 98: 520. Ulanova, 1.P. 1984) Fundamentals of toxicometry. In: Preventive Toxicology. Collection of Training Materials. Vol USSR Commission for UNE, Moscow, p. 271-28Z UNEP/IRPTS 1984) Preventive toxicology. Collection of training materials. Vols I-III. USSR Commission for UNEP, Moscow. UNEP/IRPTS/IPCS 1985) International consultation on toxicometric methodology. USSR-UNEP, Moscow, pp. 13. UNEP/ILO/WHO 1987) Report of the IPCS technical review meeting to compare CMEA and OECD approaches to toxicity testing and risk assessment. WHO, Geneva, May 1987, pp. . Ungv6ry, G. 1981) Hygienic-toxicological qualification of vinyl chloride - toxicological data for MAC value of vinyl chloride. A review (in Hungarian) Munkav6delem. No 27 10-12): 34-50. Ungviry, G. 1990) Occupational health in Hungary: Occupational safety, hygiene and health care. In: Environment and Health in Eastern Europe. Proc. Symp. On Occupational and Environmental Health during Societal Transition in Eastern Europe. P6cs. pp. 75-79. Ungviry, G. 1991) Workplace air contamination in Hungary. In: Proc. Symp. Air Pollution in Central and Eastern Europe: Health and Public Policy, Management Si. for Health, USA (Boston), Fr dek-Mistek (CSFR), p. 19-23. 161 9 XA04NO317 Pesticide Risk Assessment in the United States Richard N. Hill Environmental Protection Agency Washington, DC Pesticides by definition are dangerous substances, since they are designed to kill, mitigate, or control some type of pest - whether it is plant, animal or microbe - and unlike most other chemicals in commerce, they are deliberately released into the environment in large amounts by direct application to land, water or air. High-level exposure may be encountered by persons applying pesticides and those around treated materials as well as by non-target organisms on land and in water following pesticide treatment. Chronic and widespread but low-level exposure occurs in the general population from the consumption of pesticide residues on treated foods. Considerable benefits occur from pesticide use including such things as the production of a wholesome, plentiful and inexpensive food supply; the sterilization of medical instruments; the treatment of homes for termites; and the application of insect repellents. Thus, pesticides are agents that have both significant benefits and significant potential risks to humans and the environment. In recognition of potential risks, all pesticides distributed and sold in the United States must fulfil extensive registration requirements for the Environmental Protection Agency (EPA). Registration is a licensing procedure where industry must submit data to demonstrate the safety of pesticidal substances and products before they can be used commercially. The regulatory control of pesticides is unique among chemicals in the U.S. in that testing beyond initial registration may be imposed by the Agency throughout the commercial life of the chemical, as long as there is adequate justification. Registration requirements are gauged to the nature of potential exposures. For instance, more data are generally needed for food use registrations than for non-food uses because of direct consumption of treated foods by the whole U.S. population. Unlike pesticide practices in many countries and authorities, as in the European Community where agricultural pesticides, non-agricultural pesticides and genetically engineered microbial agents are handled by separate directives, all pesticide activities are covered in the U.S. by the Federal Insecticide, Fungicide and Rodenticide Act. This statute covers pesticide uses on foods and animal feed and a number of non-food applications like forest and horticultural uses, residential lawn care, in-home applications, and disinfectants/sterilants. Traditional inorganic and organic chemicals are covered, as well as biological agents like pheromones. Naturally occurring and genetically altered microorganisms also come under the definition of pesticides, but multicellular animals are exempt from regulation as pesticides. Pesticide registration in the U.S. as in many other countries may be a long-term, resource intensive undertaking. Not uncommonly the process from beginning to complete registration may take 4 to 10 years and cost about $10 million. To meet the responsibilities of reviewing studies, overseeing 400 active ingredients and 35,000 products, and implementing other aspects of the statute, EPA employs about 900 people. Unlike the approval process and actions to cancel pesticides in some countries, in the U.S. they are generally much more transparent. The need for specific test guidelines and the nature of the protocols are debated in the scientific community. Test data adequacy is evaluated according to 162 ...... 1 9 9 2 objective criteria, and study findings are reviewed using standard evaluation procedures and risk assessment guidelines that have been through public comment. Agency risk concerns for a registered pesticide are vetted at a meeting of its Scientific Advisory Panel, a group of technical experts outside of government, where the potential risk case is reviewed and discussed. The meeting is open to the public with opportunity for input from industry, public interest groups, and other parties. Formal Agency regulatory proposals to cancel registered pesticides are published for public review and comment, and there may be several rounds of public involvement before a final decision is reached. Even after a regulatory decision has been made, and unless a settlement has been reached, the Agency is often sued by an environmental group or industry that questions the EPA position; in such situations, the case is then transferred to the court for deliberation. Risks are evaluated for a host of different effects by the pesticide program: acute and chronic, human health and ecological. For illustration, the assessment of cancer and aquatic ecological effects is presented. Cancer assessment For about the last 20 years pesticidal chemicals have been evaluated for potential carcinogenic effects. The Agency has developed risk assessment guidelines that serve as a basis for judging the evidence for human carcinogenicity and the means of estimating cancer risks. The hazard classification is patterned after the one that was originally developed by the International Agency for Research on Cancer. At the heart of our determination is the premise that cancer effects in animals serve as a sentinel for potential effects in humans. When EPA first started making estimates of cancer risk, it used scientific information available at the time which indicated that carcinogens were reactive molecules (or were metabolized to such molecules) that interacted with the DNA of cells, caused mutations and commenced the carcinogenic process. Since these mutagenic events might occur after single chemical-DNA interactions, it was posited as a worst-case scenario that cancer dose-response relationships could be linear. On the one hand, from evaluations of laboratory animal cancer studies on numerous different chemicals in the 1980s, it became apparent that many laboratory animal carcinogens were not reactive mutagens as had been thought previously. This led some scientists to conclude that linear dose-response relationships would not be applicable. Instead, tey felt that sub-linear or threshold considerations would apply in these cases. On the other hand, statisticians argued that if a chemical carcinogen worked through mechanistic processes that occurred in the absence of chemical treatment, then the effects of the chemical would add to the background processes and cancer incidence and lead to low-dose linearity. In the mid-1980s, the U.S. government and the EPA espoused linear extrapolation of cancer risk for cases which the carcinogenic mechanism was unknown. More recently, scientists have been developing significant data that challenge the conservative principles that cancer in animals is generally always relevant to humans and that linear cancer dose-response relationships apply in all cases. For instance, EPA recently reviewed the data on the means by which chemicals produce kidney tumors in male rats via interaction with alpha-2u- globulin and concluded this tumor was not predictive of human carcinogenicity. The Agency also has proposed a science policy stating that thyroid tumors in animals that arise from the disruption of thyroid-pituitary homeostasis would show a dose-response threshold at the point where normal endocrine hormone is perturbed. 163 ...... 1 9 9 2 EPA is moving to address other changes in its cancer risk assessment practices. For instance, in the past the Agency sometimes relied upon very high "bounding" estimates of exposure to ensure that exposures were not underestimated; we are now moving to incorporate both realistic high end and central tendency estimates. For major food-use pesticides, we are also requiring the development of market basket residue data to give better indication of real ingestion exposure levels rather than relying on residue levels on raw agricultural products. In addition, a recent policy directs Agency programs to discuss the uncertainties that accompany all aspects of risk assessment: alternative interpretations of controversial aspects are presented, and a rationale is required to justify selection of a given option. It is also recognized that the carcinogenic risk assessment procedures used in the U.S. differ in some ways from those employed by other nations. EPA is revising its cancer risk assessment guidelines and will be soliciting broad scientific and science policy input on the proposals including parties outside the U.S. Even with the steps that are being taken to update the carcinogen review process at EPA based on current scientific understanding, in some cases the Agency must ban pesticides on foods that have been shown to produce cancer. Whenever the processing of an agricultural commodity results in the concentration of the pesticide residue, as in milling, cooking and mechanically drying, the 1958 Delaney clause of the Federal Food, Drug and Cosmetic Act applies. This provision prohibits the introduction of additives into the human food and animal feed supply that have been found to be carcinogenic in animals or humans. It covers any animal carcinogen, regardless of its relevance to human carcinogenicity and any animals or human carcinogen regardless of how low is the exposure and anticipated risk. A recent attempt by the Agency to challenge part of the Delaney clause and exclude animal carcinogens with estimated lifetime upper-bound cancer risks of one in a million or less, failed in court. Ecological assessment Testing of pesticides for ecological effects is tiered depending upon the potential for hazard and exposure. For instance, pesticides are evaluated for acute aquatic toxicity in an invertebrate and in freshwater fish. Depending upon test outcomes and the ratio between expected environmental exposure to the agent and the toxicity level, more in-depth testing may be warranted. Further laboratory analyses may include evaluation of aquatic reproduction and growth parameters and, in some cases, field testing is required. EPA started using field studies to confirm serious effects seen in the laboratory because of the lack of information on how to extrapolate laboratory study results to those expected in the environment. In one type of field study, an aquatic niesocosin, 12 ponds of 300 in 3volume are constructed 4 doses with 3 replicates). One set of three ponds serves as controls, while the other 3 sets receive prescribed pesticide loadings. The middle dose is considered to be a "reasonable worst case" exposure that would be expected to occur at least once in a growing season under typical pesticide-use practices. The other two dosed sets are below and above that level. Each pond receives a balanced ecosystem of plants and aquatic invertebrates and vertebrates. Pesticide effects at the top of the ecological food chain is the monitored endpoint of concern, that is, reduction in the biomass of finfish. Results are considered significant when adverse effects are noted in either the middle or low-dose groups. These study results are considered to be important because the test design incorporates real anticipated environmental exposures, dose-response relationships, the ability to evaluate the persistence of effects over time and measurement of direct ecological effects at the population and community levels. On the negative side, however, the studies are long term 2 years) and costly ($1.5 million). Now that the Agency has mesocosm 164 ...... 1 9 9 2 study results on 7 different pesticides, it plans to evaluate the results to determine whether 1) the study outputs are worthy of the investment; 2 the existing results can be used to develop a model to predict pesticide-induced effects; or 3 the present study results can be extrapolated to other pesticides. Opportunities for harmonization Pesticide regulation has developed independently in many countries with little overall coordination among most areas. With the development of the European Community it became evident that mechanisms were needed to coalesce the pesticide regulatory procedures used by member states. More recently Sweden took the initiative to host a meeting of international pesticide regulators to identify areas that may be ripe for coordination through the Organization for Economic Co-operation and Development (OECD). This seminal meeting was quickly followed by a special session of OECD earlier this year to address some of the differences among pesticide practices in various nations and the differences between information needs for pesticides and those for other chemicals. Several topical areas were identified for future development, and preliminary decisions have been reached on others. One area deals with data requirements and test guidelines. Pesticide authorities will try to identify a core set of test data that will be applied uniformly across countries, while leaving open the possibility of some unique requirements under certain circumstances. Although OECD has already made major steps in developing test guidelines that are applicable to all chemicals, it was recognized that some data needs are specific to pesticides (e.g., environmental fate). To this end, current OECD test guidelines will be reviewed as to their adequacy or pesticide registration, and lists will be made of acceptable ones, those needing modification and endpoints where new guidelines are needed. Discussions through OECD have also identified a desire to develop consistent interpretations and use of test data across authorities for both health and ecological effects. Common risk assessment practices need to be drawn up for consideration and approval; cancer assessment will be the first area for review. In regard to hazard classification and communication, there is a U.S. policy to develop a single classification system to cover all chemicals in commerce (e.g., pesticides, workplace, transportation); this goal has been affirmed by OECD and, more recently, a commitment was made at the United Nations environment meeting in Brazil to implement uniform classification systems. The International Programme for Chemical Safety will take the lead in harmonizing risk assessment and hazard classification practices. Another significant area for international coordination concerns pesticide reregistration. Currently, many countries are in the process of updating their files on registered pesticides, filling test data gaps according to current protocols and reviewing the safety of present-day pesticide application practices. The U.S. will host a meeting of regulators this fall to begin work on the coordination of schedules for reregistration reviews. Such efforts could result in a significant reduction in resources, enhanced pace of outputs and consistent application of assessment procedures across authorities. All these international efforts deserve special attention as we move into a truly global economy. 165 9 XA04NO318 An Industry Approach to the Risk Assessment of Pesticides Dr Barry Thomas Schering Agrochemicals Ltd., Chesterford Park Research Station, Saffron Walden, Essex, U.K. 1. Introduction The regulatory control of pesticides has developed over the last 40 years during which time major changes have occurred, not only in the scientific basis of risk assessment but also in the socio-political perception of pesticides and of the agricultural and chemical industries. Traditionally, and logically, the registration of pesticides has been based on a pre-marketing risk- benefit assessment but changes have occurred in the relative importance associated with the two elements of this assessment. These potential risks have assumed a greater importance as has the acceptability of such risks. By contrast the benefits of using plant protection products to increase agricultural productivity and production has assumed less importance in the light of perceived agricultural surpluses, at least in the Developed Countries. This paper will consider current and future regulatory requirements for pesticides and identify some of the key areas which are of importance to the Plant Protection Industry. It will also discuss initiatives, by both the Industry and Government, aimed at ensuring the safe use of plant protection products, how such initiatives are likely to impact on new product developments and the consequential effects on global food supplies. 2. The registration of pesticides and key issues of risk assessment It is worth stating at the very outset that the pesticide industry is one of the most heavily regulated and that data requirements for registration or re-registration are amongst the most demanding. Pesticides are therefore comprehensively tested both with regard to the quality and the quantity of data. Standards continue to increase within the industry in response to scientific advances and the increasing demands of Good Laboratory Practice (GLP) requirements. Data are generated to enable the potential risks to users, consumers of treated food, domestic animals, third parties, wildlife and the environment to be assessed. These data therefore cover a broad spectrum of scientific disciplines including physico-chemical, analytical and residue chemistry, plant and animal metabolism, mammalian toxicology and toxicokineties, environmental toxicology, environmental fate, and biology. The extensive and comprehensive data required by Regulatory Authorities to support the application for a new pesticide (or to re-register one already on the market) may take up to 7 years to assemble and represents a considerable financial investment on the part of the Company. Because of the costs of developing a new crop protection product (an estimated 50 million dollars when the cost of unsuccessful compounds is taken into account) development can only be justified on the basis of a global market or, at least, a significant proportion of that market. Consequently the data base has to satisfy a large number of different Regulatory Authorities throughout the world. Often the evaluation of this common data base leads to different regulatory decisions and by implication to different assessments of the risk associated with the 166 ...... 1 9 9 2 proposed use of the product. Whereas in some cases differences in local factors such as climate, agronomic practices, etc, may need to be given different considerations in the risk assessment process, the variability in regulatory decisions more often reflects differences in the basic assessment of the data. A number of examples will serve to demonstrate this point as well as to identify areas of concern to Industry in relation to perceived and actual risks. Thus: 2.1 Toxicology Toxicological data requirements have developed over the years until by today they have reached a fairly stable plateau, although there are indications that additional demands for data relating to neurotoxicity and immunotoxicity may be forthcoming in te future. During this development period the emphasis has clearly changed from initially a concern with acute toxicological effects to the current preoccupation with longer-term effects and, in particular, carcinogenic potential. Long-term studies are conducted in the rat (for chronic toxicity and carcinogenicity), in the mouse (for carcinogenicity) and in the dog (for chronic toxicity). This combination of studies thus provides data on chronic toxicity in two species (rat and dog) and carcinogenicity, also in two species (rat and mouse). These studies are both expensive and time consuming. For example, a 2-year study in the rat may, when account is taken of the time needed to complete the extensive histopathology and report writing, take up to 3 years to complete and will cost about 650,000 at UK prices. The aim therefore is to complete one study which will satisfy all Regulatory Authorities and which will incorporate any special protocol requirements of individual Authorities. As the USA is one of the world's major markets a prime objective with most pesticides is therefore to meet the requirements of the US Environmental Protection Agency (EPA) and in this respect the Agency's strict adherence to the concept of the Maximum Tolerated Dose (MTD) is of particular importance in the risk assessment process. Without going into the technical details of the MTD requirements, suffice to say that test animals must be dosed at the maximum level which will be tolerated without compromising the life span of the animal. In very many cases this requirement leads to dose levels which bear no relationship to the potential human exposure which will ensue from the use of the pesticide. It can therefore lead to biological, biochemical and toxicological effects which are difficult to interpret. Thus a major problem which often arises from complying with MTD requirements is that tumours occur at these high doses because of biological mechanism initiated by such doses - mechanisms which often have no parallel in either other species or man. These problems are further exacerbated by the Agency's application of complex mathematical models to the data thereby arriving at a highly conservative estimate of the carcinogenic potency of the test compound. Regulators then assign a degree of accuracy to these predictive estimates which the models do not justify but which nevertheless seem to offer a black and white decision-making process, i.e. exceed a fixed limit and no registration is possible. The weight of evidence, the qualitative nature of the effects and the use of the pesticide are thus relegated to matters of secondary consideration and instead the compound is invariably classified as a "possible" or even worse a "probable" human carcinogen. This in turn can cause regulatory problems in those countries which look to the USA for guidance but do not fully comprehend the subtleties or idiosyncrasies of the EPA's risk assessment process. Perhaps from a public point of view this concern with the carcinogenic potential of pesticides is understandable as cancer, as a disease, still holds a certain dread about it. More and more 167 ...... 1 9 9 2 people are perceived as dying from cancer whereas in reality, if lung cancer due to smoking is excluded, the actual incidence of other cancers in most countries is either static or declining (Ref 1). Certainly there has been no drastic increase in incidence since the 1940's when the modern pesticide industry first started. The whole area of the carcinogenicity of pesticides has been reviewed by Bruce Ames in a series of publications often aimed at the comparison between 4 4synthetic" pesticides and toxins naturally occurring in plants, e.g. aflatoxin, hydrazines, estragole, etc., few of which have been investigated to the extent that pesticides have. Ames has concluded that pesticide residues in food and water are likely to be of minimum concern relative to background levels of natural substances and that the intake of what Ames describes as "nature's pesticides" is likely to be 10,000 times higher than man-made pesticides (Ref 2 3. Despite significant support in the scientific community for Ames' views the perception of carcinogenic risk arising from the use of pesticides remains unchanged. 2.2 Residues Residues in food, domestic animals and animal by-products, have assumed a very high importance since the very early days of pesticide registration. In this respect little has changed except, of course, technological advances in analytical chemistry have resulted in dramatic improvement in methodology and sensitivity. Cynically one might argue that the increase in levels of detection have contributed to public concern and regulatory control in this area. The concept of parts per million or billion means little to the man in the street - if you can measure it, it must be dangerous! Over the years a vast amount of residues data has been accumulated and includes not only those results published in the literature, which in itself is substantial, but also the large body of data which has been generated by companies in support of registration and which remains unpublished because it shows the absence of residues. All this information clearly indicates that the large majority of analysed samples, and I would guess this could be as high as 98%, do not contain any detectable residues. Of the remainder then the detected levels are of no real toxicological concern. Nevertheless Regulatory Authorities continue to demand large amounts of residue data to support an application for the registration of a new pesticide or for a new use of an already registered pesticide. Data must often be "locally" derived despite the availability of relevant data from other countries indicating the absence of residues of any toxicological significance. It is as if Regulatory Authorities are constantly seeking reassurance that there really is no problem. 2.3 The environment In contrast to the data requirements in the residues and toxicological areas, those in the ecotoxicological and environmental fate areas have increased dramatically over the past few years. This is clearly indicated by the comparisons made in Tables and 2 which illustrate the differences between present data requirements and those required some 30 years ago. 168 ...... 1 9 9 2 Table 1: Eotoxicological data requirements 1950's 1990 Birds: acute toxicity Birds: acute oral toxicity 2 spp) Fish: acute toxicity Birds: 8-day dietary toxicity 2 spp) Bees: acute toxicity Birds: Reproductive effects Fish: acute toxicity 3 spp) Fish: early life stage tests Fish: hioaccumulation Daphnia: acute toxicity Daphrda: reproduction study Shrimp: acute toxicity Algae: acute toxicity Bees: acute toxicity Beneficial insects: toxicity Soil fauna: effects Earthworms: toxicity Soil micro-organisms: effect Table 2 Environmental fate requirements 1950's 1990 Soil: residues Soil: Adsorption desorption Soil: Leaching Soil: Lysimeters Soil: Photodegradation Soil: Metabolism Water: Metabolism Water: Photolysis Water: Hydrolysis Soil/Water: Sediment studies Soil/Water: Biodegradation rate One issue however which is currently of major regulatory importance and which clearly demonstrates the conflict between political objectives and scientific principles is the impact of the European Community Drinking Water Directive (80/778/EEC). This Directive set limits on the amount of pesticides in drinking water, namely 0. 1 ttg/litre (0 I ppb) for individual pesticides and 0.5 itg/litre for total pesticides. It is acknowledged by all concerned that these limits have no scientific basis and bear no relationship whatsoever to toxicological risks to consumers. What they reflect is a political decision that the mere presence of pesticides in drinking water is not acceptable. The only acknowledgement to science is that the limits were not set at zero! The implementation of this Directive has caused much debate and coincidentally much unnecessary concern to consumers, but to date there seems little chance of amending either the limits per se or of amending the Directive to reflect a more pragmatic and scientific approach to the potential presence of low levels of pesticides in drinking water. Indeed rather than an improvement in the situation what we are actually witnessing at present is an exacerbation as the limit of 01 ttgffitre is extended beyond water consumers to include the environment in general. Thus we are seeing these limits being applied to all water sources, i.e. both ground-water and surface water, when considering the registrability of new or indeed the continued use of existing pesticides. In my opinion the Directive also serves to illustrate how not to undertake a risk assessment unless, of course, one assumes that the objective of the Directive was to establish a zero risk in this area. 169 ...... 1 9 9 2 3. Harmonization of data requirements and of risk assessment The harmonization of data requirements and of the subsequent risk assessment based on these data are, in my opinion complementary although I would suggest that the former, albeit difficult in absolute terms, is easier to achieve than the latter. The harmonization of regulatory data requirements is by no means a new concept and attempts by, for example, the Council of Europe and FAO have spanned many years. Whereas some degree of uniformity has no doubt resulted from discussions at these fora, full harmonization of data requirements has proved extremely difficult to achieve in practice. Indeed even within the European Community itself where harmonization is a political cornerstone, initial attempts by the Commission failed to succeed. The influence of the Single European Act has undoubtedly added a political impetus to achieving the harmonization of pesticides and it would be gratifying to report that moves in this direction had been approached on a sound scientific basis. Regrettably this has not been the case and the EC Directive (91/414/EEC) on the Community harmonization of pesticide registration (due to be implemented in July 1993) has initially focused its attention on the procedural aspects of pesticide registration. The data requirements which are currently under discussion have, in turn, concentrated on data for the formulation and proposals for requirements for the active ingredients are not expected until the end of 1992. Similarly proposals currently under discussion for the harmonization of the interpretation of these data are most advanced in relation to the product. In effect these proposals aim to ensure that Community Member States adopt a common approach to their risk assessment of crop protection products. The basic approach is to lay down certain criteria with regard to toxicological, ecotoxicological and environmental properties which if exceeded precludes the possibility of registration unless a subsequent risk assessment indicates that registration is still possible. As an example a pesticide which has a soil half-life of more than 3 months would not be registered unless it could clearly be demonstrated that accumulation in soil does not occur, that unacceptable phytotoxic effects on, or unacceptable residues in succeeding crops do not occur, and that there is no unacceptable impact on non-target species. This "cut-off" criteria approach to the registrability of pesticides is one which causes Industry considerable concern in that it once again detracts from the weight of evidence approach to evaluation and overall risk assessment. Whatever the outcome of the ongoing discussions it is clear that a harmonized data base will eventually emerge for the European Community. This is however only the first step towards achieving global harmonization. Industry however welcomes initiatives by the OECD to aim towards this objective and thus extend the concept of global harmonization so as to include the USA, Japan, Australia, etc. Clearly this will be a major and long-term undertaking. 4. Impact on new compound development and industry safety initiatives There can be no doubt regarding the advances made in crop protection chemistry over the last thirty years or so; dramatic reductions in rates of application and increased target specificity are but two examples of the areas where significant improvements have been achieved. This period of intensive development has been accompanied by an equally intensive escalation of the regulatory control of pesticides, as escalation which indeed still continues. 170 ...... 1 9 9 2 Thus the discovery, development and marketing of a successful new crop protection product has become increasingly more difficult and also more expensive. This is most dramatically demonstrated in the decreasing number of innovative companies active in this area of business and in the fact that such companies must synthesize some 25,000 active ingredients so as to obtain any chances of success in finding that elusive new compound. One might argue that the future implementation of the Registration Directive is merely another step along the regulatory path and that Industry will adapt to its impact and implications, much as it has done with similar changes in regulatory control at the National level. Whereas there is an element of truth here inasmuch as Industry will learn, perhaps somewhat painfully, to cope with the Directive in a mechanistic sense, it would be over-simplistic in the extreme to underestimate the future impact of the Directive. This is no mere change, however significant, at the Member State level but rather a change which will apply to the whole Community and to the very large Western European crop protection market. Because of the difficulties and financial risks inherent in the development of a new compound, this development must be aimed at a world market or at least at those major parts in that market, i.e. USA, W Europe, Japan. Regulatory barriers to the registration of new compounds within the Community could therefore preclude the development of a new compound on a broader global basis. In recognizing the difficulties Industry has recognized the need to change its role in the crop protection area. Thus Industry will no longer merely "sell chemicals" but must rather market a complete plant protection service including the appropriate technical advice, and Integrated Pest Management Systems. Industry would also contend that there is also another key factor in the equation relating to the risks associated with the use of crop protection products. Thus whereas, as discussed above, much emphasis has been placed on the risk assessment of pesticide usage, the attention given to risk management has received proportionally less attention. Pesticide registration is essentially a pre-marketing exercise in which the end-point can be regarded as te availability to the user of a product wich has been fully evaluated and wich, in te form of te product label, is accompanied by instructions for safe and effective use. The actual use of the product however depends to a very great extent on the farmer and on his technical ability and knowledge. Clearly a well trained and professionally responsible user can minimise any potential risks to himself, to others and to the environment. In the United Kingdom a major part of the Control of Pesticide Regulations introduced in October 1986 was the introduction of a legal requirement for users of pesticides to be adequately trained. It is understood that the European Commission are considering the introduction of a similar requirement on a Community basis. This approach to user training is one which is fully endorsed by Industry and which has recently been extended on to an international basis by the GIFAP (the International Trade Association for the Crop Protection Industry) Safe Use Project. The intention of the project is to obtain a substantial improvement in all aspects concerned with the manufacture, distribution and use of crop protection products in the developing countries. Recent reviews of the situation in typical countries have confirmed that current practices are falling short of the standards required by the FAO Code of Conduct for Pesticides (Ref 4 to which GIFAP subscribes. The Safe Use Project will be initiated in pilot countries in each region, progressively spreading the results to other countries. The countries selected are Kenya for Africa, Thailand for Asia and Guatemala for Latin America. In each country an experienced fun time project leader has been appointed. 71 ...... 1 9 9 2 The goal of the Project is to ensure that at all levels of involvement from manufacture through packaging and labelling, distribution, transport, storage, application and disposal of crop protection products, real measurable improvements are made. Progress will be monitored to ensure this. Conclusions The Safe Use Project referred to above is important in another context in that it focuses attention on the Developing Countries, most of which do not enjoy the luxury afforded to us in the West, namely an adequate food supply. The world population continues to increase at a rate of 200,000/day and is expected to reach 6 billion by the year 2000. It is estimated that world food supply must increase by 75 % to meet this increase in population (Ref 5). An estimated 40,000 people, mainly children, die from starvation every day. It has become "fashionable" in the Western world to consume "naturally l or "organically" grown food. Whilst such foods have a place in the market for those who wish to purchase, and can afford to purchase, these foods, any suggestion that organically grown food can be implemented to meet national let alone global needs is non-supportable. The FAO has estimated that, even with the use of pesticides, 20-40% of food production is lost annually to pests and disease (Ref 6); without pesticides food shortages would occur on a massive scale. In this context it is worth it to remember tat "organic" farming was the "state of the art" in Ireland in 1846 when potato blight ed to the death of a quarter of a million people. These facts clearly support the continued need for pesticides in food production and an equally convincing argument can be made to support their continued use in the production of non-food crops such as cotton, rubber, etc., and in the public ealth area for the control of insect-borne diseases and pests such as cockroaches, fleas, etc. In conclusion therefore I would contend that food production must be the predominant factor for the future well-being of us all. The potential risks associated with the use of pesticides must be kept in perspective and the data requirements must likewise be founded on sound scientific principles and not on the misconception that "more data means safer products". Pesticides are an essential tool in ensuring an adequate food supply and we in the Western world, from where most of the anti-pesticide movement originates, must recognize that "In the developed world we should count ourselves lucky that the main risk to health we face from food is eating too much of it" (Ref 7 Would that we could say this for the whole world. References 1. British Medical Association, Living with risk. A report of the BMA, Professional and Scientific Division, John Wiley Sons, Chichester, 1987). 2. Ames, B. N., Science, 221 1983), 1256-1264. 3. Ames, B. N., Magaw, R., Gold, L. S., Science 236, 1987), 217-280. 4. FAO, International Code of Conduct on the Distribution and Use of Pesticides, FAO, Rome 1986). 5. Blaxter, Sir K., People, food and resources, Cambridge University Press, 1986). 6. FAO, Agriculture towards 2000, FAO, Rome 1981). 7. British Medical Association in (1). 172 ...... 9 9 XA04NO319 Radiation Protection Standards A Practical Exercise in Risk Assessment Roger H. Clarke Director, National Radiological Protection Board Member, International Commission on Radiological Protection Abstract Within 12 months of the discovery of x-rays in 1895, it was reported that large doses of radiation were harmful to living human tissues. The first radiation protection standards were set to avoid the early effects of acute irradiation. By the 1950s, evidence was mounting for late somatic effects - mainly a small excess of cancers - in irradiated populations. In the late 1980s, sufficient human epidemiological data had been accumulated to allow a comprehensive assessment of carcinogenic radiation risks following the delivery of moderately high doses. Workers and the public are exposed to lower doses and dose-rates than the groups from whom good data are available so that risks have had to be estimated for protection purposes. However, in the 1990s, some confirmation of these risk factors has been derived from occupationally exposed populations. If an estimate is made of the risk per unit dose, then in order to set dose limits, an unacceptable level of risk must be established for both workers and the public. There has been and continues to be a debate about the definitions of "acceptable", "unacceptable" and "tolerable" and the attributing of numerical values to these definitions. This paper discusses the issues involved in the quantification of these terms and their application to setting dose limits on risk grounds. Conclusions are drawn about the present protection standards and the application of the methods to other fields of risk assessment. Introduction The control of exposures to ionising radiation is an example of risk assessment and its application to an environmentally hazardous agent. This paper describes what is currently known about the risks of radiation exposure and how that knowledge is applied to setting standards for protection. Finally the paper considers whether the system of radiological protection can be applied to other carcinogens. It is useful first briefly to review the history of radiation protection. Roentgen discovered x-rays in 1895, and in 1896 the first paper appeared reporting radiation damage to the skin of the hands and fingers of the early experimental investigators. Radium was also used for therapy soon after Becquerel's identification of radioactivity, also in 1896, and in the next ten years several hundred papers were published on the tissue damage caused by radiation. Several countries were actively reviewing standards for safety by the start of the First World War, but it was not until 1925 that the International Congress of Radiology was formed and first met to consider establishing protection standards. This Congress established the "International X-ray and Radium Protection Committee" in 1928, which evolved into the present International Commission on Radiological Protection (ICRP). 173 ...... 1 9 9 2 The early recommendations were concerned with avoiding threshold (deterministic) effects, initially in a qualitative manner. A system of measurement o dosimetry was needed before protection could be quantified and dose limits could be defined. In 1934 recommendations were made implying the concept of a safe threshold - under "satisfactory working conditions a person in normal health can tolerate exposure to x-rays to an extent of about 02 roentgens per day"(1 - about ten times the present annual dose limit. The tolerance idea continued and in 1951 "the figure of 2 r per week seems very close to the probable threshold for adverse effects" led to a proposed limit of 03 r per week for low LET radiation(2). In considering neutrons and alpha-particles, it was stated that "anaemia and bone damage appear to have a threshold at I ACi Ra-226". By 1954 the threshold was rejected(3) and "maximum permissible doses were such as to involve a risk which is small compared with other hazards in life" and "since no radiation level higher than natural background can be regarded as absolutely 'safe', the problem is to coose a practical level that, in the light of present knowledge, involves a negligible risk". The change was brought about by te emerging epidemiological evidence of excess malignancies amongst American radiologists and the first indication of excess leukaemias in the survivors of Hiroshima and Nagasaki. The problem had become one of limiting the probability of harm and much of what has subsequently developed related to the estimation of that probability of harm and the decision on what level of implied risk is acceptable or, more importantly, unacceptable. The problem arises in both working and public environments and both issues will be dealt with in turn. First, however, it is useful to review what is currently known about the long term effects of chronic radiation exposure. Carcinogenic effects of ionising radiation The major effect of occupational or public exposure to radiation appears to be the appearance, at long times after exposure, of a small excess of cancers in any irradiated population. It is not possible to distinguish a radiation-induced cancer from one arising naturally, so that any estimate of the risk has to be made from a statistical analysis of the long term health of irradiated populations. Although there are several population groups who were exposed in the past to high doses of radiation and from which estimates of cancer risk can be made, including patients treated for ankylosing spondylitis and other medical conditions, or those exposed at work, the single most important source of information is from the survivors of the Hiroshima and Nagasaki atomic bombs. There are some 76,000 survivors of Hiroshima and Nagasaki for whom individual dose histories exist. The number of excess cancers in te population has continued to increase as the follow-up in epidemiological studies continued to the mid-1980s. Lifetime cancer experience is not yet available for any of the large epidemiological studies; for example, over 60% of the Japanese survivors are still alive. Amongst the 30,000 who died, there has so far been a statistical excess of approaching 400 radiation-induced cancer deaths in the 6000 or so cancer deaths that have occurred in the population. Therefore to project the lifetime cancer risk for an exposed population it is necessary to use mathematical models that extrapolate forward, in time, data based only on a limited period of the lives of the individuals. 174 ...... 9 9 2 The data now available provide evidence that, for radiation-induced cancer, a multiplicative model is the most likely representation of the effect of irradiation, at least for the more common cancer types. In the multiplicative model, the time distribution of the excess risk follows the same pattern as the time distribution of natural cancers, i.e. the excess (after latency) is given by a constant multiplying factor applied to the age-dependent incidence of natural cancers in the population. An implication of the use of the multiplicative model is that for the majority of solid cancers it results in an increasing risk with time after exposure, following the natural increase with age, as shown in Figure 1. There are indications, at least in some groups of a decline in the excess at long periods after exposure. This is well documented for leukaemia, but is also seen in the solid tumours from the uranium miners (lung cancer), and the patients irradiated both for spondylitis and for enlarged thymus (thyroid cancer). On the other hand, there has been no indication of a decrease in relative risk up to 40 years in the Japanese survivors, the tnea capitis patients (skin and thyroid cancer), and the North American breast cancer studies of women irradiated for mastitis or pneumothorax. Because of the use of a multiplicative risk model, estimates of risk for the Japanese A-bomb survivors have to be applied to other populations with different baseline "natural" cancer rates. ICRP has tackled this problem by calculating risks to other populations in two ways: first, the relative risks by age, sex and site in the A-bomb survivors can be applied to the "natural" cancer rates in the new population; and second, the absolute numbers of cancers to date in the Japanese are used to give absolute risks, and these risk factors are applied to the new population to estimate a set of risks relative to the baseline cancer rates of the new population. In both cases risks are projected into the future using the multiplicative model. Figure I Scheinatic representation of the ultiplicative model Baseline (spontaneous) cancer risk ------Total risk in irradiated population Annual cancer risk 401*4/ J,!0 1#40 J! 1# 20 80 2 0 8 0 Age, y Age, y 175 ...... 1 9 9 2 ICRP has used these two methods since there is no agreed method of transferring risks across populations and it is certainly true that cancer rates vary considerably between populations. The calculations were performed for five populations representative of the different parts of the world with varying natural cancer rates and the results were averaged. The most comprehensive human evidence for carcinogenesis induced by radiation is obtained from groups given single doses at relatively high levels. If observations were directly available at chronic low doses, the questions about thresholds and the shape of the dose-response curve would be irrelevant. But such observations are unlikely because the size of the study population, in order to give the same statistically significant result, must increase as the square of the reduction in dose to the population; a 100 fold lower dose needs a 10,000 times larger population. As a result, theoretical considerations and experimental results have to be considered to estimate risks at low doses and low dose rates. In summary, after transferring risks across populations and extrapolating to low doses and dose rates, the nominal risk factor is 1-5 MSV-1 for a population of all ages As a perspective on this number, natural background whole body radiation gives a dose of about I mSv per year. The total risk varies by less than 50% between the sexes, the higher values applying to women. The variation with age is more marked - the estimated risk for ages 020 years is more than twice the figure for all ages. For a population of working ages, defined as ages 18-64 years, the risk figure is about 80% of the fatal cancer risk for a population of all ages. Some support for these risk factors has come from the first analysis of the National Registry for Radiation Workers which covers some 95,000 occupationally exposed workers in the UK(4). Acceptable levels of risk The existence of doses in all parts of the body from natural sources of radiation decreases the importance of the shape of the dose-response relationship at doses close to zero. Small doses are always in addition to the natural background dose. For moderate doses above background, a linear relationship between the incremental dose and incremental probability of a deleterious effect will be an adequate approximation whatever the true shape of the relationship between dose and risk. It follows that the discussion on the limitation of risk to individuals must include not only the risk per unit dose but also what risks are acceptable. All human activities or lack of activities carry some risk. Some of the activities are accepted by most people even if the risks are rather high, e.g. traffic accidents. Other activities are not accepted because the risks are considered unjustifiably high in relation to the ensuing benefits even after reasonable attempts at risk reduction. Many attempts have been made to set an upper level of risk to an individual, i.e. a level of risk which would not be acceptable even if it could not be futher reduced. The source of the risk is important and there is a difference here between voluntary and imposed risks. For radiation protection purposes the relevant circumstances would be normal occupational or private life in what might be considered a safe society. A report on Risk Assessment by a Study Group of the British Royal Society was produced in 1983(5) and was used by NRPB in giving guidance in 1987(6). In 1988 the Health and Safety Executive produced a rationale for tolerable risks(7) in response to a recommendation by Sir Frank Layfield, the Inspector at the Sizewell B Pressurised Water Reactor Inquiry. In 176 ...... 1 9 9 2 order to discuss the issues, it is first necessary to clarify the terminology and this has been done by ICRP in their latest recommendations(8). ICRP has found it useful to use three words to indicate the degree of tolerability of an exposure or risk. They are necessarily subjective in character and must be interpreted in relation to the type and source of exposure under consideration. The first word is "unacceptable". which is used to indicate that the exposure would not be acceptable on any reasonable basis in the normal operation of a practice of which the use was a matter of choice. Such exposures that are not unacceptable are then subdivided into those that are "tolerable", meaning they are not welcome, but can reasonably be tolerated, and "acceptable", meaning that they can be accepted without further improvement, i.e. when protection has been optimised. In this framework, a dose limit is set at a level of risk selected at the boundary in the region between "tolerable" and unacceptable" for the situation in which dose limits apply, i.e. the control of specified practices. The limit then protects the individual from all sources under control by ensuring the total risk is not unacceptable. There will also be a level of risk that is trivial, and the source will automatically be considered acceptable. If the risk is above the trivial level then optimisation of protection from the source must be undertaken. It follows that the optimisation process for any single source must be constrained to the maximum acceptable individual risk so that the risk from that single source does not cause concern and the combined risk from all sources under control does not become unacceptable. The relationships are shown diagrammatically in Figure 2. Figure2 Schematic diagram of the acceptability of risk - unacceptable dose limit tolerable ...... maximum acceptable risk from a single source A acceptable after optimisation V ...... trivial level of risk acceptable without optimisation Having decided on the vocabulary to be used, it is possible to review the Royal Society and HSE reports in an attempt to derive the numerical values that correspond to each word. The considerations are different for workers and members of the public. (a) Occupational risk For workers, the Royal Society group concluded that a continuing annual probability of death of I in 100 would be clearly unacceptable, since the individual would be almost certain to die from the occupation. On the other hand, an annual probability of death of in 1000 could hardly be called totally unacceptable provided the individual at risk knew of the situation, judged he had some commensurate benefit as a result, and everything reasonable had already been done to reduce the risk. The HSE report concluded that, broadly, a risk of death of I in 1000 per year is about the most that is ordinarily accepted under modern conditions for workers in the UK, and adopted it as the dividing line between what is just tolerable and what is intolerable (unacceptable). 177 ...... 1 9 9 2 Both the Royal Society and HSE arrive at the same judgement on the level of unacceptable occupational risk of I in 1000 per year, although the former argued down from the clearly unacceptable to what was hardly totally unacceptable, while the latter argued up from the norm to what was bordering on intolerable. There does not seem to have been much discussion on acceptable risks in industry. Clearly a wide range of risks applies across different occupations. Annual fatal risks of I in 10,000 seem widely accepted across industry so that the maximum that might be accepted is probably between 1-4 and 1-3 y-1. The tolerable level of risk must therefore be just below 10-3 y-1, but probably not as low as 10' per year. (b) Public risk In the case of risks imposed on members of the public, the Royal Society and HSE reports are more divergent in their views. HSE argue down, stating that if the maximum tolerable risk for workers is I in 1000 per year, the maximum level that they would be prepared to tolerate for members of the public is not less than 10 times lower, i.e. I in 10,000 per year. Such a level, they observe, equates to the average risk of dying in a traffic accident. The Royal Society Study Group on the other hand, argues up from the lower level of what is a negligible risk to individual members of the public. The Study Group felt that a risk as low as I in 1,000,000 per year was commonly regarded as trivial, and while in some circumstances it could be ten times lower (I in 10,000,000), it may be as much as ten times higher (I in 100,000). For members of the public, the Royal Society Study Group said that there was a widely held view that few people would commit their own resources to reduce an annual risk of death which was already as low as I in 100,000; however, if there are grounds for suspecting a real risk, at an annual level of I in 10,000, the imposition of that risk is likely to be challenged. In its 1987 guidance the Board stated: "The maximum acceptable annual risk for a member of the public is around -51, and ,, Risks much beyond 10-5 y-1 probably verge on the unacceptable for involuntary risks tolerated by members of the public." The question of the maximum level of dose to be tolerated by members of the public was debated at the Public Inquiry into the Hinkley Point 'C' Pressurised Water Reactor. In his report, the Inspector, Michael Barnes QC, said on this point:(9) "When cross-examined, Dr. Clarke of the NRPB said that he thought a figure of in 10,000 per year as the maximum tolerable level of risk for members of the public was a rather high figure. He had previously stated in writing that he considered a risk of 3 1-5 3 in 100,000) per year for the public as verging on the unacceptable. I do agree that it is difficult to see how certain of the public documents which were quoted (e.g. the Royal Society Study Grup and the Recommendations of the ICRP) support a level of tolerable risk for members of the public as high as I in 10,000 per year." The Inspector concluded by saying that he would not have recommended consent be given to the construction of the PWR if, inter alia, the risk of fatal cancer to any member of the public exceeded I in 100,000 per year. It seems therefore that a reflection of society's present view, as expressed by the Hinkley Inspector, is that a risk of I of 100,000 per year is the maximum acceptable risk from a single source for members of the public and that, in summary, it can be concluded that for risks imposed on members of the public 178 ...... 1 9 9 2 an annual fatal cancer risk greater than 3XIO-5 per year is unacceptable an annual fatal cancer risk between 3X IO-' and 10-5 per year is tolerable an annual fatal cancer risk of 10-5 per year is the maximum acceptable for a single source It is also important to note that NRPB advice(10) on the criteria to be used for the land disposal of solid radioactive waste include a maximum risk constraint, on the exposure of an average member of the group representative of those most highly exposed, of in 100,000 per year from a single repository. If two or more disposal facilities are planned at the same site, the constraint applies to the sum of the risks. Control of doses to occupationally exposed individuals In order to recommend limits to exposure and constraints on optirnisation, it is necessary to compare the risks of exposure with the risk acceptance criteria derived above. The attributable fatal cancer rate as a result of working from age 18 to 65, i.e. 47 years at annual doses of 10, 20, 30 and 50 mSv are shown in Figure 3 The attributable fatality probability as a function of age tends to follow the probability of death from cancer for an individual aged 18 because of the use of a multiplicative model. The peak risk rate therefore arises at an age in the ate-70s. The question to be asked is what is to be compared with the numerical criterion that the level of unacceptable risk is 10-3 per year. This criterion would be exceeded at an age in the mid-50s for someone receiving 50 rnSv y-' and in the early-60s for someone receiving 20 mSv y-'. Exposures would need to be kept below 15 niSv y-'so as never to exceed an annual risk of 10-'. But is a peak risk in the later years of life as important as added risks earlier in life? In Table I various attributes of the risk are given for a range of rates of exposure: the annual attributable fatal risk at age 75 is compared with the average annual risk to age 75, and the lifetime attributable risk of fatal cancer. There are also non-fatal cancers and hereditary effects of radiation, which ICRP have weighted for their severity and added to the fatal cancer risk to give a measure of detriment from radiation exposure. There has been no agreed method of combining the different health effects of radiation into a single detriment quantity. In Publication 60 ICRP has produced a protocol for weighing the various effects and it does not seem unreasonable to accept this procedure at present. Detriment calculated in this way is included in Table . The relative importance of these different attributes has to be judged when making a decision on acceptability. On the basis of the data presented above, ICRP recommends dose limits of an average of 20 mSv y-1 over years (100 mSv in years) widi no more than 50 mSv in a single year. At this rate of exposure, the lifetime risk of induced fatal cancer is nearly 4, which with added weighted allowances for non-fatal cancers and hereditary defects is %, which may be compared with the lifetime risk, inferred by the maximum tolerable risk of 10-3 per year, of 4.7 % for work frorn age 18 to 64, and the natural risk of dying of cancer of towards 25 The average annual attributable fatal cancer risk is 7 10-' and detriment 10-3. These levels of risk seem to correspond to the most that will be tolerated and therefore mark the borderline of unacceptability. The Board has advised that workers exposures should be controlled so as not to exceed an average dose of 15 mSv y-'. Tis level of exposure should ensure risks in the long term are tolerable to those exposed occupationally. In many cases, worker doses will not need to be this high and constraints will be set at considerably lower levels of dose. 179 ...... 1 9 9 2 Figure 3 Annual risk of death from radiation induced cancer for 10, 20 30 and 50 mSv y -I received from age 18 to 64. 0.003 - 7 0.002 - 30\ CZ 20 CZ Annual risk LL 0.001 ------ 0 0 50 100 Age: y Table 1: Attributes of detriment for occupational exposure Annual dose (mSv) 10 20 30 50 Attributable lifetime probability of cancer death (%) 1.8 3.6 5.3 8.6 Detriment %) 2.5 5.0 7.4 12.0 Attributable annual fatal cancer risk at age 75 6.5 10-' 1.3 10-' 2.10 -3 3.3 10-' Average annual attributable fatal cancer risk from 18-75 3.5 10-4 6.9 10-4 1.0 lo-, 1.6 10-3 Control of doses to members of the public For members of the public, the same approach as for occupational exposure has been used to consider the different results of exposure over a lifetime received at 02 03, 0.5 and I mSv y - . Figure 4 shows the time distribution of fatal cancer r;sk and Table 2 the peak and average annual attributable fatal risks, lifetime attributable fatal risk and detriment. Although suggestions of upper limits to acceptable levels of imposed risk have been made, it is clear that again judgements have to be made about whether the time at which that risk is received is important. Added risks late in life may be less important than risks added in earlier years. On the basis of considering these risk levels and the variation in natural background radiation (excluding radon for high levels of which intervention is recommended), ICRP recommends a dose limit of I mSv in a year. The lifetime fatal cancer risk at this rate of exposure is 04% which represents an increase of about 15% of the natural probability of dying of cancer. The average annual risk up to age 75 is just under 3 10--. This corresponds to the level of unacceptable risk outlined above, and represents the ICRP judgement on an unacceptable level of risk to the individual. The Board is proposing that the constraint on optimisation from a single source should be no more than 03 mSv.y-I corresponding to an average annual fatal risk of 10-6 which has been said is about the most that is acceptable. The associated detriment would be just over 1 - 180 ...... 1 I...... 9 9 2 per year on average. For most sources of public exposure, lower constraints are expected to be set for the optimisation of protection. There is a further reason to consider that the maximum constraint of 03 mSv is a robust figure. It represents about 10% of the average exposure from all sources of radiation, principally natural radiation which itself has a variation that is much greater than 0.3 mSv across the country even when excluding radon exposures (which can be very high and for which remedial actions are recommended). Even if risk factors change, it would seem unlikely that it would be justified to reduce the maximum constraint to below 03 mSv y-' on the basis of its being a small fraction of background radiation. Figure4 Anual risk of death from radiation induced cancer for 02 03, 0.5 and I mSv y -I over a lifetime. 1. - 1 mSv 6 Exposure from age to 100 1.25 - (average for both sexes) 1.00 .20 0 75 - co 0.50 -02 D 0.25 Annual risk 1 in 105 < ------ 0 0 20 40 60 80 100 Age at expression (years) Table 2 Attributes of detriment for public exposure Annual dose (mSv) 0.2 0.3 0.5 1.0 Attributable lifetime probability of cancer death 0.08 0.12 0.2 0.4 Detriment %) 0.12 0.2 0.3 0.6 Attributable annual fatal cancer risk at age 75 2.8 10-1 4.2 10 -1 7.0 10 -5 1.4 10-4 Average annual attributable fatal cancer risk to age 75 5.3 10-6 8.0 10-6 1.3 -5 2.7 10-1 Conclusion First an assertion - people are becoming more risk conscious. Both the workforce and members of the public are able to express their concerns more numerately. They demand better standards of protection and some, admittedly extremists, would like a risk-free society. This zero risk demand, cannot be met by defining radiation or other environmental agents as having zero risk at low levels. This paper has tried to explore where society would today set the levels of unacceptable, tolerable and acceptable risk for both workers and the public. Using these values, the radiation protection 181 ...... 1 9 9 2 standards recommended internationally and in the UK for workers and the public are robust. Fatal cancer risk estimates now seem unlikely to be underestimated and recent evidence from occupationally exposed groups suggests they are not significantly overestimated. However, this is a time of change. Even if the risk per unit dose is thought to be fairly certain, society may seek higher safety standards that would require further restrictions on doses received. Radiation protection philosophy has shown imaginative conceptual thought leading to a system of control that may be better than for any other environmental agent. The future is uncertain, but it may be that other environmental pollutants should be assessed and controlled in the way developed for radiation. If so, a better balance of perceptions will result. References 1. International recommendations for X-ray and radium protection. BJR 7 695-699, 1934. 2. International recommendations on radiological protection. BJR, 24, 46-53, 1951. 3. Recommendations of the ICRP, BJR Suppl 6 1955. 4. Kendall, G. M., Muirhead, C. R., MacGibbon, B. H. et al. Mortality and occupational exposure to radiation. BMJ, 304, 220-225, 1992. 5. The Royal Society. Risk assessment: A Study Group Report. 1983. 6. NRPB. Interim guidance on the implications of recent revisions of risk estimates and the ICRP 1987 Como statement, NRPB-GS9, 1987. 7. Health and Safety Executive. The tolerability of risk from nuclear power stations. HSE, 1988. 8. ICRP. 1990 Recommendations of the International Commission on Radiological Protection, Publication 60. Annals of the ICRP, 21 13, 1991. 9. Barnes, M. The Hinkley Point Public Inquiries. Vol. 4 para. 35-36, 1990. 10. NRPB. Radiological protection objectives for the land disposal of solid radioactive wastes. Does. of the NRPB 33,1992. 182 ...... 9 9 2 Panel Discussion 12.30-13.00 Dr J. Skarka (Co-rapporteur, Session 3) Madam Chairman, ladies and gentlemen. In the first half of this morning's session, we discussed problems of classification of chemicals in the chemical industry and we listened for contributions from this branch. First, Patrick Murphy of the EC presented the general approach to the risk assessment of chemicals. He discussed problems of necessary data and information on chemicals in connection with the hazard identification and hazard effects for human health and for the environment. It was stated that environmental effects for a natural ecology system is a highly complex problem. Concerning conclusions and the most important points, there could be stated the following ideas. First, that the quality of any risk assessment is highly dependent on the quality and quantity of the data available and on the available information. The second idea could be that the safety data requirements or the requested data reflect the kind of use of the potentially hazardous chemicals and depend also on the character of the potential hazards. Point 3 could be for new chemicals. We have usually only basic information relating to possible hazards but there are no specific data on real exposure effects. Exposure has to be modelled using reasonable worst case scenarios and often pessimistic coefficients to be on the sure side in conclusions. Point 4 could be the effects which have no thresholds. Concentrations and figures raise often very different questions for risk assessment and respective risk management. Point 5: for environmental risk assessment, pathway and analysis is often very important. Point 6 absence of exposure data is usually the most serious weakness in carrying out the environmental risk assessment. Another point could be that, compared with the human health effects, our knowledge of environmental effects is often poor. Next the harmonised classification and labelling requests need harmonised and sophisticated criteria and also a greater consensus on what the risks are. In the second contribution we listened to two authors. It was a mutual contribution of Professor Bencko from Czechoslovakia and Professor Ungvary from Hungary on risk assessment practices and approaches in the Central and Eastern European countries. The contributions were focused mainly on vinyl chloride monomer, air pollutants, C02, arsenic compounds and heavy metal problems. Simultaneously results from original research work were included. If we look for the conclusions it's possible to say first of all that the risk assessment procedures in Central and Eastern European countries were not radically different from those used in OECD countries but the risk management procedures were often quite different. A second point could be that over the next few years it is envisaged that an increasing and important adoption of procedures and techniques developed in the OECD countries could be applied in the Central and East European countries. Point 3 monitoring systems are important but they have not been introduced and applied in Central and East European countries in adequate levels. They should be for necessary monitoring, introduced especially in the environmentally critical territories of the countries. Another point is that field monitoring of wild life can be a powerful identification of environmental barm but sometimes proving cause and effects can be difficult. Biological monitoring in people can provide valuable information for both risk assessment purposes and for checking the effectiveness on the risk management. The last pint could be indicated as 183 ...... 1 9 9 2 poor risk management and generally poor attention paid to the environmental programmes in Central and Eastern Europe in previous regimes resulted in worse environmental conditions. Central and Eastern European countries shall need help from the international community to improve this situation. The other two contributions were focused purely and especially to pesticides and one was bringing the views of US and from US side, the other from European or UK situation. We agreed with Dr Norman King from the UK Department of Environment that he will present his comments for this very special professional area. Thank you for your attention. Dr Nornmn King (Co-rapporteur, Session 3) These two papers I think are a very useful perspective, one from the regulator's point of view, the other one from the regulated point of view. I think the thing that surprised me as the rapporteur was how much common ground there was in the papers although they didn't always perceive the problems in exactly the same way. Pesticides are designed to be dangerous to some organisms and are among the most regulated of all chemicals - not surprisingly so. Both speakers dealt with aspects of cancer risk and the USEPA has used a conservative model for risk assessment of possible arcinogenesis but their thinking is evolving to take account of what is known about the mechanism of carcinogenesis now. Given the very great influence which the USEPA's thinking and data requirements have on the pesticides field generally I think we must all look upon this with great interest from now onwards. Dr Thomas also raised carcinogenesis and in particular drew attention to the problems that arise from the use of maximum tolerated dose - that this can distort or complicate risk assessment for carcinogenesis and in worst cases the effect found could be claimed to be a manifestation of the test method itself. National legislation was raised by Dr Hill on the one hand and EC legislation by Dr Thomas on the other. I think the way in which these were raised leads back to a point which Dr Fisk made yesterday about the way in which legislation, which perhaps may be out of date or produced from a particular veiwpoint, can qualify or even remove the ability to conduct risk assessment. Dr Hill touched on the Delaney clause and how that prohibits the risk assessment approach being used in certain areas. Dr Thomas on the other hand pointed to the EC Drinking Water Directive where again risk assessment is virtually ruled out because of the way in which the criteria have been applied. He also hoped that the EC Harmonisation Directive on pesticides wouldn't fall into the same trap. Both speakers agreed that there is a need for international harmonisation. I think they both agreed that harmonisation of data generation and data requirements is a very desirable goal and one that can be achieved albeit with some difficulty and there are moves afoot internationally to do this. Harmonisation of classification and labelling is also an achievable goal it was felt. Harmonisation of risk assessment is certainly desirable but will be much more difficult to achieve. Dr Thomas pointed out that the agrochemicals industry is moving away from simply supplying chemicals and towards the provision of a full plant protection service to its customers, again a point which I think Dr Fisk touched on in his talk yesterday, that people are moving away from simply being suppliers to being servers of their customers. Dr Thomas, I think, made a very valuable point about the problems associated with pesticides that come down not to the way in 184 ...... 9 9 2 which we register or approve them but to the way in which they are used. At the end of the day many of the problems of exposure come down to making sure that the pesticides are used properly. This is a particular problem in developing countries and I welcomed the news that industry recognises this and is working to reduce the problem particularly in those countries. The final point that Dr Thomas made is, I think, a salutary reminder to all of us that we must not throw away the benefits that pesticides can give us simply in an attempt to achieve totally unrealistic levels of risk. Thank you Madam Chairman. Professor Victor Ivanov (speaking on behalf of Professor Anatoly Tsyb, Co-rapporteur, Session 3) Madam Chairman. In our very brief report we would like to point out only the main question. For us as experts in radiation medicine and as a member of the Russian Commission for Radiation Protection (Professor Tsyb is the Head of this commission) it is more interesting to know about the development of risk assessment in this field, in radiation protection, in radiation medicine and we have today a very nice presentation by Dr Clarke. It's very important to stress the problem of risk assessment for low dose range and low dose rate. As you know radiation risk is estimated on the basis of Japanese data from the Radiation Foundation in Hiroshima and Nagasaki for dose range from more than 100 rems in one year and at the same time we have an extremely important problem now in Russia of risk estimation for doses in the range of up to 30 rems. Why only two examples? After the nuclear power accident in 1986 in Chernobyl te Ministry of Public Health of Russia adopted a large-scale programme for the creation of a register of people affected by radiation after the Chernobyl accident; and now we have a data base with 600,000 people with different doses. Half of them, 300,000 are emergency accident workers, clean-up workers, and the mean external dose for this group of people is about 12 rems. Emergency accident workers in 1986, the first year after the accident, received a mean external dose of approximately 20 rems. Mean age of this population is 33 years old. And of course it's very difficult to use now the risk assessment approach to estimate the long-term and short-term effects of radiation action. The same situation for children which received a high level of radiation in Belorussia, approximately 300 rems, and of course it's a big problem to estimate the stochastic cancer risk percentage in the future for this so-called low level of radiation. In his very interesting presentation Dr Clarke discussed today several fundamental radiation protection and risk assessment definitions: acceptable, unacceptable and tolerable level of risk and numerical values to these definitions. It's clear and we agree with Dr Clarke that standards should be based on tolerable risk. We also agree with Dr Clarke that the public tolerability of risk from radiation changes with time and we know this has happened during the last 25 years. We think there is a well-developed technique for estimation of radiation risk which may be very effectively used for other environmental contaminants. Thank you very much. Question and Answer Session Comments kom Dr Ellen Silbergeld (University of Maryland, USA) I'd like to put on a hat if I may as a member of the Carcinogenicity of IOSA Review Panel of the National Toxicology Programme in the United States and make some comments about comments that were made by Drs Hill and Thomas concerning the qualitative hazard identification of carcinogens. There has been a great deal of conversation in the US concerning 185 ...... 1 9 9 2 the use of the maximum tolerated dose and its potential distorting impact on the interpretation of the results of the IOSA and I would refer you to several fairly intensive reviews of the data base on this subject within the NTP test compounds. In fact the use of the MTD does not really overly distort the data base in a qualitative sense. Relatively few chemicals are in fact identified only on the basis of statistically increased incidence of tumours observed at the MTD. In fact in most cases of chemicals identified as carcinogens, including pesticides, by the NTP, there is evidence of carginogenic response at considerably lower doses. There is also considerable care taken in the setting of the MTD. Conversely it is the case that a number of chemicals that exert a considerable amount of target organ toxicity are not carcinogenic so that an easy equation between cell or organ level toxicity and carcinogenicity, even within the confines of the rules of the NTP system, is I think a bit facile. The second point that I would make is that if we move to - and I think everyone applauds this - if we move to a much more mechanism-based approach to either the hazard identification phase, as Dr Hill suggests with his discussions of alpha 2-microglobulin and thyroid carcinogens, or to quantitative risk assessment which is undergoing a great deal of attention in the States, I think we have to be careful of two potential side effects, as it were, in adopting this philosophy. One is, as I tried to suggest earlier, that the data requirements will increase enormously for us to reach even preliminary decisions as to hazard. Not only win we have to demonstrate biologic effects in whatever animal model or test system we choose, we will have to put forward and determine some kind of mechanistic basis for tose effects; this could impede preventive regulation considerably. The other is much more molecularly based and that is that, if we start to look at molecular explanations for what we might think are species-specific responses or of species-specific dose responses in the animals that we use, we have to be very careful to understand the comparative molecular biology in humans. A great deal has been made of the purported absence of alpha 2-microglobulin in humans without recognition that there is in fact a very similar molecule induced in humans which may well perform very similar molecular functions. So we must be very careful, if we're going to wave the flag for molecular biology, and understand that it is complicated, expensive and may in the long run not suit our purposes but if in any event we're going to use it we had better understand it well. Ron Haigh (DGV, European Commission, Luxembourg) Q: I was interested that the faults of the Directive are all the fault of the Commission! I would remind Dr Thomas that in fact it's governments who adopt legislation, not the Commission, we just propose it. It's a small point ... Dr Murphy, who also works for the Commission, might also be sympathetic to that idea. The more substantial point I wanted to raise is that when I made my intervention yesterday I tried to say more clearly in my speech that there was risk management, risk assessment and hazard assessment and I felt that there was some kind of overlap between the three and that it was very difficult to decide where one started and one finished - although this conference is on risk assessment. This morning I began thinking that risk assessment was a scientific part of the process until I heard from Dr Thomas that the risk assessment should be developed so that regulatory control should not impede development - impede development, I presume, of sales of pesticides. And then I heard Dr Clarke referring to tolerability in terms of risk assessment... tolerability set by the judge or the investigator during a certain number of court cases. The rapporteur also talked about tolerability in terms of risk assessment. Under these circumstances therefore, could I ask the question of the panel - is it therefore not reasonable to say that risk 186 ...... 1 9 9 2 assessment is political in its end result and there isn't a kind of scientific component and if you're going to work into tolerability, work into risk management, then we should not be trying to differentiate hazard, risk assessment and risk management. We should have hazard - which I suppose we can all recognise - and then there should be a continuum where the political framework is far greater than at least I imagined when I came in yesterday. Thank you. Dr Barry Thonzas (Schering Agrochemicals Ltd, UK) A: I stand corrected for implying tat te Commission bad anything to do with the Water Directive. It was of course nothing to do with them! I was not trying to imply for one minute that the degree of regulatory control should be tailored, in a sense, to reach an absolute minimum so that we can actually proceed with the development. What I was trying to suggest was that it should be balanced ... that the risk assessment process should not necessarily be excessive for pesticides merely because pesticides in a sense are almost an easy target. Because you have an enormous amount of data for a start and in a sense it sort of lends itself to risk assessment. What I was trying to bring out was that some of the risks that are perceived are not necessarily tied up with the actual risks. And I think that the point you made about risk assessment being - if I understood you correctly - in a sense politically driven is absolutely true. Politics plays a major part, I think, certainly in the regulatory control of pesticides. What I would like to suggest is that we should be honest about this. We should actually make the statement that, for political reasons, pesticides are going to be controlled in such and such a way, but not try to hide political decisions under the mantle of scientific risk assessment. I think this is the point I was trying to get across. Ms Dieta Heding (Science Environmental Journalist) Q: I can see that we are very quickly analysing ourselves to extinction here because nothing will be dangerous, nothing will be deadly, if we follow up the line tat we have already made. So I think that you have to come down to earth and don't beat about the bush. So I have a few questions. What is a free dose of anything ... of radiation ... or chemicals? Who decides ... researches the free doses! And who researches the co-effects, the synergetic effects, of chemicals? And since all the substances in nature are not verified how can anything be safe? And Frank Barnaby, the atomic nuclear professor, said that there is no such thing as a safe dose. An extra dose is a dose too much whether it is radiation or chemicals. So who adds up all these small acceptable doses that we get everyday? Who adds them up? One dose may not be dangerous but if you get 20, 30, 40 microdoses every day in different foods and so on that is not any longer a microdose. So who checks that? And most of you here are men. Do you know that men's fertility has been reduced almost by 50 % since 1930? And do you know why? ... so the whole chemical industry can find out why ... because we know why. So this should really, I think, concern you. Not the older men here but the younger men. Most men are too old to decide, I think, in environmental business. And I do not know which planet you are living on, Dr Barry Thomas, but I know that Carson's book, 'Silent Spring', has come true to a degree which is almost a disaster already. Plants and species and people are dying and we know it. So, tell me, what is a safe dose? Who decides what is a safe dose? 187 ...... 1 9 9 2 Dr Kerstin Niblaeus (Chairman, Session 3) Thank you very much. These are really very important questions. I have a feeling that that might be something to consider at the very last panel discussion on Friday when we are supposed to discuss how well risk assessment serves our needs. Question from John Ballard (Occupational Health Review) to Dr Barry Thomas Q: Ifthechemicalindustrydidi thavetospendsomuchmoneyonriskassessmenthowwould society actually benefit, in particular society in developing countries? Dr Thomas (Schering Agrochemicals Ltd, UK) A: If I understand your question correctly, what you are saying is that if we didn't spend so much money on generating data how would society benefit? Well I think that we would be able to put more money into the research side so that we would be actually developing better pesticides. The state of the art is such at the moment that basically you can try, for example, and design a pesticide such that you can actually control an enzymatic reaction in a plant so that you can be highly specific. So you're actually tailoring a pesticide to deal with a particular problem whether it's in a plant or in a pest; so if you can put more money into that area then I think the end point would actually be, in a very broad sense, a safer pesticide. But the point is, because you actually have to put so much money into development, then the proportional amount of money that you can put into research is obviously lessened. 188 ...... 1 9 9 2 Tuesday, 6 October 1992 Risk assessinent in praetitee 189 9 9 XA04NO320 Food and Drinking Water Safety: Can Risk Assessment Help Us To Get Our Priorities Right? Dr. TV H. B. Denner Chief Scientist (Food), Ministry of Agriculture, Fisheries and Food 17 Smith Square, London SWIP 3JR UK Introduction Huge resources are devoted worldwide by governments and food producers to ensure that food and water are produced with due regard to the safety of consumers. This inevitably involves some form of risk assessment but in the field of food safety a formalised process of risk assessment has been slow to develop. An ad hoc mosaic of approaches has evolved which varies not only between countries but sometimes within countries as well. This may not be unexpected considering the vast variety of kinds of food hazards (Table 1). Table 1. Examples of food-related and drinking water-related hazards Potential hazards iherent in food: Inherent plant toxins Nutritional imbalances Malnutrition Over-nutrition Novel foods Potential hazards related to food production and processing: Food processing contaminants Products of biotechnology Novel food processes Veterinary drug residues Pesticide residues Water treatment agents Processing aids Potential hazards accidentally associated with food: Algal toxins in water Food spoilage microorganisms Diseases transmissible through food and water Marine biotoxins Environmental contaminants Mycotoxins Food contact materials Radioisotopes Potential hazards added to food: Colours Preservatives Flavourings Sweeteners Malicious contamination Not only do food-related hazards themselves vary widely, so do the effects which they can cause. For example microorganisms can cause mild stomach upsets or death within a few hours depending upon the organism involved. For chemical contaminants in food the potential effects are usually less acute although no less serious. Many of the chemicals of concern are believed 190 ...... 9 9 2 to be carcinogens whose effects might only be realised after many years of exposure. Nutritional imbalances may result in an increased risk from diseases, such as heart disease and cancer, which can also arise from other causes. In these latter examples it is often difficult to relate cause to effect even when extensive epidemiological evidence is available. It is important to understand the enormous diversity in possible food-related hazards before describing the assessment of risks associated with them. This great diversity makes it unlikely that any single approach to risk assessment can suit all situations. This means that making comparisons between risks from different hazards is extremely difficult. In fact trying to allocate resources in a logical way between all the different kinds of food-related hazards is a major problem in itself. For with finite resources there is always the danger of finding that focusing on one area of concern results in a potential risk elsewhere being neglected. The aim of this paper is to take a general look at some of the issues facing those with responsibility for controlling food-related risks today and to ask whether a more formalised approach to risk assessment could help to get priorities right. Food safety is not a new subject and it is worth examining history first before going on to look at some current issues and then attempting to draw some more general conclusions. Historical perspective In human history, concern for food safety has always been apparent. In ancient Egypt, we know that the Pharaohs had professional "tasters" who were responsible for primitive "risk assessments" and the notion of poisons was well-established in antiquity. However, the ancients had no way of telling whether the hazard was, for example, the rats themselves or the diseases they carried, or "natural" toxins or poisons added by their enemies. Nevertheless, in many ancient societies there was an awareness of the importance of purity and cleanliness which was reflected in customs and religion. Some religious practices may well have been based upon specific experience of avoiding risky food and water supplies. Prohibition of pork and shellfish may reflect the fact that they were common sources of food poisoning. For example, Trichinellaspiralis carried by pork can form cysts in human muscle and shellfish can accumulate pathogenic microorganisms and toxins from water. The writing of early food law was, however, largely driven by the need to protect consumers from fraud. In England, King John prohibited the adulteration of bread in 1202 and in 1266 a law was passed to protect against short weight and unsafe meat. Hence an erroneous connection may have been made by the public between freedom from adulteration and freedom from risk. Risk assessment of foods The practice of risk assessment, in the context of food and water, is currently conducted in different ways in different countries. The USA generally uses the standard definition of risk - that is the statistical probability of some adverse event occurring - widely in the practice of risk assessment for food. In practice most other countries use an approach which seeks to minimise the risks from food without actually quantifying what these are. Of course in the majority of cases the risks associated with food are so small that it would be virtually impossible to make any reasonable kind of estimate of risk. 191 ...... 1 9 9 2 There is however a similarity in most risk assessments which are applied when considering hazards added to food. Essentially a three stranded approach is employed, although the terminology will vary in different systems (Figure 1). Figure . General scheme for food and water risk assessments Risk Assessment Hazard Evaluation Occurrence Assessment -fIntake Estimation ]--I Consumption Estimation The three essential elements of the risk assessment process are the hazard evaluation, where all of the available toxicological or epidemiological data are considered, the occurrence assessment which seeks to establish the concentration or incidence of the hazard of concern in foods and the consumption estimate which seeks to establish how much of the foods of concern consumers actually eat. The occurrence assessment and consumption estimate are brought together to estimate the intake of the hazardous agent by consumers of the foods in question. Finally the intake estimate and the hazard assessment are brought together to estimate the risk. The risk assessment is then applied in risk management. In some circumstances the hazard evaluation will provide a dose-response relationship so that the estimated intake can be read off against the likely risk. In other cases the hazard evaluation will generate an acceptable level of intake against which the estimate of actual intake can be compared. In the case of food chemicals which are intentionally added the actual level of risk is minimal because the assessment process is based upon a level of exposure at which no adverse effects are expected. It would be a mistake though to say that the level of risk is zero. Although safety factors are included to allow for scientific uncertainty there is always a residual risk which is due to uncertainties which cannot be fully accounted for in the hazard evaluation process. This is not the level of "acceptable risk" since the level of risk is not being assessed. The best we can say is that the level of risk at the ADI is negligible based on our experience of toxicology and food- related risks of this kind. Of course very few consumers ever ave a level of intake which approaches the ADI and this provides an even greater margin of safety. However, presenting information to consumers on this level of risk is very difficult because of their expectation of zero risk. There is a great danger in referring to this level of risk as "zero" because there is always the possibility that new scientific evidence will emerge which will require a change in the regulatory position. Under these circumstances it is very confusing for consumers to be told that something which they were previously told was "safe" is now not safe and this leads to mis-trust in the system. It is therefore necessary to introduce consumers to the idea of minimal risk associated with food. One small group of consumers who are already familiar with managing residual risks from foods are those who experience intolerance reactions. Although many more people believe that they are intolerant of certain foods than can actually be shown in clinical studies, people with genuine 192 ...... 1 9 V 2 food allergies need to take special care over what they eat. Most intolerance reactions seem to be related to natural constituents of foods. However a small proportion can be associated with food additives. These reactions are so idiosyncratic that it is impossible to take account of them in the risk assessment process. Instead, foods are clearly labelled with the ingredients they contain so that consumers with known intolerances can avoid eating them. In addition, in the UK there is a Food Intolerance Databank which provides advice to dietitians in planning suitable diets for their patients. Development of analytical capabilities As our understanding of chemisty, biology and the other sciences has developed over the past centuries, we have been able to establish even more links between cause and disease and have developed ever more sensitive methods for the detection of hazards in food and water. While this is a triumph for science it can present us with serious difficulty vhen we wish to present information on risks to consumers. For example, some of the techniques for analysis of chemical contaminants are now so exquisitely sensitive that scientists hav- difficulty in explaining to the lay public the meaning of their results. Even having convinced consumers of the very small amo. its involved, psychological research has shown that consumers are unwilling to accept any kind of "contagion" or "contamination" of their food by alien substances. Therefore by reducing analytical detection limits ever lower we apparently reveal more and more previously unknown (and therefore uncared about) hazards. The public are not sufficiently knowledgeable about toxicology to understand that very small traces of contamination do not necessarily present a significant hazard. They assume that the merest presence of a contaminant indicates harm. Risk assessment can help to evaluate these hazards and put them into perspective against other every day risks. However it may not be so easy to overcome the psychological factors underlying public perception which dictate that no level of contamination is acceptable. Risk assessment for drinking water Nowhere is this problem of trace contamination more prevalent than in the risk assessment for drinking water. Consumers in the developed world have come to expect high standards of drinking water safety and quality and continue to demand higher standards. Such demands are based in part upon the improving capability for micro-pollutant analysis and also on expanding information about their toxicity. There is also much more public awareness, and in some quarters concern, about water quality issues today. It is, therefore, of great importance that standards for drinking water be set on a scientific and rational basis in order that the public can be given reassurance in clear and unambiguous terms. It is also imperative to have clear standards in order to avoid uncertainty about the significance of the standards and consequent inappropriate expenditure on regulation and control. The current philosophy of many pressure groups is that absolute safety is a desirable and attainable goal for drinking water. One approach to dealing with demands for zero risks associated with drinking water is to set a level of risk which is indistinguishable from background risks. The application of the USA Environmental Protection Agency's Surface Water Treatment Rule to giardia is a good example of this risk assessment approach. Under the Rule water suppliers would need to ensure that the maximum of giardiasis was one case per year per 10,000 of the population. To guarantee this level of risk additional water treatment would be required. 193 ...... 1 9 9 2 Meanwhile in Europe there are concerns about the European Community's Drinking Water Directive regarding standards for pesticides which do not relate to health effects. Other standards in the Directive may also not adequately distinguish between health and aesthetic standards and such ambiguity can lead to unjustified consumer concern. The Directive does not have the transparent basis which is evident in the US EPA approach which I have just described. Without any rational basis to setting standards for drinking water, all standards tend to be regarded as "health" standards and consumers become concerned if any are breached. The concern is greater because, apart from two of te standards, the Directive requires total compliance in every sample taken. This is so for colour, turbidity and iron as it is for other aesthetic parameters. This presents particular problems for water suppliers since in the absence of risk-related standards the supplier is unable to make any judgement as to whether a breach represents a serious risk to health - which justifies immediate provision of an alternative supply - or whether water can continue to be supplied pending the completion of remedial measures. This is an example where there is clear evidence that standards based upon rigorous risk assessment could improve the management of risks. Risk assessment of food chemicals Food chemical products such as food additives and pesticides are presented for risk assessment by their manufacturers and the risk management strategy here is to either approve, restrict or prohibit their use. It is the responsibility of the company making the submission to provide information on toxicology and conditions of use and to demonstrate the safety of the product. Approval for use is granted by the regulatory authorities within established legislative frameworks. In such cases the risk assessment can be comprehensive and any risks carefully balanced against any potential benefits to consumers. In addition to those chemicals which are intentionally added to food many chemicals can arise in food as a result of contamination through the food chain. Substances such as heavy metals and dioxins are ubiquitous environmental contaminants and as a result traces of these chemicals can be detected in many foods. Chemical contaminants in food present slightly different problems to food additives: The presence of trace levels cannot be said to bring any direct benefit to the consumer and so there is limited scope for balancing risks against benefits. Also in many cases no-one can be identified as being directly responsible for, or can be said to benefit from, the presence of these substances in food. For chemical contaminants the most common approach to risk management is to reduce levels of contamination to the lowest levels practicable. Tolerable daily intakes can be set for contaminants which are not genotoxic carcinogens in the same way that acceptable daily intakes are set for additives. However it may not prove practicable to achieve such low levels of intake since the same mechanisms for control are not always available. One factor often called upon when determining levels for chemical contaminants in food is the analytical limit of determination. It is impracticable to set a maximum tolerable concentration of a chemical in food which cannot be monitored. The management of chemical contaminants is often aimed at targeting the main sources contributing to intakes and reducing concentrations to the lowest levels practicable. Of course there is a need for risk assessment and management techniques in deciding what concentrations are the lowest practicable. Often it is valid to estimate what the background level is in the diet and then to consider whether any particular source makes a significant contribution over and 194 ...... 9 9 2 above this. Another possible approach to determining tolerable levels in food is to seek to identify the point at which a small decrease in adverse health effects no longer justifies the additional cost of control. However establishing tolerable concentrations of chemical contaminants in food is likely to be a very delicate process because both the very small health consequences at low levels of contamination and the true costs are extremely difficult to determine. Once against formal procedures of risk assessment and management will need to be developed if this is to be successfully accomplished. In addition to chemicals which are alien in food there are thousands of substances which can arise quite naturally. Chemical compounds which occur naturally in food, however, are rarely given formal risk assessments unless safety concerns arise through food poisoning incidents or are identified by research. One subject which has received close attention is the production of mycotoxins by fungal contaminants in food. Stored food is very prone to attack by fungi although the presence of the organism may not be obvious to the naked eye. Sophisticated chemical analyses are therefore often required to detect the presence of mycotoxins. The problems of mycotoxins have been recognised for many years and many countries have introduced legislation to control concentrations to parts per billion levels. In Asia, sub-tropical Africa and certain other parts of the world, fungal contamination of stored foods is widespread because the climatic conditions encourage fungal growth. If all food contaminated with mycotoxins above parts per billion levels were to be prohibited in these countries then food shortages could be the result. In this example it is clear that a risk assessment carried out under one set of conditions may not be relevant if applied elsewhere. It is interesting to note that there is a "folklore" amongst some consumers that "natural equals good". It seems logical to them to assume that foods which we have been eating for generations should be intrinsically safe. This notion is appealed to over and over again in food advertising and so the idea is continually reinforced in peoples' minds. This belief in the "goodness" of natural things and suspicions about anything altered or synthetic perhaps underlies our heavy emphasis on te regulation of substances added to, or contaminating food. However there are many chemicals of toxicological concern present naturally in our everyday diet. A typical menu of everyday foods can illustrate this (Figure 2 This includes only a few of the natural toxicants which have been identified and there may well be many others present in these foods. One group of compounds of particular interest are the heterocyclic amines. These compounds are produced during high temperature charring of proteinaceous foods and are therefore found at high concentrations in grilled, fried and roast meats. This class of chemicals is related to some of the most carcinogenic known. Many scientists now question the relative priorities assigned to "synthetic" and "natural" chemicals in food and water in our present approach to risk assessments. In the USA, Ames has estimated that 99.99% by weight of pesticides in the diet are "natural" - in that they are produced by plants as natural defences against insect attack - and he adds that only a small proportion of these natural pesticides has received any toxicological evaluation. However the evaluation of all natural food chemicals for toxic hazard by traditional animal-based methods would be an enormous task since there are many thousands of natural plant substances. It is also questioned whether the traditional methods of hazard evaluation would be even relevant since many naturally occurring toxicants occur with protective agents like the anti-oxidant vitamins 195 ...... 9 9 2 which could limit the toxic effects. Toxicological assessment of inherent plant toxins therefore needs to take place in the context of the foods in which they occur. Protective agents in the diet may prove to be a very important factor in the risk assessment of naturally occurring toxicants. For example naturally occurring carcinogens appear to be so common in the diet that one would expect a much higher incidence of cancers of the digestive tract than is actually observed. The nutritional status of individuals may also have a significant impact upon their susceptibility to natural toxicants and risk assessment methodologies will need to take all of these factors into account. Ultimately it will only be through epidemiological studies that the hazards of natural toxicants will be linked with any adverse health effects. One major problem - of assessing the exposure of individual consumers to natural toxicants in their diets - may be overcome by the use of biomarkers wich provide an indication of the level of a marker compound in a bodily tissue or fluid. The risk assessment of inherent toxicants is therefore an area where there is vast scope for development of new testing methodologies and management strategies. Only through the application of appropriate technologies can we hope to make sense of the many potential toxicants which occur naturally in the diet. One interesting approach to food chemical risk assessment which Ames has proposed is the Human Exposure/Rodent Potency (HERP) index for assessing cancer risks. This is a measure of the ratio of the estimated human exposure to a toxin to the results of toxicological studies on rodents (Figure 3. Figure 3 Calculation of the HERP index HERP = Human exposure index Rat Potency index The higher the HERP value, the greater the risk. Ames regards the index as a means for rationally setting priorities although the HERP has been the subject of criticism amongst toxicologists. Nevertheless, the approach provides some interesting comparisons between some natural and synthetic cancer risks (Table 2. 196 M I I 0 - I ...... 9 9 2 Table 2 HERP Index for Possible Carcinogenic Potency Average or Reasonable Carcinogen and Dose Daily Human Exposure per 150-Pound Person HERP Substance Amount Carcinogen Dose 4.7 Wine 8 oz Alcohol 30 ml 2.8 Beer 12 oz Alcohol 18 ml 0.1 Basil I g dry leaf Estragol 8.8 mg 0.1 Mushroom 0.5 oz Hydrazines 0.06 Saccharin One diet cola Saccharin 95 mg 0.03 Comfrey tea I cup Symphytine 38 ug 0.03 Peanut butter One sandwich Aflatoxin 2 ppb 0.003 Baconcooked 100 grammes Nitrosamine 0.04 ug 0.001 Chlorinated tap water I litre Chloroform 8 ug 0.0004 Grain products Daily intake EDB 0.42 ug 0.0002 PCBs Daily intake OCBs 0.2 ug EDB = Ethylene dibromide PCB = polychlorinated biphenyl (From "Food Safety" by J. M. Jones, Eagan Press, 1992.) Microbiological risk assessment In spite of the importance of microbiological risks associated with food, formalised risk assessment has not been applied to decision-making in tis field. This is largely attributable to the unique problems posed by te assessment of the risk of microbiological disease. Estimating risks from microbiological azards is complicated by the wide variability of the virulence of strains of organisms, the foods in which contamination can occur and the susceptibility of the populus. It is therefore very difficult to make any numerical estimate of the degree of risk. The degree of hazard may also change dramatically as a food passes through the food chain so that a satisfactory assessment at one point in the chain is only meaningful if appropriate measures to maintain microbiological quality are taken at all stages later in the chain. Many experts rate the risks from microbiological hazards in foods to be just as important as those from chemical hazards, but allocation of regulatory resources is only just beginning to reflect this. Radioactivity in food The risk assessment of radioactive contamination of food, by contrast, is a relatively well developed science. It occupies a unique place in our considerations, because te risks to the consumer of a wide variety of natural and man-made contaminants can be related in terms of the single concept of radiation dose. Radioactivity also has a high profile in public and media concern, much higher than it merits in terms of any objective assessment of risks. Although the assessment of risk is the subject of a wide degree of international agreement there remain small areas of disagreement within the scientific community. These disagreements, however, are in matters of detail that are minute compared to the uncertainties regarding many other forms of contamination. 197 ...... 1 9 9 2 This is one area where risk assessment is used in a direct, quantitative way in the control of contamination of food and drink. Because exposure to radiation can be expressed in a single dose and the relation between dose and effect is well established, it is possible to make probabilistic estimates of risk which are reflected in the regulatory system used in the United Kingdom to control man-made radioactivity in the environment and the food-chain. Both the control of radioactive discharges by nuclear sites and the limits on contamination following a nuclear accident are based on the concept of acceptable or tolerable radiation dose, which is in turn derived from the risk per unit dose calculated by the International Commission for Radiological Protection and interpreted for us in the UK by the National Radiological Protection Board. As with other forms of contamination, the debate on what level of risk is acceptable or tolerable is a matter of widespread concern, but it can at least be based on a reasonably solid scientific foundation. Risk assessment in developed and developing countries In developing countries, microbiological hazards are often of greater importance than other kinds of food-related hazards. A supply of drinking water which is free from pathogenic micro- organisms is often jeopardised by natural conditions and there may be arguments for accepting risks associated with the chemicals used to treat water in order to control acute microbiological hazards. For example, a decision made on the best of intentions by Peruvian officials not to chlorinate much of the country's drinking water seems to have led to a cholera epidemic of major proportions. The decision appears to have been based upon studies by the US EPA which showed that the chlorination of organic matter in water may create a cancer risk from the formation of trihalomethanes. Studies had shown that, at 100 ppb levels, these compounds could pose a cancer risk of in 10,000 lifetimes. It would be very difficult to justify the introduction of alternative treatment methods in the USA to reduce such a small risk since it is doubtful that the benefits could in any way outweigh the immense costs involved. However in the South American situation - where there was no possibility of replacing chlorination with other forms of treatment - the risks from cholera were far in excess of the risks from cancer due to chlorination. A risk assessment of the removal of chlorine from South American water supplies would probably have resulted in a very different decision being made. Managing uncertainty in food risk assessments Risk assessment and risk management of foods and water are often complicated because there are a large number of factors contributing to variability in the human population (Table 3. Table 3 Factors contributing to variability in the human population Genetic diversity of man Age-related differences in response Medical conditions Regional and ethnic variations in diet Overall nutritional status. In addition to these factors is the wide variety in traditional diets and in what individuals choose to eat from the foods available to them. 198 ...... 1 9 9 2 Toxicological testing procedures also introduce uncertainties into the risk assessment process. In particular the interpretation of animal studies and the extrapolation of animal data to man are great sources of uncertainty. In order to allow for these uncertainties two approaches are adopted: Firstly only data from the most susceptible animal species are considered; secondly safety factors are introduced to allow for the uncertainty. In some systems further conservative measures are used especially when quantitative risk assessment (QRA) computer models are used to extrapolate from high doses in animals to low doses in humans. It is impossible to select a model which accurately represents the underlying biological cancer induction mechanism while this mechanism remains unknown. Therefore many systems adopt a "worst case model" approach. Furthermore, in some systems the risk estimate is based on te upper confidence limit of the dose-response curve selected. Where the quality of data is poor this can introduce a very large element of systematic error. Overall, whilst QRA techniques appear to offer great potential for improving risk assessments their validity seems limited until they can be sown to reliably model the underlying biological mechanism. A conservative approach is also adopted when making estimates of the likely intakes of consumers. Here worst case estimates are often employed to estimate the occurrence of a hazardous agent in the food supply when the true occurrence is not known. Also food consumption estimates are designed to ensure that the diet of even the most extreme consumer is considered. It is furthermore assumed that high level consumers will maintain this level of consumption throughout their lifetime. The problem with such a conservative approach is that the margins of error which are quite properly introduced to allow for uncertainties in the data are compounded at each stage in the risk assessment process. The final effect can be to produce a margin of safety which far exceeds the realms of possibility. The reason for this is that although each correction for uncertainty is valid, it is most unlikely that unpredicted events associated with all of the sources of uncertainty would arise in any one incident. This phenomenon has been called "creeping conservatism" and its effect is to over-estimate the true degree of risk associated with food hazards. This in turn leads to sub-optimal risk management decisions so that the cost of control far outweighs any benefit that might be gained from the action. This is an area of food risk assessment where considerable further development is needed. Nevertheless it is important not to under-estimate the need for a conservative approach and any strategy adopted for dealing with uncertainty in a rational way must make adequate allowance. For example the wide variety in perfectly normal dietary patterns means that it is not valid to estimate the average risk on a population basis from exposure to a hazard through food. It is not sufficient just to estimate societal risks - the risks to actual consumers must be assessed. Public perceptions of food-related risks While regulatory authorities go to great lengths to accurately estimate the risks to consumers the public's perceptions of the relative risks associated with food and water have often been found to differ markedly from those of the experts on food safety. This is not a unique phenomenon but is of particular importance in food safety because people are naturally very sensitive about an issue as intimate as what they put in their mouths to eat. It is no use saying that consumers are irrational about the decisions they make and therefore need educating because consumers ultimately have the right to demand food of the nature and quality they expect. Instead it 199 ...... 1 9 9 2 is necessary to begin to understand the factors which consumers take into account when making decisions about food. Once these factors are understood then it will be possible to take them into account when framing information for consumers and even in some cases in the risk assessment process itself. As part of the risk management process it is necessary to allocate resources rationally between the many competing demands for risk assessment for the various food hazards listed in Table . There is also a need to understand the systems which governments currently adopt to do this. The balance seems to be about right most of the time but it is apparent that the allocation is not strictly based upon the scientific data alone. For example regulatory authorities will need to consider whether they can take any action about a problem. It is far easier to control pesticide residues through the centralised process of product approvals than it is to improve food hygiene in millions of domestic kitchens. But what is also important is that the public feel that they are able to control hazards which might occur in their own homes. In contrast they rely upon governments to control hidden hazards such as pesticides. The public is also concerned about hazards which affect vulnerable groups like children and are concerned that the costs of a particular hazard might not be borne by those who reap the benefit. These kinds of factors, in addition to scientific judgements about the likely risks have driven our priorities in the past and there is a need to develop more formal ways in future. The scientific process of risk assessment should be undertaken within a social context if it is to be successfully accomplished. Food irradiation An example of the public's perception of risk sometimes differing from that of the scientific community is that of the risk associated with food irradiation. Over the last 40 years food irradiation has been studied more than any other food preservation process. The risk has been assessed by a number of international organisations such as the WHO and FAO, including consideration by an independent UK Safety Committee (the Advisory Committee on Irradiated and Novel Foods). Yet despite reassurances that the process is without identifiable risk consumers are still reluctant to accept irradiated food and the potential benefits of the process. Regulations have recently been introduced in the UK to allow the controlled use of the process under a strict licensing system. To date only a single licence has been issued for a range of herbs and spices. This to some extent reflects the hesitancy of food retailers and their perception of the public's attitude to irradiated food. Nevertheless, other applications may follow and food irradiation will no doubt find its niche amongst the armoury of preservation methods developed over the centuries to enhance food safety. In this case the Government's decision was to permit the use of food irradiation provided that the final product was appropriately labelled. This gives the final decision on whether to consume irradiated food or not to the consumer. In other circumstances this would not be so easy. For example there are frequently calls to label foods with the pesticides used in their production. This would be an impossible task when you consider imported foods which may have been mixed and blended and passed through several suppliers and the problem of composite foods is even more daunting. Consumers are however able to select "organically produced" food and thus make a choice about the use of agricultural chemicals in that way. Even though consumers express high levels of concern about pesticides in their food the level of take-up of organically 200 ...... 1 9 9 2 grown foods is surprisingly small. This may be due to the high prices charged for organic food but is probably also because consumers are against the principle of the use of agro-chemicals and their potential environmental effects rather than being concerned about specific adverse health effects. In this case no amount of revision of the risk assessment process would make the use of pesticides on food crops more acceptable to consumers. Biotechnology In the UK guidelines have been published on labelling requirements for products of biotechnology. The US Food and Drug Administration, on the other hand, has recently decided to treat foods created through plant biotechnology just like any other foods derived from plant breeding. The merit of labelling is that it enables the consumer to make an informed choice where there has been material alteration to the final food. However, this should not be used as a primary education tool. In order to avoid conveying over-simplistic and potentially misleading messages a wider public information campaign would be required. For this to be effective there is a need to understand better the factors which underlie consumers' concerns about the products of biotechnology. Nutritional hazards In many parts of the world food shortages and nutrient deficiency diseases are still major problems. In such regions sophisticated risk assessments are not required to show that what is needed is simply more or better food. In western countries the supply of food is not a problem and here the main diet-related diseases are due to over-nutrition and the inappropriate choice of food. It has been found that many people seem not to be willing to change their diets in the light of information provided to help consumers adopt a more healthy diet. It is thus apparent that we need to provide better quality information about the major food-related risks to consumers if we are to have any significant impact on dietary change and improvements in health. Risk assessment can help in the development of policies on the issues and in presenting information to consumers in a clear and consistent manner. Conclusions Given the enormous diversity of food and drinking water-related hazards no single approach can address all of the problems encountered and this means that it will probably always be difficult to directly compare the results of one risk assessment with another. This in turn makes it very difficult to prioritise risks and thus allocate resources in a rational way between them. We will probably therefore always have to rely upon expert assessment of the available information in order to judge between different risks and get our priorities right. This does not mean, however, that there is no role for formalised risk assessment and no need for improvements in risk assessment methodologies. There are many examples where risk assessment can help us to identify and evaluate the factors involved in food-related risks. Tere is also vast scope for improving commonality between different procedures. For example, a common language of risk is vital if we are to be able to make any judgement about the results of different risk assessments. There is also a place for a common philosophy - for instance the expression "as low as reasonably practicable" has been well defined in the context of occupational health and this approach might find application in food risk assessments. Furthermore it should be possible to develop consistent standards so that the meanings of 201 ...... 1 9 9 2 acceptable and tolerable risk when associated with food are well defined. This does not necessarily mean that they will always reflect the same probability of harm since what is acceptable in one set of circumstances may not be acceptable in another. What it should mean is that a common strategy has been applied to determine the acceptable risk in any particular context. In conclusion it should be said that risk assessment is a management tool and its value should therefore be assessed in the terms of the results that are achieved with it. Food safety should always command a high priority above other considerations but there will always come a point at which the additional benefit to be achieved from providing an extra degree of freedom from risk is far outweighed by the sacrifices required to achieve it. We must always aim for the optimum and carry public opinion with us. Risk assessment can help us to achieve this. If risk assessment procedures take account of the context in which they are to be applied, if they use clear and unambiguous language and if they adopt a conservative approach without exceeding the limits of probability then they will provide a valuable tool for the management of food-related hazards and ultimately help us to get our priorities right. It is important that those carrying out risk assessments of food hazards should make them accurate and consistent and use methodologies which are open to public scrutiny so that the public can understand the decisions taken. In turn consumer lobby groups must be sensible and rational in their demands for safety standards. Otherwise we might as well adopt a strategy where we ban everything which looks remotely hazardous, aim for the impossibly absurd standard of absolute safety and starve as a result! 202 ...... 1 9 9 2 Figure2 Naturally occurring toxicants in food DINNER MENU Fresh Orange Juice Methanol Mushroom Soup Agaritine and other mushroom hydrazines Seafood Cocktail ParalyticshelUish poisoning Diarrhreticshel lsh poisoning Roast Beef Heterocyclic amines or Devilled (Pig) Kidneys Ochratoxin A, heterocyclic amines Horseradish sauce or mustard Glucosinolates Broccoli Glucosinolates, allyl isothiocyanategoitrin Broad Beans Vici.ne, convicine, leclins Potatoes Glycoalkaloids,phytoestrogens or a vegetarian alternative Vegetarian Lasagne Deoxynivenol,fumontsins ohratoxin A tomatine, gycoalkaloids, apsacins, safrole, estragole, aflatoxin MI Lemon Sorbet Psoralens Plums in Brandy Ethyl carbamate, yanogenetic glycosides Pistachio Nut Ice-cream Aflatoxins Bread Rolls Ethyl carbarnate Red Wine Alcohol, ethyl carbamate, tyramine Coffee Caffeine Herbal (Comfrey) Tea Pyrrolizidinealkaloids 203 9 XA04NO321 Microbiological Risk Assessment and Public Health Dr. Roger Skinner Department of Health Introduction Despite the advances made in risk assessment in the past twenty years, in areas as diverse as toxicology and offshore engineering, the risk assessment approach has made little impact on those addressing the microbiological aspects of public health. In this paper the advances which have been made are discussed and the difficulties preventing the wider application of microbiological risk assessment (MRA) to public health are considered. The term microbiological risk is used here to mean the probability of contracting a disease caused by a microorganism. I intend to demonstrate that the dynamic nature of microorganisms and the unique nature of the relationship between a pathogen (a microorganism which causes disease) and its host create special challenges for those involved in MRA. Although these problems are difficult they are not intractable. Indeed in some cases partial solutions have already been found and applied. It is hoped that this paper will help stimulate further thought and consideration in a variety of disciplines so that these challenges can be met, thereby allowing MRA to fulfil its potential. The Complexity of Pathogenic Microorganisms There is an enormous range of pathogenic microorganisms. However all pathogenic microorganisms have certain things in common. They are invisible to the naked eye, they can reproduce themselves, they can change, they can die and they are biologically adapted to infect a host or range of hosts. A lay-person's description of microorganisms is beyond te scope of this paper, but anyone keen to know more may wish to consult a book such as Microbes and Man (Postgate, 1992). Some of the complexities peculiar to MRA can be illustrated by considering the possible fate of a salmonella bacterium in a sandwich. If the sandwich is kept at room temperature the bacterium multiplies. Being at room temperature (rather than body temperature) is a signal to the bacteria that they have not been eaten. Consequently they cover themselves with a type of "hair" or fimbriae which can bind very effectively with some of the sugars that line the oral cavity. When the sandwich is eaten, the bacteria it contains are ready to colonise the mouth. In response to the rise in temperature and the drop in pH (the mouth has a pH of about 5.5) the bacteria switch on special mechanisms which allow them to tolerate the very low pH of the stomach. Thus prepared the bacteria are able to pass through the stomach and enter the intestine. Once inside the intestine the salmonella bacteria compete with the indigenous flora. If the host has recently encountered this strain of bacterium, plentiful and specific antibodies in the intestinal mucus may prevent the bacteria from attaching to, and penetrating, the cells which line the intestine. This process describes some of the complexities that we know about. Even though enteric pathogens of this type are comparatively well characterized, there is still much to learn about how salmonella organisms initiate an infection. 204 ...... 9 9 2 We know very much less about most other pathogenic microorganisms. There is also a large range in host response which is imperfectly understood. The interaction of host and pathogen is variable and unpredictable and consequently poses a challenge to the development of reliable and valid MRA. Successes and Challenges in Microbiological Risk Assessment Despite the complexity of microorganisms and ost pathogen interactions, a small group of microbiologists in the USA have developed a methodology for assessing the risk of waterborne disease (Regli et aL, 1991). The keystone of this work is a number of dose response relationships, ingeniously deduced from studies with human volunteers. These have enabled researchers to calculate the probability of infection from a single organism. In order to infer what level of treatment is required to ensure that a given water source provides drinking water of an acceptable, quality, these dose response relationships have been combined with assumptions about the distribution of te pathogen of interest in water, the amount of water consumed and the acceptable level of waterborne transmission of the organism (Rose et aL, 1991). This approach has had a degree of success and te US Environmental Protection Agency (EPA) has used it to help formulate their Surface Water Treatment Rule. It is currently being used in te development of forthcoming EPA regulations covering groundwater disinfection and disinfection by-products. It would be wrong to assume that because MRA has been applied to one area, the data and methodology for widespread and routine MRA are within easy grasp. Those applying the MRA to drinking water have certain advantages. For example, most people drink some water and water is a relatively homogenous medium. Consequently, reasonably valid assumptions about consumption and distribution can be made quite easily. However, the principal advantage in undertaking a MRA with water is the availability of results of human infectious dose trials (experiments in which volunteers consume different quantities of a pathogen) for a number of the relevant micro-organisms. Human trials, the technical limitations of which are discussed below, were only conceivable because the organisms of interest were not believed to cause serious illness. The availability of these kind of quantitative data meant that a quantitative risk assessment was possible. These forms of data will not be readily available for many future assessments of microbiological risk. This is partly because such trials are inconceivable with any pathogen which may cause serious adverse effects and partly because of an increasing reluctance to conduct such experiments under any circumstances. The difficulty of obtaining dose response data may mean that many future MRAs are more likely to be conducted on a qualitative basis. In spite of the advantages of conducting MRA in water the waterborne models do have certain weaknesses. The limitations of the waterborne models include the uncertainties surrounding the assumptions about the applicability of the data, the consumption of water and the distribution of pathogens in the water. These problems are common to most considerations of MRA and will be dealt with below as part of a broader consideration of issues awaiting resolution. Issues that have yet to be resolved in microbiological risk assessment To consider and compare the different issues that will need to be resolved in MRA and related areas it is useful to break down the process into a number of possible steps: hazard identification, risk characterization, risk estimation, exposure and option evaluation. Issues which cover the entire process are also discussed. The use of these steps is meant to facilitate discussion rather than to prescribe a particular sequence for future MRAs. 205 ...... 1 9 9 2 Hazard identification Hazard identification in this context means the identification of a microorganism associated with a particular disease. Since 1876, when Koch found that a microbe (subsequently named Bacillus anthracis)caused anthrax, a wide range of diseases has been found to be caused by microorganisms. It seems likely that many more discoveries have yet to be made. For example, only about half of the cases of presumed infectious diarrhoea are of known aetiology and it seems likely that many more pathogens await discovery. Rotaviruses and a type of bacterium called Campylobacter,both of which were first isolated from humans in the 1970s, may account for as much as a quarter of all gastrointestinal disease and probably have done so for a long time. Epidemiology, discoveries in the veterinary world and painstaking laboratory work have all helped us discover new microbial causes of disease. In the future the pace of discovery may quicken as novel molecular techniques are applied to this field. These new techniques allow the detection of the nucleic acid of pathogens that are undetectable by traditional methods. This may mean that the discovery of microorganisms will be far less reliant on growth of the organism in the laboratory, a method pioneered by Koch. Risk characterization There are special considerations deciding how a microbial risk should be expressed (risk characterization). Risk managers and assessors must decide which risk to consider as well as deciding how a risk should be expressed (for an individual, a population and so on). Even if a person is exposed to a pathogen they may not become infected, if they do become infected they may not become ill, and if they do become ill they may not die. The choice of endpoint (exposure, infection, disease or death) can have very important consequences. For example, since infection usually occurs more frequently than disease, a risk assessment based on the former will be more conservative (that is, give a higher estimate of risk) than one based on the latter. Risk assessors and managers must consider which is best for their purposes. In this context it is important to emphasize the difference between infection and disease. One may become infected, rid oneself of a microorganism and acquire immunity without ever being aware of one's encounter with a microbe. This is important since much dose response data from man or animals may measure infection rather than disease and an assessment based only on infection may cause us to use resources reducing a risk which has no consequences for our health and well being. However, in certain circumstances a person who is infected (whether they are ill or not) may spread disease to others. This is called secondary spread and is an important factor to consider in controlling infectious disease. The choice of endpoint is to some extent a risk management decision; however the decision may be guided by a knowledge of the chances of an infected person becoming ill or infecting another person and the chances of an ill person dying. In certain circumstances this sort of information may be gleaned from existing public health records and previous epidemiological studies. Risk estimation Estimating the risk is undoubtedly the most problematical area in MRA. Methods available include epidemiological studies, where the risks are calculated from studying disease in the population, and dose response studies, where the risks are estimated from trials with animals or human volunteers. 206 ...... 1 9 9 2 Epidemiological studies. Many microbial diseases are relatively common and epidemiology provides an opportunity to gather data about the likelihood of disease from real population based events. Factors such as immunity, host variability and pathogen mutability are automatically taken into account; this can remove at least some of the uncertainty inherent in making assessments of risks. Nevertheless, other factors must be considered when using epidemiological studies. The ability to estimate a risk depends on the frequency of occurrence of the adverse event (that is the risk) and the ease with which the event can be identified. For example, the risks of botulism are well known because although the disease is very uncommon it rarely goes undiagnosed. However, some diseases would require large studies at a cost out of all proportion to the problem in hand. It should also be remembered that most epidemiological studies will not give the sort of dose response data required for a quantitative model of the kind used by the EPA to formulate drinking water regulations. Much can be done to overcome these difficulties. There are good mechanisms in the UK for disease surveillance. These mechanisms may be used or adapted to obtain a considerable amount of information about microbiological risks. Also a number of studies being planned by the Department of Health and others should give valuable data about microbiological risks relating to food safety. Possible future developments include the combination of novel immunological tests with sophisticated methods of data analysis which may make large and sophisticated surveys possible at a lower cost. It is also possible that, in the future, a more exhaustive and detailed investigation of outbreaks may give us better dose response data. Experimental data. These are the kind of data required by the current quantitative model used in water MRA. Whether the experiments are conducted on animals or humans the data can be interpreted in one of two ways: either as a dose response relationship, in which it is assumed that just one organism can initiate an infection and the probability of acquiring an infection rises with the number of organisms to which one is exposed; or, in terms of a minimum infective dose (MID) in which it is assumed that a certain minimum number of organisms (usually more than one) is required to initiate an infection. Given that, by definition, a pathogen must be able to reproduce itself in its host there would seem to be no reason why only one organism cannot initiate an infection. The notion of a MID arose from the use of limited numbers of animals or volunteers in early infectivity studies. As the number of organisms to which the subjects were exposed was decreased the probability of infection declined. When the number of organisms used was smaller than the number required to have a reasonable chance of infecting at least one subject it was concluded that the subjects had not been given "an infective dose". However, it is quite plausible that if the number of subjects were incr eased one might find that a lower number of pathogens would initiate an infection in at least one of the subjects. The dose response curve that usually gives the best fit to experimental data is called the beta distribution, a distribution which has as its inherent assumption that one microorganism can cause disease (Haas, 1983). In interpreting experimental data one must also be aware that the experimental exposure to the microorganism may not accurately reflect what happens under natural circumstances. For example, protozoa or bacteria are capable of varying significantly in response to their environment. 207 ...... 1 9 9 2 Human experimental data are available for about 13 enteric pathogens (Rose and Gerba, 1991). It is these data which have formed the basis of the assessments of the risk of water-borne disease referred to above. It is unlikely, for the reasons already discussed, that new trials will be conducted to expand upon this database. Animal experimental data may also be used in MRA. This form of data can probably only give an indication of the risk of infection rather than the risk of disease. The validity of animal data depends on the extent to which the infection in the animal model reflects infection in humans. Consequently, if animal models are to be used, it will be important to discover as much as possible about the limitations inherent in the models. Recent advances in our understanding of the mechanisms underlying pathogenicity may help decision making and aid in formulating safety margins. However, much work is still required in this area. Exposure Data on how frequently an individual or a population is exposed to a risk may seem technically easy to obtain. However exposure is often intimately related to human behaviour and consequently very variable. This variability may be accounted for by using "worst case" assumptions (for example, that we all consume 3 litres of water or two raw eggs a day) but this may lead to an overestimate of the risk. More accurate and detailed exposure information will reflect the variation in human behaviour. Such data can be obtained by using the techniques developed by market researchers. A lot of information is already available for some forms of exposure; for example, the Dietary and Nutritional Survey of British Adults (Gregory et aL, 1990) gives a breakdown of food consumption by age, sex and social class. The numbers and distribution of the pathogen may also be very variable as they are affected by growth and death, which in turn are dictated by the organism's environment. Option evaluation Once an assessment of a risk has been made a decision must be made about what, if anything, must be done to reduce the risk to an acceptable level. The acceptability of a risk depends on how the risk is perceived. Although quite a lot of resarch has been published on risk perception in general, the perception of microbiological risk has not yet been adequately addressed. It is possible that the "one off" and immediate nature of some microbial infections may mean that they cause less anxiety than cumulative chemical effects. Nevertheless, the general principles of risk perception most probably apply. For example, exotic microbial risks often get more attention and resources than more familiar microbiological problems. If a number of options are available, or two conflicting risks have to be reconciled, then some form of comparison must be made. The risks may be compared with other risks, the benefits or cost. The difficulties of making such comparisons in a reliable and equitable manner are not peculiar to MRA. However such comparisons can give some interesting insights. For example, the eradication of smallpox, a triumph of public health, was an extremely effective use of resources. The entire campaign cost approximately 300 million US dollars, yet the eradication of smallpox is thought to save billion US dollars each year. Not all cost-benefit analyses are so clear cut and the moral and practical difficulties of attributing costs in a process as complex as a microbiological disease are formidable. 208 ...... 1 9 9 2 Issues relating to the process overall Much of what has been considered so far is dauntingly complex. There is therefore a need to develop procedures to help us accommodate the whole process and to validate the results to find out if they are at least approximately correct. Computer based decision making systems (expert systems) should help in the handling of complex decision making processes. Expert systems have found many uses in other areas of complex decision making such as engineering design, but it will not be possible to employ them in MRA until appropriate procedures have been developed. Whatever procedure is used the result of the MRA must not be uncritically accepted. As noted above, many of the events to which MRA might be applied occur quite frequently. This can be exploited by, wherever possible, validating the risk assessment by comparing it with reality. The assessment can then be refined, possibly as part of an iterative process. Conclusions From the above it should be clear that a lack of both information and a detailed conceptual framework are factors inhibiting the widespread use f MRA. The Department of Health has therefore asked the Government's Advisory Committee on Dangerous Pathogens to address the issue. It is hoped that the Committee's findings will form a framework within which data may be collected, processed and implemented with consequent widespread benefit both nationally and internationally. References Gregory, J., Foster, K., Tyler, K. and Wiseman, M., 1990). The dietary and nutritionalsurvey of British adults. HMSO, London UK. Haas, C. N., 1983). Estimation of risk due to low doses of microorganisms: a comparison of alternative methodologies, American Journal of Epidemiology, 118, 573-582. Postgate, J. R. 1992). Microbes and Man, Cambridge University Press, Cambridge, UK. Regli, S., Rose, B., Haas, C. N. and Gerba, C. P. 1991). Modelling the risk from Giardia and viruses in drinking water,Journal of The American Water orks Association, 83, 76-84. Rose, J. B. and Gerba, C. P. 1991). Use of risk assessment for development of microbial standards, Water Science and Technology, 24, 29-34. Rose, J. B., Haas, C. N. and Regli, S. 1991). Risk assessment and control of waterborne giardiasis. American Journal of Public Health, 81, 709-713. 209 ...... 1 9 9 XA04NO322 Chemical Safety of Food and Drinking Water M. Younes and C A. van der HeiJden WHO European Centre for Environment and Health Bilthoven, The Netherlands Introduction Food and drinking water are major sources of human exposure to a large number of chemicals added intentionally for technological reasons or present unintentionally due to contamination. On the other hand, there is a public demand for an essentially risk-free supply of food and drinking water. The concern over the presence of chemicals in the human diet received further emphasis through the development of toxicological and analytical methodology with increased sensitivity over the years. In order to minimize the potential health hazards to the consumers, standards have been established which indicate levels of consumption that are - according to scientific evidence - considered safe and which, consequently, permit control measures to be taken. In this context, public perception of a particular risk, may not always be in line with what might be considered a "real" risk. Thus, while in the public opinion risk associated with smoking or over-nutrition might be accepted or underestimated, certain food chemical related risks may not be accepted and are sometimes perceived as alarmingly high. Nature of chemicals present in food Chemicals present in food can be broadly divided into two categories: those added intentionally to food, i.e. food additives, and those present unintentionally, i.e. contaminants. In fact, the use of direct food additives predates recorded history, when methods for preservation of food abundantly available at harvest time were developed to ensure adequate food supply during nonagriculturally productive period, or to preserve game collected at peak hunting periods. Examples of early methods of food preservation include the use of salt and potassium nitrate (saltpeter). The direct food additives used at present include: (I)processingaids, e.g. anticaking agents, which are intended to aid in the processing of foods during production and after purchase; 2) texturing agents, e.g. thickeners, which provide specific food items with a desirable consistency and texture; 3 preservatives, e.g. antioxidants and antibacterials, which prevent degradation of foods during processing and storage; (4)flavouringand appearanceagents, e.g. flavour enhancers, flavouring agents and surface-finishing agents, which either enhance existing flavours or add a flavour to foods, and which are capable of improving their appearance; (5) artificial sweeteners; 6) nutritional supplements, e.g. vitamins and trace minerals, which are added to replace essential nutrients lost during processing or to supplement existing levels; and (7food colours, both of natural and synthetic nature. Substances that are not natural constituents of food and have not been added to food for a technological reason are considered to be contaminants. They may become constituents of food items during production, processing, packaging or storage. Major food contaminants include residues of animal feed additives and veterinary drugs, mainly antibiotics and growth-promoting agents, substances migrating from packaging material, natural toxins of microbial origin or those that are produced by moulds, and environmental chemicals like pesticides (insecticides, fungicides, herbicides, etc.), polychlorinated biphenyls, dibenzodioxins and dibenzofurans, heavy metals and radioactive compounds. 210 ...... 1 9 9 2 Nature of chemicals present in drinking water Antimicrobial substances are the major agents added intentionally to drinking water with the aim of improving its bacteriological quality. In this respect, chlorine is the most widely used disinfectant because of its ready availability and cheapness, as well as the ease with which it can be measured in water. When added to water, chlorine may react with reducing agents, e.g. ferrous ions or hydrogen sulfite, to yield chloride ions, or with ammonia, amines and other organic compounds, resulting in the formation of a variety of chlorinated compounds including chlorinated hydrocarbons and a large variety of high-molecular weight chlorination products of, e.g., humic acids. Besides, residual chlorine will usually be present in drinking water after treatment, as it is generally added in excess to ensure its antimicrobial action. Another agent, which is added intentionally to drinking water in some countries, is fluoride, because of its known protective action against caries. The presence of chemicals in drinking water is often a consequence of ground water contamination. Major sources of such contamination include disposal of industrial wastes and industrial impoundments, solid waste disposal sites and agricultural activity. Major contaminants include organic compounds, like chlorinated ydrocarbons and pesticides, as well as inorganic constituents, like heavy metals and anions (e.g. nitrate). Finally, drinking water may contain some radioactive constituents. Safety evaluation Safety evaluation involves the phases of hazard identification, as well as risk estimation and evaluation. The first stage is hazard identification, where hazard is an inherent property (or a set of inherent properties) of a chemical substance or mixture enabling it to cause adverse effects when a particular level of exposure is reached. A second stage is the estimation of risk, where risk is the predicted or actual frequency of occurrence of an adverse effect from a given exposure. In fact, the terms hazard and risk have been used incorrectly as synonyms. To put it in a simple way, hazard is the potential to harm, while risk is the chance that harm will occur. The determination of this probability of a harm to occur and the subsequent consequences forms the basis of safety assessment. The aim of evaluating risk is to determine a level at which risk is minimized to an acceptable level. At an international level, WHO plays a leading role in establishing guideline values for the intake of chemicals present in food and drinking water which, according to the scientific data available, represent no health hazard to humans. In the case of food additives and contaminants, the Joint FAO/WHO Expert Committee on Food Additives (JECFA) and the Joint FAO/WHO Meeting of Expert Bodies on Pesticide Residues (JMPR) carry out safety assessment leading to the establishment of acceptable daily intakes of food additives or tolerable intake values (on a daily or weekly basis) for food contaminants. Substances assessed by both committees undergo a periodic review procedure to ensure that new scientific data are considered whenever available. An Environmental Health Criteria document on "Principles for the Safety Assessment of Food Additives and Contaminants in Food" has been published in 1987 On the level of the European Community, periodic evaluations of food additives and contaminants are conducted through the Scientific Committee on Food (SCF). In the case of drinking water, WHO has provided guideline values for major constituents (both of microbiological and of chemical nature) of drinking water in the "Guidelines for Drinking Water Quality" as a basis for countries to develop standards which will ensure the safety of drinking water supplies. 211 1...... 9 9 2 Toxicological Evaluation Prior to toxicological evaluation, data on the chemical nature of the compound tested and on the nature and levels of impurities, as well as on the stability and the breakdown products should be known. These data are useful both for identification of the chemical to be used in the toxicological studies to follow and for the determination of the types of toxicological studies to be performed. Biological data usually collected include the following: 1) Biochemicalstudies, which describe aspects of absorption, distribution and elimination of the chemical tested, and provide information on its biotransformation and on its effects on biochemical parameters, e.g. enzyme activities; 2) Acute toxicity studies, which describe signs and quantitative aspects of toxicity following acute exposure; 3) Short-term toxicity studies, which - in general - assess the toxic effects of the chemical investigated following its administration for a period of up to three months; 4) Long-term toxicity studies, during which the exposure period covers the entire life span of the test species or over 50% of it; 5) Special studies, dealing among others with pharmacological, neurotoxic, carcinogenic (if not part of long-term toxicity studies), mutagenic and teratologic effects, as well as with observations related to the reproductive function; 6) Observations in man, which usually include biochemical studies dealing with absorption, distribution, metabolism and elimination of a chemical, as well as clinical studies of poisoning cases and epidemiological studies. Metabolic studies are usually thought to complement toxicological tests through facilitation of extrapolation from animal data to man in two ways. One is - via determination of absorption rates and sites of distribution (and, possibly, accumulation) and of routes and rates of excretion - by defining the animal species in which exposure to fluxes of the chemical under investigation and its metabolites most closely resembles exposure in man. In this respect it must be noted that following long-term administration of a compound in the diet, a steady state will be reached; thus, marked differences seen in different species following single dose administration might change or even disappear. Another factor to be considered is that under certain experimental conditions, the use of high doses in toxicity testing may lead to an overload of metabolic routes and lead to effects unrepresentative of the situation expected at the actual levels of exposure. The second way by which metabolic studies contribute to the overall assessment is via the determination of the mechanism of toxicity. This is of special importance in the case of compounds which exert carcinogenic effects in animals, as - depending on the mechanism of carcinogenic action - different approaches towards risk evaluation should be used (see below). If a compound exerts a toxic effect through displacement of endogenous substrates from carrier proteins or receptor sites, or via covalent binding to a vital macromolecules, comparative binding studies in different species (also in vitro) will be of great value for interspecies extrapolation. In the case of experimental toxicity studies, the major endpoints can be grouped into the following categories: - Functional manifestations (weight loss, laxative effects, etc.), - non neoplastic lesions with morphological manifestations/organ-directed toxic effects, - neoplastic/carcinogenic manifestations, and - reproduction/developmental manifestations. 212 ...... 1 9 9 2 For a number of potential adverse effects, adequate and quantifiable toxicological endpoints are lacking. This is the case, for example, with neurotoxic and behaviouraleffects. Toxic effects on the nervous system may be detected by means of pathomorphological and/or electrophysiological techniques. The latter lack, however, the capability of providing information on the underlying mechanism(s) responsible for a change of function that might be observed. Complementary biochemical techniques are usually needed for this purpose. Another point to be mentioned is the fact that mechanisms of toxicity towards sensory organs are poorly understood. There is therefore a need for the development and validation of standard techniques for the investigation of neurotoxic effects, and for incorporation of these techniques into standard toxicological test procedures. Bebavioural toxicology is a young branch of toxicology and needs further development and refinement in terms of both the methodology and the provision of basic background data. The same also applies, in principle, to the detection of effects on the immune system, both of immunotoxic and of allergenic nature. Although a number of test procedures for the detection of adverse effects on the immune system both in vivo and in vitro have been developed, further improvements are needed to permit their incorporation into standard protocols for toxicity testing. Besides, the choice of test animal species is problematic, as, on the one hand, there are large species differences in responding to known immunomodulators, and, on the other hand, little is known about the background of these differences. Besides, it is difficult to distinguish between direct adverse effects on the immune system and secondary effects on it resulting from general toxicity following exposure to a substance with a broad spectrum of toxic actions. With allergenic effects, similar problems arise with respect to the choice of animal species, as the commonly employed test species are often insensitive towards such effects. Also, a large number of animals could be needed in order for an allergenic effect to become apparent. Finally, difficulties may arise in the selection of immunological and physiological parameters to be assessed. In the case of reproductive toxicology, further research is needed for a better understanding of the effects observed. The male reproductive system is vulnerable to toxic effects of compounds that induce infertility by affecting the developing gamete directly or that act in one or more of the physiological compartments that form the pituitary-testis hormonal axis. Our limited understanding of basic processes such as spermatogenesis and sperm maturation, and the lack of in vitro tests of sperm function are restrictive towards the development of adequate test strategies. Similarly, there is a need for basic research to elucidate mechanisms of toxic actions towards the female reproductive system. Another area of concern is that of developmental toxicology. Besides the need for appropriate refinement of clinical observations and functional tests with animals exposed pre- and/or perinatally to toxic chemicals, the establishment of an adequate scientific basis for testing effects on behavioural development is ultimately required. A special situation is encountered in long-term testing for chemical carcinogenicity.In fact, a large number of chemicals proved to cause neoplastic lesions in animal species under the existing test procedure of a long-term exposure to partly extreme doses. However, biological responses to very high doses may be unrepresentative of the situation to be expected with the actual levels of exposure. For example, at high doses the kinetics of a chemical compound may be completely altered leading to an unusual distribution or to the saturation of metabolic processes involved in the bioactivation and/or detoxification of chemicals and reactive metabolites. Besides, some chemical agents may lead first to necrosis at high dose levels, and carcinogenic" effects may be secondary to such an action. Therefore, data on the influence of 213 ...... 1 9 9 2 dose levels both on the metabolism and on the toxic effects of a particular chemical in different species should be generated. In addition, the potential mechanisms) of carcinogenic action of the compound should be known to assess the relevance of an observed outcome on the one hand, and to differentiate between genotoxic carcinogenic effects (having no threshold) and epigenetic effects on the other hand. This is also of importance for the safety assessment procedure to be employed, as will be discussed later. In this context, it should be mentioned that standard procedures for the assessment of tumour-promoting effects needs to be developed and validated. Exposure levels The total intake of a chemical by human beings is defined as exposure. Assessment of exposure is of great value, as a compound might be extremely hazardous but, at the same time, pose little risk if the exposure to it is low. On the other hand, a compound that is less hazardous might impose a higher risk of an adverse effect if it is consumed in large amounts. With food additives, low levels of residues from extraction solvents used in extracting fats and oils, in defatting processes and in decaffeinating coffee and tea may be present in food due to incomplete removal. Besides, residues of enzymes and immobilizing agents for them may still be detectable in a final food product. Food flavouring agents, both of natural and of artificial origin, will generally be used in low amounts. By way of contrast, other food additives, e.g. xylitol, sorbitol or modified starches, might be used in large amounts. Human studies Human data, especially those from controlled human exposure studies, are valuable in several ways. They are useful in confirming safety levels as indicated by animal studies and in facilitating their re-evaluation as more data become available. With several contaminants, as well as with food additives during their pre-marketing stage, data relating to human exposure can be obtained through health monitoring of workers coming into contact with these chemicals in the laboratories or in the manufacturing plant, although the routes of exposure are different. Epidemiological studies are usually less sensitive than well-designed animal studies, but this relative insensitivity is somewhat ameliorated when the number of individuals studied becomes very large. In general, such studies are conducted to confirm an adverse effect observed in experimental animals. With food additives, only studies of this kind have been conducted. Saccharin is a good example for the usefulness of epidemiological data. Both, retrospective epidemiological studies and case-control studies revealed negative results with respect to human bladder cancer, in contrast to animal data. In contrast, long-term low level exposure to nitrate, being a ubiquitous contaminant, proved difficult to study epidemiologically due to the fact that subpopulations with little or no exposure to it are hardly ever found. A number of adverse reactions related to intolerance reactions towards certain ingredients (mainly allergic reactions) cannot be detected by satisfactory animal models. One example of such a reaction is the induction by the food colour tartrazine of urticaria and bronchoconstriction in asthmatic patients. Another example is the "Chinese Restaurant Syndrome" which is largely characterized by severe headache and which mainly affects certain sensitive subpopulations; although the mechanism is unknown, it has been suggested to be caused by the consumption of monosodium glutamate a flavour enhancing agent used extensively in oriental cuisine. Recent studies, however, indicate that the Chinese Restaurant 214 ...... 1 9 9 2 Syndrome is due to IgE-mediated type I food allergy caused by the consumption of shrimps, peanuts or spices, in particular those of the parsley family (e.g. coriander). To be able to demonstrate food intolerance reactions, human data of controlled epidemiological as well as challenge-feeding studies are needed. Similar difficulties as with intolerance reactions arise with a certain "perceived" human effects, including neurotropic and behavioural reactions. Again, no adequate animal -modelsexist and results of controlled human studies are needed to evaluate whether there is a scientific basis for a certain perceived effect or not. In order to obtain more reliable human data on adverse effects of food additives and to enlarge the database, the introduction of some mechanism of "post-marketing surveillance" might be useful. In fact, adverse effects of pharmaceuticals and other chemicals are being recorded after their introduction to the market. Reporting on adverse effects of food additives and contaminants would allow the identification of possible toxic and/or intolerance reactions not observed in animal studies and provide the possibility of conducting more adequate retrospective epidemiological studies. Safety assessment The practice of safety assessment depends largely on the type of toxicological endpoint observed. If a certain lesion is considered to have a threshold, which means that if exposure to the chemical investigated is below a certain level, the adverse effect(s) observed at higher dose levels will not occur, the concentration of the chemical yielding no adverse reaction is determined and an acceptable level of intake is calculated for compounds added intentionally to food and drinking water, or a tolerable level of intake is established for contaminants. This approach assumes the ability to define a negligible or no-risk intake in humans. With some types of lesions, mainly with genotoxic carcinogens, it is assumed that any level of exposure represents a certain risk. In the case of a chemical showing such an effect, exposure sould be reduced to the lowest practicable level, and other procedures of risk assessment are employed. In general, safety assessment procedures have to deal with two types of uncertainties: one is related to the extrapolation from the high levels of exposure showing adverse effects as employed in animal studies to lower levels which reflect actual exposure. The second uncertainty is related to the extrapolation from animal to man, as, obviously, "humans are not rats". This is true both in a quantitative as well as in a qualitative sense. The same type of response may occur in animals and in man, though at largely different doses in some instances. Some toxic reactions are limited to a certain species. For example, some synthetic antioxidants lead to proliferative reactions in rodent forestomach. As humans have no such organ, the significance of such a finding for human exposure has to be carefully assessed. In defining safe levels of consumption, any type of evaluation should relate calculated health risks arising from exposure to certain additives or contaminants to those stemming from natural constituents of food and substances formed during the cooking procedures. The threshold approach In general, the safety of chemicals added intentionally to food is evaluated by setting 4 4acceptable daily intakes" (ADIs). With contaminants, it is assumed that their presence in food and drinking water is undesired. Therefore, tolerable intake levels are set based on daily or weekly consumption for chemicals that do not accumulate or those that do, respectively. As a result, "tolerate daily intakes" (TDIs) or "tolerable weekly intakes" (TWIs) are calculated. The 215 ...... 1 9 9 2 term tolerable is used to stress permissibility rather than acceptability for the intake of contaminants at levels which are unavoidable when other-wise wholesome and nutrious food and/or drinking water are consumed. This approach has been followed by JECFA and JMPR as well as by several national and international bodies, including the SCF. JECFA usually adds the term "provisional" to tolerable intake levels of contaminants to exoress the tentative nature of such evaluations in view of the paucity of reliable data on the consequences of human exposure at levels of concern. With trace elements that are essential nutrients at low levels, but which pose health risks at higher concentrations, a range is usually given, the lower value being the level of essentiality and the higher being the maximum tolerable level. The first step in safety assessment is the determination of the "no observed effect level" (NOEL). Usually, toxicological studies involve multiple exposures of a certain test species to the chemical investigated. The doses are ideally chosen in a way which ensures that at least one dose will induce no overt toxicity. For the establishment of NOELs, the most sensitive species is considered. Usually, the most serious adverse effects will be observed in long-term studies, but this is not necessarily the case with every chemical. Some substances may exert toxic effects at lower doses in special toxicity studies, e.g. those designed to detect adverse effects towards the nervous or the reproductive system. Some observed effects in toxicological studies may not necessarily be of adverse nature, e.g. changes in the composition of intestinal flora, laxative effects or cecal enlargement. Reference is therefore often made to a "no observed adverse effect level" (NOAEL). Effects which cannot be readily ascertained as being adverse or not, e.g. a decrease in body weight gain, cause problems. JECFA considers an effect as being adverse whenever there is doubt about its toxicological significance. With chemicals present in food, the NOEL determined is then divided by a safety factor to yield the amount of the chemical, expressed on a body weight basis (usually in mg per kg body weight), which can be ingested daily over a lifetime without appreciable health risk, i.e. the ADI or the TDI in the case of food additives or contaminants, respectively. In general, a safety factor of 100 (10 x 10) is used, to account - on the one hand - for differences in sensitivity between laboratory animals and man, assuming that humans are 10 times more vulnerable to toxic effects than the most sensitive animals species, and - on the other hand - for differences in response to toxic insults by single individuals of a population. Larger safety factors may be applied, e.g. when the database is inadequate or when irreversible developmental effects are observed. If toxicity and dose-response relationships of a compound in humans is known, a lower safety factor might be considered. With food additives, JECFA allocates - apart from regular ADIs - temporaryADIs, mainly for chemicals that are in use but for which the available data are not fully adequate. A date is generally set by which the lacking data must be provided. Conditional ADIs are allocated to chemicals to be used under specified conditions. Group ADIs are given to groups of chemicals with similar chemical and toxicological properties. Besides numerical ADIs, and ADI may be not specified, which means that on the basis of available data, the total daily intake of a certain substance arising from its use at the levels necessary to achieve the desired technological effect and from its acceptable background in food, does not represent a health hazard. Where available data are inadequate, no ADI is allocated, whereas if the available data clearly indicate serious toxicity a not to be used decision is made. The same approach applies in principle to chemicals present in drinking water. Chemical contaminants in drinking water are not usually associated with acute effects; this fact places 216 ...... 1 9 9 2 them in a lower priority category than microbial contaminants. In setting guideline values for chemicals in drinking water, these are related to an average level of exposure to account for the long-term nature of any hazard associated with chemical constituents. Occasional small excesses are acceptable subject to detailed local consideration of their implications. Guideline valuesare based on scientific criteria defining dose-response relationships for substances, on analytical data about the frequency of occurrence and the concentrations of substances commonly found in drinking water, and on the potential application of suitable control techniques to remove or reduce chemical contamination. The objective of developing guideline values is to define a quality of water tan can be consumed by everyone throughout their lifetime with no health risk. The levels defined indicate tolerable concentrations and must not be interpreted as defining final targets for water quality. In setting guideline values based on toxicity data, the safety factor approach for total intake of the chemical is employed. A proportion of the ADI/TDI is then allocated to drinking water as compared to other sources of intake, assuming an average consumption of water of 2 litres per person per day and an average body weight of 70 kg. In the WHO Guidelines for Drinking Water Quality, the percentage of the ADI/TDI allocated to drinking water was determined as an inverse function of the capability of the chemical to accumulate in food chains. For chemicals such as chlorinated pesticides, which accumulate readily, only I % of the TDI is aocated to drinking water, while chemicals that show a lesser tendency to accumulate are allocated a greater proportion. Despite the lack of a scientific justification for the safety factor generally employed from a quantitative point of view, its use has been and is still being widely accepted. One of the major criticisms has been that poor toxicological investigations are rewarded because they tend to lead to a higher NOEL and, consequently, to a larger ADI. This assumption is incorrect. In fact, data from studies with inadequate number of animals and/or with inappropriate measurements of toxic effects usually result in no ADI or a smaller ADI due to the use of a larger safety factor. Nevertheless, as more data (including human data) on interspecies and interindividual variations in toxicokinetics and toxicodynamics become available, and with the further development of mechanistic and in vitro studies, a scientific basis for the use of quantitatively adequate safety factors will be established and other procedures of safety assessment based, e.g. on dose-response effects in various species, will be validated. Non-threshold effects Safety assessment of chemicals showing non-threshold toxic effects, mainly genotoxic carcinogens, is less straightforward. With such chemicals, it is assumed that some risks may be associated with any level of exposure. In general, exposure to such chemicals should be reduced to the lowest level practicable. In the special case of food additives found to be non-threshold toxic agents, their use sould not be permitted. Safety assessment of chemicals the exposure to which cannot be avoided relies in general on some kind of "residual risk" approach. With such an approach, The quantitative dose-response relationship determined at igh doses is extended to estimate the incidence of at lower levels. As a result, a "virtually safe dose" VSD) is estimated. Several models exist which assume that there is a finite risk from any exposure, however small, and that the risk is proportional to dose. They are generally designed to estimate the highest possible upper limit of incremental (excess over background) risk from a lifetime exposure to a particular daily amount of a chemical. Based on socioeconomical judgemental criteria, a risk of I in 100,000 per lifetime is in general arbitrarily selected as "acceptable". If the no-threshold assumption of such a model is invalid, the actual risk could also be "zero". 217 ...... 1 9 9 2 The mathematical models employed for dose extrapolations include the one-hit model, which assumes that cancer can result from a single "hit" such as irreversible damage or interaction with a receptor, the multistep model, which pays credit to the fact that the development of cancer is a process involving multiple stages, as well as a number of pharmacokinetic models, which have been developed to account for kinetic and metabolic aspects such as the dose reaching the target organ, thresholds and interspecies differences in responsiveness, to name but some. All models have advantages and disadvantages, which cannot be discussed in detail in this context. Application of frontline developments in toxicological research Although, in general, thorough consideration of available "classical" data on a particular chemical, including information on chemical and physical properties, toxicology, kinetics, metabolism and mechanism of action, form the basis of safety evaluation, there are instances at which additional studies which are based on frontline developments in reseach may be the key feature of the safety evaluation process. Two examples should illustrate such situations: a) Plasticizers are used in large amounts during the production of polymers. As they are not covalently bound to the resin, they may diffuse from polymers used for food packaging into foodstuffs. Di(2-thylhexyl)phthalate (DEHP) a widely used plasticizer, was shown to induce liver tumours in mice and rats at high concentrations. DEHP was shown to be non- genotoxic. Like other phthalate esters, DEHP was shown to induce the proliferation of hepatic peroxisomes in rodents, and its ability to induce liver carcinoma proved to be associated with its capacity to evoke peroxisomal proliferation. There is evidence from the analysis of liver biopsies of dialysis patients receiving DEHP through blood packaged in PVC bags that peroxisonial proliferation can also occur in man, though available data suggest that humans are less sensitive to such an effect than rodents. Since there is likely to be a threshold for peroxisomal proliferation, it may be concluded that there is also a threshold for carcinogenicity. Consequently, tolerable daily intakes could be calculated based on the no-effect level for peroxisomal proliferation (e.g. by the EEC Working Group on Packaging Materials in 1979). b) Halogenated aromatic compounds, typified by the polychlorinated dibenzo-p-dioxins (PCDDs), dibenzofurans (PCDFs) and biphenyls (PCBs), are industrial compounds and byproducts which have been identified in environmental media including foodstuffs. They exert a number of toxic effects including body weight loss, thymic atrophy, impairment of immune response, hepatotoxicity and porphyria, dermal lesions (e.g. chloracne), tissue-specific hypo- and hyperplastic responses, carcinogenesis, teratogenicity and reproductive toxicity, though large differences in response exist between strains and species. These compounds are also potent inducers of drug-metabolizing enzymes, particularly cytochromes P-450 IAI and P-450 IA2 with their associated nionooxygenase activities (AHH, EROD). Scientific evidence emerged showing that a number of toxic and biochemical responses to these compounds were related to their interaction with a specific cytosolic receptor, the Ah-receptor. The identification of this receptor provided a basis for the study of species and strain differences in response to PCDDs, PCI)Fs and PCBs and for the screening of relative potencies of congeners within each class of compounds, facilitating thereby the development of "toxic equivalency factors" (TEFs in particular and safety assessment in general. Emerging issues Apart from the problems associated with the development of toxicological methodology to assess 218 ...... 1 9 9 2 a number of adverse effects, with exposure assessment, with interspecies and interindividual extrapolation, with the incorporation of toxicokinetic and toxicodynamic data into the safety factor approach, and with the application of methods to assess the safety of chemicals showing non-threshold toxic effects, several issues require attention in the process of chemical safety assessment in food and drinking water. These include: 1) Availability of data. The number of chemicals wich occur in food and drinking water is very large and the changes they undergo during processing is too diverse to allow a comprehensive toxicological evaluation. In order to identify potential hazards and to prioritize testing, there is a need for the development of valid toxicological testing methodologies which provide rapid information. The appropriateness of in vitro methods for this purpose needs to be validated. The same holds true for quantitative structure-activity relationships (QSAR). 2) Data on populationsub-groups. In assessing human exposure, average consumption of food items and drinking water is generally estimated. Problems emerge with foods which are eaten rarely or are consumed by small numbers of people. Another problem is the fact that within a population there are groups of individuals which are particularly vulnerable to some adverse effects of chemicals, not to mention the obvious differences in responding to allergens. There is therefore a need both for data on dietary habits of population sub-groups and for identification of population groups at special risk for certain adverse effects. These include diabetics, women during pregnancy, babies and small children and aged people, to name but a few. 3) Products of biotechnology. In the area of food, novel methods of biotechnology have been introduced, on the one hand to produce novel foods which are natural products, e.g. single cell protein, or to develop new strains of food plants or to modify the characteristics of animals used in food production, and, on the other hand, to improve the production of certain food additives, including amino acids, citric acid, vitamins, enzymes and polysaccharides by microorganisms and plant cell cultures. In the context of this paper, only the latter products of biotechnology can be shortly discussed. In dealing with these chemicals or chemical mixtures, a distinction must be made between intrinsic properties of the chemical and those due to contaminants or residues left by the manufacturing and purification process or to antigenic variation or reversion to the wild type of a living organism. Precise definition of the product and its manufacturing process and control of both the organisms used and the residual processing chemicals are necessary information for toxicological evaluation. In general, the same toxicological procedures are applicable in safety assessment of biotechnology products. In addition, special attention must be paid to the possible emergence of antigenic activity and, consequently, to immunological reactions, to exclude the possibility of enhanced pathogenicity or toxin production by genetically modified organism cells, and to ensure the adequacy of DNA excluding and degrading stages in the manufacturing process. Conclusions Safety evaluation of drinking water and food involves multiple steps including chemical identification, metabolic studies, toxicity testing, exposure assessment and risk evaluation. As our knowledge of mechanisms of toxic actions progresses, improvements need to be introduced into the standard methodologies of toxicity testing and safety evaluation with the aim of guiding policy in the field of consumer protection. In carrying out safety assessment procedures, it should be borne in mind that while we may all agree that a "zero risk" environment is neither attainable nor desirable, tere is a poor correlation between the risks considered to be important 219 ...... 1 9 9 2 by "experts" and the risk concerns of the public. It is therefore necessary to communicate to the consumers that there are different levels of risk, and that there is a need to set the safety of chemicals present in drinking water and food into the context of other risks. Bibhography IPCS: Principles for the Safety Assessment of Food Additives and Contaminants in Food. Environmental Health Criteria 70, WHO, Geneva, 1987 Lovell, D. P.: Risk Assessment of Chemicals. In: Experimental Toxicology. The Basic Principles (Eds. D. Anderson D. M. Conney), The Royal Society of Chemistry, London, 1988, pp. 414-435. Lu, F. C.: Acceptable Daily Intake: Inception, Evolution and Application. Regul. Toxicol. Pharmacol 845-60 (1988). Vettorazzi, G.: Advances in the Safety Evaluation of Food additives. Food Addit. Contam 4 331-356 1987). WHO: Guidelines for Drinking Water Quality. Vols 13, Geneva, 1984. WHO; Strategies for Assessing the Safety of Foods Produced by Biotechnology, Geneva, 1991. 220 9 XA04NO323 Risk Assessment in the Context of EC Directives on Genetically Modified Organisms P J. van der Meer, Ministry for the Environment, the Netherlands This introductory paper focuses on three general questions: 1. What are GMOs according to EC directives? 2. Why risk assessment? 3. How risk assessment? 1. "at are genetically modified organisms (GMOs)? The definition of a GMO used in EC directives is: 6 a GMO means an organism in which the genetic material has been altered in a way that does not occur naturally by mating and/or natural recombination." In order to understand this definition, it is useful to place it in the context of a human activity that has been carried out for thousands of years: selective breeding. Ever since man started to produce his own food, he tried to modify plants, animals and micro- organisms in such a way that they could better serve his purposes. This activity was employed for many different purposes, for example in order to get better and more producing live stock, to get better and more producing plants for food production, and to get better equiped micro- organisms for the production of beer or cheese. If we compare the original species of plants and animals with the results of centuries of breeding and selection, we can see that the results of this activity of man are impressive. For example, over hundreds of years man increased the weight of a maize ear thirty times; and over the milk production by cows was increased from less than half a litre of milk a day to thirty litres daily. However, there is a natural barrier which sets limits to conventional breeding: plants can only cross with certain related plants and animals can only cross with certain related animals. Dogs do not cross with sheep, and tulips do not cross with oak trees. An important step forward in the understanding of the mechanisms and limitations of breeding was delivered by the work of scientists such as Gregor Mendel in the last century, and Watson and Crick in this century. Through the work of these scientists, it became clear that breeding depends on the exchange of genetic information, called DNA. With the development of an understanding of genetics, a new science developed, molecular biology. The first time this term was used was in 1938. With the use of this new technology, scientists attempted to overcome the limitations in breeding set by natural barriers. In the early 70s, the first results were reported. By using the "recombinant DNA" technique, it was possible to "cut and glue" pieces of DNA of one organism, and with the use of a vector, to place it in another organism. Although many technical problems have to be solved, with the use of the recombinant DNA technique, it is in principle possible to bring genetic material of any organism into any other organism. Not only 221 ...... 1 9 I...... 9 2 from any plant into any other plant, or from any animal into any other animal. With this new technique, genetic information of micro-organisms can be introduced in plants, genetic information of human beings can be introduced in micro-organisms (for example to produce insulin), and so forth. In principle, there are no limitations. Since the first recombinant DNA application, several other techniques are developed with which it also is possible to make new combinations of genetic material which are not likely to be produced in nature. Some of those techniques are: - protoplast fusion, or hybridisation: with this technique cells are "fused" after their cell wall is removed; - micro-injection; with a microscopic small pipet DNA is injected into the nucleus of a cell; - DNA particle gun method: tungsten bullets coated with DNA are "simply" shot through cells, leaving the DNA i the cell behind; - organel transplantation: where for instance the nucleus of one cell is transplanted into another cell. These are only examples. Other techniques have been developed, and will be developed. 2. Why risk assessment? The introduction of these new molecular technologies initiated an international discussion on the safety in biotechnology. In 1974 one of the pioneers of this new technology, Paul Berg, expressed his view on te potential risks of recombinant DNA applications in the famous "Berg letter", leading to a self-imposed moratorium on certain experiments. Following the Berg letter and the Asiloniar convention, much international attention has been given to the question of safety in biotechnology. This attention resulted in hundreds of documents, research programmes, guidelines and regulatio 1 vs. This resulted, among others, in two EC Directives on genetically modified organisms: the EC Directive 90/219/EEC on the contained use of genetically modified micro-organisms, and Directive 90/220/EEC on the release of genetically modified organisms. These directives lay down a system for harmonization of risk assessment and risk management with regard to the safety for human health and the environment. Before discussing the general outlines of the risk assessment laid down in those two directives, let us first focus on a fundamental question: why risk assessment? Are GMOs inherently dangerous? No. The technique of genetic modification is eutral, and the resulting organisms (GMOs) are neither inherently dangerous nor inherently safe. So, why risk assessment? The rationale for risk assessment is "accountability beforehand". This is a matter of common sense, nothing more and nothing less. This conference, where risk assessment in the area of chemicals is an important issue, offers a good opportunity to place the rationale for risk assessment of GMOs in a broader context. Formerly, all chemical substances could be placed on the market without further ado. This meant that a chemical substance was only taken off the market when harm had been done. Pesticides being a case in point. 222 M MM ...... 1 9 9 2 Nowadays in the EC, and I am speaking of the period from 1981 onwards, a new chemical substance can only be brought onto the market if it has been established beforehand whether or not harmful effects are to be expected. The procedure would seem to be perfectly logical, but it has taken years and involved a great deal of environmental damage before we reached this stage. We have learnt our lesson from what has happened with new chemicals. With the use of the new molecular techniques, (totally) new combinations of genetic traits can be made with which there is relatively little or no experience. Analogous to new chemicals, this unfamiliarity or uncertainty is the rationale for specific attention. The preventive approach. In addition, it should be emphasized that risk assessment does not imply that risk exists, but simply that it is necessary to establish, by some means or another, whether a risk exists. Summarizing: the rationale for risk assessment is "accountability beforehand". The community at large can only benefit maximally from the potentials of this new technology if it is developed and applied judiciously, in order to avoid to the extent possible negative side effects that have diminished the potentials of many new technologies in the past. This concept has been recognized and laid down in the Agenda 21 Programme "Environmentally sound management of biotechnology" of the United Nations Conference of Environment and Development. Agenda 21 was signed by most countries in Rio in June this year. 3. How risk assessment? In the European Community, the EC Directives mentioned before lay down a general outline for risk assessment. This general outline is presented in the form of "points to consider". Whether or not a certain application of a GMOs, may involve a risk, depends on the characteristics of the GMO involved and the way it is applied. Taking the last item first: the way in which a GMO is applied will influence the potential for risk. In a very general way, a distinction can be made in "contained use" of GMOs and "released into the environment". This distinction can be recognized in the two EC Directives: one directive deals with the contained use and the other deals with the release into the environment. Contained use is described as those applications where certain physical, or physical and biological, barriers are used to limit the contact of the GMOs with the environment. Examples of contained use are: applications in laboratories, research greenhouses and process installations. A release into the environment is defined as every application which is not contained use. With regard to the potential for risk for human health or the environment, it is obvious that there is a difference between the use of organisms in laboratories under strict conditions, and for example a large scale release in the environment. In the first case, the risk assessment may rely primarily on the level of containment (e.g. is the building sufficiently "closed"?). Fifteen years of experience with contained use provided us with a detailed categorization of containment levels and safety procedures. 223 ...... 1 9 9 2 For releases into the environment, the characteristics of the GMO are the key issue. Here, experience and knowledge is relatively limited. The question of describing the characteristics of the GMO, is more complicated. To start with, we should recognize that there are two fundamental questions to be answered: 1) Does the GMO itself, because of the modification, pose a risk to human health and the environment? 2) Can the genetic material of the GMO be transferred to other organisms, which - as a result - can pose a risk to human health or the environment? Let us take again the second question first. It should be recognized that it is very well possible that a gene does not constitute a safety concern in the organism in which it is introduced, but that there may be a safety concern if that gene is transferred to other organisms. For example, the introduction of antibiotic resistance in a soil bacteria does not a priori make that particular bacteria harmful for human health. However, if that antibiotic resistance is transferred to a bacteria which is pathogenic to man, there is a safety concern, because it may have the effect that patients infected with the (resistant) bacteria can not be treated with that antibiotic. TurDing to the first question "does the GMO itself pose a risk to human health and the environment", it should be recognized that much of the methods used to answer this question are still very much of a qualitative nature. The characteristics of a certain GMO have to be known, in order to decide whether the GMO itself poses a risk to human health and the environment. The characteristics of the GMO can be derived from: - The characteristics of the original organism, also called the host of parental organisms). For example, it may be relevant to know whether we are dealing with a sterile culture crop plant which is not able to survive outside agricultural setting, or with a plant which can hybridize with wild relatives and which has weedy characteristics. - The characteristics of the vector, the carrier, may be relevant, if that vector has harmful DNA or RNA sequences. - The characteristics of the introduced genetic material and the related traits. In addition to these elements, empiric data on the GMO, derived from earlier developmental stages or testing, are relevant. In this context, it should be recognized here that organisms are developed and evaluated in a "step wise fashion", through an appropriate continuum, for example from the laboratory stage, through stages of field testing, to final (e.g. commercial) application. Progression through these developmental stages entails in general a reduction of containment, whilst size is often gradually increased. Examples of some issues in the risk assessment of GMOs Some examples may illustrate the complexity of this area of risk assessment. As an example, the main issues will be presented of the risk assessment regarding the introduction into the environment of genetically modified potatoes in the Netherlands. Potatoes are chosen as an example, because potatoes do not cross with the wild flora in the Netherlands. 224 ...... 9 9 2 This keeps the examples surveyable. In the Netherlands, several genetically modified potatoes have been subject to risk assessment for an introduction into the environment. Acknowledging that the characteristics of potatoes and the proposed releases were well described, the risk assessment focused on the introduced new genes and the traits related to this introduced genetic material. The new traits introduced in the potatoes were: - virus resistance - herbicide resistance - insect resistance - bacteria resistance - fungi resistance - antibiotic resistance - non selectable markers - change in starch composition - diminished bruise sensitivity - Erwinia (soil disease) resistance The main issues related to these new genetic traits in these potatoes were very diverse: recombinations of viral material leading to viruses with a new host range, toxicological effects, development of resistance against certain biological pest control agents, effect on non-target or beneficial organisms, frost sensitivity, increased "weediness", and so forth. In the accompanying table, the safety concerns related to the introduced new genes and traits are summarized. This table summarizes the main safety issues that were considered during the risk assessment procedure. This does not imply that after the risk assessment, these issues were reason for concern. For a good understanding of this table, it should be recognized that consideration of some of those issues depend on the scale of the application. For example, if there is uncertainty about potential impacts on beneficial soil organisms, the conclusion may be that such uncertainty is acceptable on a small scale, but not on a large scale. Here again, we should remember the "step wise" development and evaluation of organisms. This example should make clear that this area of risk assessment is still very much in the qualitative stage. This does not imply that the risk assessment methods can not supply decision makers with sufficient information to take decisions in individual cases of releases. However, in order to move forward to more quantitative assessments, more attention should be given to (long term) monitoring and to potential long term ecological effects. 4. Conclusions Turning to the objective of this Conference, several conclusions can be made: - The community at large can only benefit maximally from the potentials of a new technology if it is developed and applied judiciously, in order to avoid to the extent possible negative side effects that have diminished the potentials of many new technologies in the past. - GMOs are neither inherently risky nor inherently safe. However, with the use of the new molecular techniques, new combinations of genetic traits can be made with which there 225 ...... 1 9 9 2 is relatively little or no experience. This unfamiliarity or uncertainty is the rationale for specific attention. - Unlike the chemical and nuclear industry with recognized quantitative methods for risk assessment, assessment of environmental risk of the introduction of organisms is still very much qualitative; - However, fifteen years of experience with contained use provided us with a detailed categorization of assessment procedures and containment levels; - Risk assessment for releases into the environment is still in its early stages of development. Nevertheless, this area of risk assessment is rapidly evolving and already provides decision makers with sufficient data to take decisions in individual cases; - Two major areas for further work in this area are: 1) improve (long term) monitoring methods and standards; 2) stronger focus on potential long term ecological impacts. Examples Trait Safety Considerations * Insert Virus Resist. - Recombinations Viruses with a new host range? * PUX/PLRU Herbicide Resist. - Toxicological considerations * Basta - Increased herbicide use Insect Resist. - Resistance against BT * CRYI(A)/BT - Effect on non target insects Bacteria Resist. - Effect on beneficial soil micro-organisms * Apeadicine IB * Cecropine B Fungi Resist. - Effect on beneficial soil micro-organisms * Osmotine II Antibiotic Resist. * NPT II, HPT, CAT Non Select. Mark. * GUS nt) Starch Composit. - Frost sensitivity * A.S. cDNA Diminished Bruise - Alkaloid Compounds * A.S. cDNA S. Brevidens Fus. - Alkaloid Compounds Erwinia Resist - Volunteers 226 ...... 9 9 XA04NO324 Risk Assessment for Federal Regulatory Decisions on Organisms Produced Through Biotechnology John H. Payne, Ph.D. Terry L. Medley, J.D. Biotechnology, Biologies and Environmental Protection Animal and Plant Health Inspection Service, U.S. Department of Agriculture* John Payne is the Associate Directorand Terry Medley is the Directorof the Biotechnology, Biologics and Environmental ProtectionDivision of the Animal and Plant Health Inspection Service, United States Department of Agriculture, whose main offices are located at 6505 Belcrest Road, Federal Building, Hyattsville, Maryland 20782. *The views in this chapter are solely those of the authors and do not necessarily represent those of the United States government. I. Purposes and History of Risk Assessment: Application to Biotechnology IL Framework in the United States for Decisions on Organisms Produced through Biotechnology III. Choosing from among potential approaches to Assessment A. Exposure Assessment does not equate to Risk Assessment: What are the Hazards? B. Setting Risk Assessment Priorities C. "Quantitative" Environmental and "Quantitative" Ecological Risk Assessments D. Ecological Risk Assessments based on Biological and Ecological Principles IV. The Bases for Good Regulatory Decisions 1. Purposes and History of Risk Assessment: Application to Biotechnology In one of the seminal treatise on risk assessment, Risk Analysis: A Guide to Principles and Methods for Analyzing Health and Environmental Risks, by Cohrssen and Covello (1989), risk analyses are seen to have the principal purpose of leading to better decisions. Methods of analyzing risk have been developed to organize complex risk information to lead to help decision makers determine environmental and health problems associated with a variety of activities and substances, compare the effectiveness of different control and mitigation techniques designed to reduce risks of an activity, and set management priorities, include priorities for regulatory action (Cohrssen and Covello, 1989; Lave, 1987; Russell and Gruber, 1987). There have been recent reviews of the application of risk assessment techniques to the use of organisms modified through biotechnology. The approaches have been quite different and often led to different conclusions. In the preface to Assessing Ecological Risks of Biotechnology, Lev R. Ginzburg writes, "Risk assessment involves predicting certain outcomes and estimating the probability associated with each of these outcomes, as well as assessing their consequences. Such quantitative estimates can only be made with mathematical models" 1991). 227 ...... 1 9 9 2 In contrast, in the preface to Risk Assessment in Agricultural Biotechnology: Proceedings of the International Conference, James J. Marois and George Bruening suggest a more tentative position with respect to the precision that can be expected in risk assessments, "Ecology is not an exact science, and often the application of ecological methods is well ahead of advancements in ecological theory. Also many areas pertinent to biotechnology have little or no ecological basis. Although the population biology of many plants and animals is well understood, most general concepts are under constant refinement. For example, ecologists have a very difficult time predicting the eventual impact that an introduced species will have on an ecosystem. Even the quantification of the attributes that make a species invasive is not completely understood" 1991). To carry out mandates to protect the environment from the introduction of any organism that could be deleterious, regulatory agencies that give approvals for the use of organisms, especially uses that involve release to the environment of organisms which are either new to an environment or which are significantly modified, must ensure that decisions are based on the best information possible and on procedures that lead to good decision making. Given the very different opinions of those expert in risk assessment, what then is an appropriate model for risk assessments to support regulatory decisions? In the following sections of this paper, we will explore the regulatory structure in the United States for review of organisms modified by biotechnology. Compare approaches to risk assessment, and describe a framework for assessment based on principles of biology and ecology. 11. Framework in the United States for Decisions on Organisms Produced Through Biotechnology The U.S. Federal policy has been, and continues to be based on several conclusions: (1)the products of biotechnology will not differ fundamentally from unmodified organisms or from conventional products; 2 the product, rather than the process should be regulated; 3) regulation should be based on the end use of the product and conducted on a case-by-case basis; (4) and that the existing laws provide adequate authority for regulating the products of biotechnology. An important corollary to this policy is the Federal commitment to promoting the safe development of the products of biotechnology. Each Federal agency is committed to ensuring protection for public health and the environment from any potentially harmful effects of the technology. The policy was developed in response to renewed public concern aroused in the early 1980's by proposals to test and use genetically engineered organisms in the environment. The National Institutes of Health (NIH) Guidelines, which were originally written to deal with NIH grantees doing biomedical research in the laboratory, were increasingly proving to be inadequate to the task of reviewing applications for testing in the environment of the broad spectrum of genetically engineered organisms and commercial products (NIH, 1986). The policy was published in June 1986 as the "Coordinated Framework for Regulation of Biotechnology" (OSTP, 1986) and contained final policy statements by the U.S. Federal agencies that share a major responsibility for regulating the products of biotechnology. The agencies are the Food and Drug Administration (FDA), the Environmental Protection Agency (EPA), and the U.S. Department of Agriculture (USDA). 228 ...... 9 9 2 It is important to note that the authority of these three agencies for regulatory biotechnology is based on statute, and that te implementing regulations are published in the U.S. Code of Federal Regulations. The NIH Guidelines by contrast are not based on a statutory basis but rather on a contractual basis. In recognition of this fact and to avoid duplication, the NIH adopted amendments of the Guidelines to provide that if certain experiments are submitted to another Federal agency for approval or official clearance NIH review is not required (NIH, 1987). The following discussion, focuses on the activities of U.S. Federal agencies with specific statutory authority for regulating biotechnology products. FDA FDA, which is part of the Health and Human Services Department of the U.S. Federal Government, regulates foods, human and animal drugs, cosmetics, and medical devices under the authority of the Food, Drug and Cosmetic Act of 1938, as amended. Key amendments to the Act have given FDA authority to require premarket approval of the food and color additives used in food, and premarket safety and efficacy testing for all drugs. FDA does not consider that recombinant derived products require a special or unique review procedure based on process. FDA's organizational units review products developed through many different processes, with attention to scientific concerns, specific tests, and "points to consider". "Points to Consider Documents" have been made available on such subjects as interferons, monoclonal antibodies, recombinant DNA-derived products, and use of mammalian cell lines. Some FDA actions are subject to the requirements of the National Environmental Policy Act of 1969, or NEPA, as amended, and require production of an assessment of the risk to human health and the environment before approval is granted for marketing. Such assessments examine the alternatives to a given action and evaluate data on the potential risks of the favored alternatives. The Food and Drug Administration issued a policy statement on foods derived from new plant varieties, including genetically engineered plant varieties, in May of 1992 (FDA, 1992; Kessler et aL, 1992). In short, that document stated that most substances added to food as a result of genetic modification are substantially similar to substances commonly found in food and therefore should not be subject to premarket "food additive" regulation under section 409 of the FFDCA unless "objective characteristics raise questions of safety sufficient to warrant formal premarket review and approval". Although the document suggests liberal consultation with FDA to determine the regulatory status of foods derived from genetic modification, the implication of the document is that most substances added to food as a result of genetic modification would be considered as "generally recognized as safe" (GRAS) and would not require regulatory approval as a food additive. The FDA food policy clearly establishes that sub 'stances in plants that properly meet the definition of "pesticide" under FIFRA will be addressed through traditional tolerance setting mechanisms with EPA. EPA EPA regulates pesticide products, including those produced through biotechnological techniques, under the authority of the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) of 1972, as amended. Under FIFRA, as amended, EPA has the authority to regulate the development, sale, distribution, use, storage and disposal of pesticides. FIFRA broadly defines 229 M ...... 1 9 9 2 pesticide as "any substance or mixture of substances intended for preventing, destroying, repelling or mitigating any pest, or intended for use as a plant regulator, defoliant, or desiccant' I The development of pest resistant plants through biotechnology has raised the issue on whether FIFRA applies to substances produced in plants. Although EPA has not issued an official policy, EPA officials have stated that pest control substances produced in plants are pesticides within the meaning of FIFRA and therefore potentially subject to regulation under FIFRA. Section 3 of FIFRA generally requires that a pesticide must be registered before sale or distribution. However, under 25(b) of FIFRA it is possible to exempt certain pesticidal substances from regulation if EPA determines that they are adequately regulated by another Federal agency or are not of a character to require regulation. The Toxic Substances Control Act (TSCA) of 1976, provides EPA with the authority to regulate chemical substances, except those used as pesticides, food, and food additives, cosmetics, drugs, and medical devices. The applicability of TSCA to the regulation of microbial products, including those that are produced by genetic engineering technology, is based on the inclusion of "microbes" in the definition of "chemical substance". Because EPA reviews are considered the equivalent of the environmental assessment required under the National Environmental Policy Act NEPA) EPA is generally considered exempt from the procedural requirements of the NEPA. Under both FIFRA and TSCA, EPA is required to consider both the potential benefits and risks of a product. EPA published policy statements in October and December 1984, and in June 1986, that established the interim review procedures for microorganisms subject to FIFRA and TSCA. EPA has not published proposed regulations applicable to microbial products of biotechnology under FIFRA and TSCA, but has operated under the policy in place since 1986 (Rogul and Levin, 1991). USDA USDA has broad regulatory authority to protect U.S. agriculture against threats to animal health, to protect against the adulteration of food products made from livestock and poultry, and to prevent the introduction and dissemination of plant pests. This authority is applicable to genetically engineered animals, plants and microorganisms (Cordle et aL, 1991). The USDA policy on the regulation of biotechnology, consistent with the overall Federal policy (OSTP, 1992), does not view genetically engineered organisms as fundamentally different from those that have traditionally been isolated from nature and introduced into new environments or that have been produced through programs or breeding and selection. The organisms and products produced through the new techniques of biotechnology are regulated under existing laws that apply to naturally occurring organisms and products of traditional technologies. To address the need for specific information necessary for the assessment of the products of the new technologies, a few new regulations have been promulgated and some old ones have been updated. In the area of animal health, the Virus-Serum-Toxin Act VSTA) of 1913, as amended, provides USDA's Animal and Plant Health Inspection Service (APHIS) with the authority to regulate all veterinary biologies that are imported into the United States, shipped or delivered for shipment interstate, intrastate and that are exported. USDA also has enforcement mechanisms such as the 230 ...... 1 9 9 2 power to detain and seize products. The VSTA is administered by APHIS in the same manner for genetically engineered and naturally occurring organisms and products. APHIS issues U.S. Veterinary Biological Product Licenses after satisfactory completion of all requirements to assure purity, safety, potency, and efficacy. Veterinary biological products produced by recombinant methods are evaluated on a case-by-case basis using the same stringent standards for licensing employed for conventionally produced biologics. APHIS has responsibility under the Plant Quarantine Act PQA) of August 20, 1912, as amended, and the Federal Plant Pest Act (FPPA) May 23, 1957, as amended, to regulate and permit the movement of organisms which are or may be plant pests. APHIS published regulations June 16, 1987, that pertain to genetically engineered organisms that are plant pests or that present the potential for plant pest risk (APHIS, 1987). Between the summer of 1987 when the regulations were issued and the fall of 1992, APHIS approved over 350 permits for field trials of genetically engineered plants, with over 650 actual field test sites, involving 20 different plant species. Field trials ave been approved for 35 States and Puerto Rico. Regulations should prevent or at least mitigate risks and not inhibit safe innovation or utilization of new technologies. Regulation and any mandatory review requirements must be balanced and, to the fullest extent possible, commensurate with risk (NRC, 1987; NRC, 1989). The APHIS regulatory structure and its permit requirements are initiated to adequately consider plant pest and environmental risk while facilitating the safe movement for research and other purposes of organisms (McCammon and Medley, 1990; Shaw et aL, 1992). Ill. Choosing from among potential approaches to Assessment A. Exposure Assessment does not Equate to Risk Assessment: What are the Hazards? An underlying assumption to environmental risk assessment that must be dealt with early on in the assessment process is the assumption that hazard exists. Classical risk assessment protocols have been widely developed and used for evaluation of nuclear power installations Okrent, 1987) and for the use of hazardous chemicals in the environment Cohrssen and Covello, 1989). In these applications the hazard is often self-evident in that specific dose-response calculations can be drawn to describe the hazard. A classic formula for characterizing risk assessment for these types of applications is: Risk Hazard Exposure. When the hazard is well understood, most of the assessment then is dependent on descriptions of routes of exposure and determination of potential doses as a result of probable routes of exposure. In this model, as a result that hazard is assumed a priori, identification of routes of exposure leads to the conclusion that there is risk. When applied to the use of genetically modified organisms this model can greatly overemphasize risk, therefore, because biological interactions are complex and, absent a hazard such as pathogenicity, predation, or competition, exposure may not equate to risk. Analysis for hazard is an important component of assessments for organisms modified through biotechnology. Many of the potential types of changes should be expected to lead to safer rather than more hazardous organisms, especially when genes responsible for pathogenesis or other negative traits are deleted or inactivated, such as in many viruses being developed as vaccines. In fact, exposure may lead to risk to the organism being introduced if the introduced organism serves as a host or prey for the organisms it comes in contact with. 231 N ...... 1 9 9 2 Several authors have analyzed exposure assessment. Kareiva 1990) has examined mathematical models of population spread for analyzing results of field releases and has applied these to spread of pollen in crops such as cotton. Supkoff 1990) has developed a model for air transport of microorganisms when those microorganisms are sprayed on vegetation in a field, and Upper and Hirano 1991) have examined aerial dispersal of bacteria in a broader context. Similarly, models of water transport of microorganisms on the soil surface have been developed by Moore 1991), and the transport of microorganisms in soil and groundwater has been considered (Berry and Hagedorn 1991). The discussion treating these studies and exposure assessment, or in some cases a theoretical approach to exposure assessment, is not meant to minimize their proper role in risk assessment. They are essential when a hazard from an organism has been identified, to estimate risk. They are, however, inappropriately used when, absent a hazard, a prediction of exposure leads to the assumption of risk. B. Setting Risk Assessment Priorities If risk assessment is to be used as an effective decision making tool, some mechanism must be used to focus the analysis on the higher risk areas or, with the growing number of uses of organisms, the risk assessment effort will be diffused and wasted on analysis of low risk applications. In the United States, William Reilly, Administrator of the Environmental Protection Agency, has proposed a national debate on risk and priority setting, and several science advisory committees have examined priority setting at regulatory agencies (EPA, 1991). At the USDA we have been examining the priority setting on the basis of risk with respect to the organisms we regulate. We have put forward to the Agricultural Biotechnology Research Advisory Committee the concept that most field crops, modified with a number of the safer traits, if grown under an appropriate set of conditions based on performance standards, could be presumed to be safe and not require a formal assessment before field testing. The committee has endorsed the scientific basis for that approach and will provide detailed comments that we will take into account when implementing that approach. The use of simple notification, rather than more formal assessment and permitting of the safer applications of organisms modified through biotechnology would allow resources to be focused on those with relatively more risk. C. "Quantitative" Environmental and "Quantitative" Ecological Risk Assessments There have been a proliferation of concepts for risk assessment of research in the environment. An environmental risk assessment has been described as a "scientific enterprise in which facts and assumptions are used to estimate the potential for adverse effects on human health or the environment that may result from exposures to specific pollutants or other toxic agents" (Thomas, 1987). As is suggested by the previous definition, the emphasis of environmental risk assessments has often been human health with much of the discussion being devoted to various environmental routes of exposure to humans of toxic materials. When the emphasis is on potential impacts to the environment or its component parts, rather than impacts on humans, the term "ecological risk assessment" has been used (North and Yosie, 1987). Ecological risk assessment has been characterized as "the development of a formal approach to characterize the scientific knowledge of the risk to ecological systems following exposure to environmental contaminants" (Thomas, 1987). 232 ...... 1 9 9 2 When ecological risk assessment has been described by some who approach it from a background in environmental risk assessments of chemicals, they have often described ecological risk assessments as being analogous to a series of additive environmental risk assessments. In other words, the approach would be to develop a numerical dose-response curve for each organism in the ecosystem that might be significantly affected. When we consider the great difficulty that has been expressed in developing this type of data for most most ecosystems for even well-studied chemicals (Silbergeld, 1987; Cohrssen and Covello,, 1989), the lack of success for this type of additive, organism-by-organism numerical model seems evident (Tiedje et aL, 1989). As discussed earlier, an underlying assumption to environmental risk assessment that must be dealt with early on in the assessment process is the assumption that hazard exists. Many recent analysis of environmental and even ecological risk assessments have provided elaborate models for exposure assessment, and purport to tereby measure risk, without carefully assessing what traits in the organisms could logically, let alone be demonstrated, to result in hazard. When the chemical model for risk assessment is used for ecological studies, as it has been described above, ecological risk assessments are often equated to being analogous to a series of additive environmental risk assessments. That approach sometimes leads to a quantification of a few well-studied interactions rather than identification of likely interactions; the reasons being several. There is in the chemical model an over-reliance on quantifiable measures. Since much of the data base for biological interactions is qualitative and qualitative discussion does not fit into the model, it may be overlooked or under-utilized. Another fault of using quantitative risk assessments for complex interactions of biological organisms is that many of the interactions cannot be quantified with any certainty. In many assessments, wen there is uncertainty with a numerical component or the component is expressed as a range, the number used for risk calculation is the most conservative or provides the most risk-averse decision. When several quantitative elements must be estimated in an assessment then the resulting numerical measure of risk may seem to have precision, but has no predictive value or utility to guide informed decisions. Therefore, we conclude that many of the models for "environmental risk assessment" or 64ecological risk assessment" that have been described are not appropriate to provide models for evaluation of the use of modified organisms. Such models are often based on oversimplified models that were developed for use of hazardous chemicals in the environment. Many uncritically assume that hazard exists and seek to characterize the risk from that hazard through exposure assessment. Most do not allow for qualitative data based on experience and observation, instead providing principally for numerical analysis. At APHIS we have sought to identify review strategies that will provide a mechanism for support of decisions on the use in experimental field testing of genetically modified organisms, the release of exotic biocontrol organisms, and the documentation of those decisions. We hve avoided over-reliance on chemical models of risk assessment. The environmental analysis used takes advantage of the more flexible, multi-disciplinary approach of the environmental assessment mandated by NEPA as discussed in the next section. Site-specific, or application specific environmental assessments have proven to be an effective vehicle for ecological risk assessment on decisions for experimental field uses of genetically modified organisms and releases of exotic biocontrol organisms. 233 ...... 1 9 9 2 D. Ecological Risk Assessments based on Biological and Ecological Principles The environmental analysis that is developed for a genetically modified organism, rather than being simply a "risk assessment", and "environmental risk assessment", or an "ecological risk assessment" per se, may contain one or several risk assessments. The environmental assessment might be characterized as containing a biological assessment in a specific ecological framework. That is, the biology and natural history of the organism that is to be introduced is carefully examined, both data from the behavior of the organism in its natural environment and specific host range (or other biological test) done in contained environments like the laboratory or greenhouse. The environment expected to be accessible to the organism when it is introduced is described. Predictions are then made about the expected behavior of the organism in the environment in which it is to be introduced based on familiarity with the ecology of the recipient organism, taking into account the biological functions that have been introduced or changed. The potential impacts to the accessible environment are described and mitigation procedures for impacts are noted. It should be clear that this approach is not fundamentally different from the types of review that have traditionally occurred for introductions of exotic biocontrol organisms (Lima, 1988; Coulson and Soper, 1989; Charudattan, 1990). The principal difference is that the questions are posed more formally than they have been in the past so as to provide a structured review and analysis of the organism, and potential impacts that may occur from its introduction, either positive or negative. Additionally, information about the changes to the organisms have been factored into the review. The major difference between considerations for the introduction of genetically modified organisms and biological control organisms new to an environment, is the behavior of the genetically modified organism is usually much more predictable, most recipient organisms that have been engineered tend to be faimilar organisms of agriculture with which we have much experience. The NEPA is our basic national charter for protection of the environment. It establishes policy, sets goals, and provides means for carrying out the policy. NEPA requires all Federal governmental agencies to prepare a "detailed statement" for major Federal actions significantly affecting the quality of the human environment (Bausch, 1991). This detailed statement is known as an environmental impact statement. An agency determination that the proposed major action would not result in any significant environmental impact must be supported by an environmental assessment. Two fundamental principles underlie NEPA's requirements: Federal agencies have the responsibility to consider the environmental effect of major actions, significantly affecting the human environment and the public has a right to review that consideration. Specifically, where appropriate, an environmental assessment (EA) is prepared in accordance with the National Environmental Policy Act (NEPA) of 1969, as an aide to the decision making process. The EA provides a template for ecological risk assessment for the decision, and ensures that the full range of risks associated with release of the organism have been addressed and properly characterized. This decision making model may provide a valuable template for analogous Federal permitting activities requiring ecological risk assessment. The complexity and uncertainty of ecological risk assessment as compared to assessment of human health risks; and compliance with the intent and procedural requirements of NEPA as a mechanism for documenting ecological risk assessments and assuring informed decisions. 234 ...... 1 9 9 2 The structure of the environmental analysis provided for the following types of questions and considerations: 1. What is the organism and how was it modified? 2. What is the biology of the unmodified (or recipient) organism in its natural environment? 3. What is the biology of the donor or vector genetic sequences or gene products? 4. What are the characteristics of the environment? 5. What other organisms is it expected to significantly interact with in the new environment and what are the potential outcomes of this interaction? 6. What procedures or controls are available to mitigate any negative impacts identified? It is important to note that the questions are of a form that would allow the use of valid qualitative information as well as quantitative data to provide the answers. The use of these types of questions are then analyzed as described in more detail below. The environmental assessment is a key component in the review process. It contains a thorough accounting of the Agency's analysis leading to a decision. It is a public document, made available to anyone who requests it, free of charge. Each environmental assessment is made up of a summary describing the purpose of the document, Departmental regulations, the conditions under which the permit is issued or denied, the background biology of the organisms, and the possible environmental consequences of the field test. The environment that could be affected by the field test is described and the precautions developed for protecting that environment, including field plot design, field inspection and monitoring, test plot security, and termination plans are analyzed. The environmental consequences of the test are examined from a possible perspectives. Consideration is given to the biology of the recipient, donor, and vector, and to the potential for biological containment based on knowledge of this biology. Any possibility of risk to native flora or fauna is evaluated, with special consideration of organisms which are threatened or endangered. Any potential impact on human health is examined. It is through the environmental assessment that the public can be assured that APHIS has fully considered the possible consequences of releasing the regulated article into the human environment. The use of the NEPA environmental assessment as the mechanism for this environmental analysis provides the necessary structure for decision making without imposing inappropriate chemical-based models of risk assessment. It avoids te overemphasis on data tat can be expressed numerically that is a fault of some traditional approaches to risk assessment. IV. The Bases for Good Regulatory Decisions We conclude that, for risk assessments to support fully informed decisions, that they must be based on the best available data and should be based on the method that has the best predictive power. It is important that hazard should be carefully identified and separated from more generalized concerns that may be based more in emotion than fact. Given the level of the characterization biological and ecological interactions, and their complexity, quantitative systems are usually not feasible for prediction of risk from organism based systems. In biological systems, strictly quantitative systems of assessment tend to be over- conservative, leading to risk averse decisions. Strictly quantitative systems tend to cause information useful to risk assessment to be ignored or under-utilized in decision making. Assessments that are based on careful evaluation of the biology of the organism with the application of ecological principles, and that use comparison with experience with similar 235 ...... 1 9 9 2 organisms has proven to be the most useful approach to risk assessment for biotechnology. Biological assessments in the context of an environmental assessment allow the use of both quantitative methods and qualitative measures as they are available or appropriate. Environmental assessment also allows a discussion of what can be called "risk concerns", those concerns which have been expressed with regard to organisms modified through biotechnology that may be based on hypothetical scenarios or other than scientific experience. These concerns may fall outside risk assessment per se, but it is important that in risk communication that all concerns, whether endorsed by experts or not, that are expressed by the general public are appropriately addressed. This allows for communication of relative lower risk of many of these concerns as a basis for discussion. Regulatory Agencies must ensure that the decisions it makes are based on the best information and utilize appropriate procedures. The use of the structured environmental assessment process as a component in the review of organisms modified through biotechnology provides an appropriate structure for the informed and responsible decision making. This pre-decisional tool allows for flexibility of inputs, including qualitative information as well as quantitative data. The environmental assessment provides a template for ecological risk assessment for release decision, and ensures that the full range of risks associated with release of the organism have been addressed and properly characterized. This decision making model should provide a valuable template for analogous Federal permitting activities requiring ecological risk assessment. Literature Cited Animal and Plant Health Inspection Service, U.S. Department of Agriculture 1987. Introduction of organisms and products altered or produced through genetic engineering which are plant pest or which there is reason to believe are plant pests. Federal Register 52: 22892-22915. Bausch, Carl. 1991. Achieving NEPA's Purpose in the 1990's. The Environmental Professional, Vol. 13 (2): 93-184. Berry, Duane F., and Charles Hagedorn 1991. Soil and groundwater transport of microorganisms. In: Assessing Ecological Risks of Biotechnology. (ed., L. R. Ginzburg) Butterworth-Heinemann, Boston. Charudattan, Raghavan 1990. Release of Fungi: Large-scale Use of Fungi as Biological Weed Control Agents. In: Risk Assessment in Agricultural Biotechnology: Proceedings of the International Conference, (eds.) James J. Marois and George Bruening. University of California, Oakland. Cohrssen, John J., and Vincent T. Covello 1989. Risk Analysis: A Guide to Principles and Methods for Analyzing Health and Environmental Risks. U.S. Council on Environmental Quality, Executive Office of the President. Cordle, M., Payne, J., and A. Young 1991. Regulation and oversight of biotechnological. applications for agriculture and forestry. In: Assessing Ecological Risks of Biotechnology. (ed., L. R. Ginzhurg) Butterworth-Heinemann, Boston. Coulson, Jack R., and Richard S. Soper 1989. Protocols for the Introduction of Biological Control Agents in the United States In: Plant Protection and Quarantine, Vol. 1, "Special Topics", (ed) Robert P. Kahn. CRC Press: Boca Raton. Environmental Protection Agency 1991. Setting Environmental Priorities: The debate about risk. EPA journal 17 164. Food and Drug Administration 1992. Statement of policy: Foods derived from new plant varieties; notice. Federal Register 57: 22984-23005. 236 ...... 9 9 2 Ginsburg, Lev R. 1991. Preface. In: Assessing Ecological Risks of Biotechnology. (ed., L. R. Ginsburg) Butterworth-Heinemann, Boston. Kessler, D. A., Taylor, M. R., Maryanski, J. H., Flamm, E. L. and L. S. Kahl 1992. The safety of foods developed by biotechnology. Science 256: 1747-1832. Kareiva, Peter 1990. Using models of population spread to analyze the results of field releases. In: Risk Assessment in Agricultural Biotechnology: Proceedings of the International Conference, (eds.) James J. Marois and George Bruening. University of California, Oakland. Lave, Lester B. 198Z Health and safety risk analyses: Information for better decisions. Science 236: 291-295. Lima, Philip J. 1988. United States Department of Agriculture (USDA) safeguards for introducing natural enemies for biological control of weeds. Proceedings of the VI International Symposium on Biological Control of Weeds 6-11 March 1988, Rome, Italy), E. S. Delfosse (ed.). Marois, James J., and George Bruening 1990. Preface. In: Risk Assessment in Agricultural Biotechnology: Proceedings of the International Conference, (eds.) James J. Marois and George Bruening. University of California, Oakland. National Institutes of Health 1987 NIH guidelines for research involving recombinant DNA molecules. Federal Register 51: 16958. National Institutes of Health 1987 Notice. Actions under guidelines. Federal Register 52: 31848. McCammon, S. L. and T. L. Medley 1990. Certification for the planrpd introduction 'oftransgenic plants into the environment. In: The Molecular and Cellular Biology of the Potato. (eds. M. E. Vayda and W D. Park) CAB International, Wallingford, U.K. Moore, James A. 1991. Surface transport of microorganisms by water. In: Assessing Ecological Risks of Biotechnology. (ed., L. R. Ginzburg) Butterworth-Heinemann, Boston. National Research Council 198Z Agricultural Biotechnology: Strategies for National Competitiveness. National Academy Press, Washington, DC 205 pp. National Research Council 1989. Field Testing Genetically Modified Organisms: Framework for Decisions. National Academy Press, Washington, DC 170 pp. North, Warner and Terry F. Yosie 1987 Risk assessment: What it is; how it works. EPA Joumal 13: 13-15. Office of Science and Technology Policy 1986. Coordinated framework for the regulation of biotechnology. Federal Register 51: 23301-23350. Office of Science and Technology Policy 1992. Exercise of federal oversight within scope of statutory authority: Planned introductions of Biotechnology products into the environment Federal Register 57: 6753-6762. Okrent, David 1987. The safety goals of the U.S. Nuclear Regulatory Commission. Science 236: 296-300. Rogul, M. and Levin, M. 1991. Regulation of Biotechnology by the Environmental Protection Agency, In: Assessing Ecological Risks of Biotechnology. (ed., L. R. Ginzburg) Butterworth-Heinemann, Boston. Russell, Milton and Michael Gruber 1987 Risk assessment in environmental policy-making. Science 236: 286-290. Shaw, J. J., Beauchamp, C., Dane, F., and R. J. Kriel 1992. Securing a permit from the United States Department of Agriculture for field work with genetically engineered microbes: a non-prohibitory process. Microbial Releases 1: 51-53. Silbergeld, Ellen 1987. From the outside: An environmentalist's view. EPA Journal 13: 34-35. Supkoff, David 1990. Air transport of microorganisms. In: Risk Assessment in Agricultural Biotechnology: Proceedings of the International Conference, (eds.) James J. Marois and George Bruening. University of California, Oakland. Thomas, Lee M. 1987 Environmental decision-making today: An interview with Lee M. Thomas. EPA Journal 13 25. Tiedje, J. M., Colwell, R. K., Grossman, Y. L., Hodson, R., Lenski, R. E., Mack, N. and Regal, R J. 1989. The planned introduction of genetically engineered organisms: ecological consideration and recommendations. Ecology 70: 298-315. Upper, Christen, D. and Susan S. Hirano 1991. Aerial dispersal of bacteria. n: Assessing Ecological Risks of Biotechnology. (ed., L. R. Ginzburg) Butterworth-Heinemann, Boston. 237 ...... 1 9 9 2 Panel Discussion 1730-18.00 Dr Norman King (Rapporteur, Session 4) I think the main theme of the afternoon has been the difficulties of risk assessment in areas where probably the risks were very low but there is a great generality about them and in some instances we may all form part of the risk group and have relatively little choice on whether to accept the risks or not. So it has been a difficult area for your rapporteur to start pulling general points out of the discussion but let me try. Dr Denner made a plea for, first of all, a common language and that's something which has been raised several times over the last two days and I think it is a point that Dr Harbison may wish to note for his session on Friday. More difficult but very important Dr Denner wanted to see developed a common philosophy in strategies which can accommodate various interests which are clear to everybody concerned, are self-consistent and err on the side of safety. I think that's quite a big hope but nevertheless I think there's a great deal of sense in that. Dr Denner also reminded us that risk assessment is not an end in itself, it's a tool to help us set priorities, to optimise our approaches to ensuring safety. Furthermore, he argued that if we follow that line then we should be able also to measure its effectiveness. I think that's a very valid point. There's no point in developing very sophisticated strategies if at the end of the day you can't actually determine whether they are delivering the goods or not. Finally, Dr Denner reminded us that not all risks in his field are man-made and again, a very important point, that people are very variable. And this variability is further compounded by our own ability to choose, for example in terms of the kinds of food we choose to eat and the amounts we choose to eat. Human variability is a very major factor in this whole area. Dr Skinner's talk emphasised the difficulties of assessing microbiological risks to man. Like Dr Denner, he, too, would like to see developed a conceptual framework in this area. In this area there's no doubt the risks are real. After all, epidemiology began as a study of communicable diseases and many such diseases are still out there waiting for us to lower our guard. I found his exploration of possible end points quite fascinating. How do you handle the situation where infection doesn't lead to disease in a particular individual. Indeed, it renders that individual immune to further infection yet he or she is a focus and a threat to other people, a very difficult concept I think to handle. Dr Younes touched on themes which have already again been raised but nevertheless are worth emphasising yet again: the difficulty of extrapolating from high to low exposures; extrapolating from animal models to man; the argument over thresholds. He also pointed us towards emerging issues on the effects side of the risk equation where neurotoxicity, immunotoxicity, reproductive toxicity and the products of biotechnology are all issues which need to be addressed. He pointed to the need here to develop and to validate (and that's an important point) methods to look at those end points. He, too, picked up the variability argument and pointed to the occurrence of special groups which raises a question in my own mind which perhaps the panel would like to think about and that is, should we be setting out to protect everyone even where a peculiar life-style lies at the route of that particular risk? Should we allow people to take risks in some areas where we ourselves might not be keen to do it? Moving onto the biotechnology. Dr Van der Meer pointed out that GMOs are neither inherently dangerous nor inherently safe and, in fact, that risk assessment is not a threat. It doesn't imply 238 ...... 1 9 9 2 that there is a risk, it simply implies that we ought to look and see if there's a risk. He made a very valuable point that our science base in this area of releasing organisms to the environment is not strong and we need to improve it. We're feeling our way and he emphasised the step by step and case by case basis that is being adopted in many countries; looking at the results of the previous step before you move onto the next one, learning as you go along if you like to think of it that way. On the contained use side of GMOs there's a much happier position. We've got 15 years of experience with a very good safety record indeed, well developed guidance on both risk assessment and risk management. But I would point out that many of the organisms used in contained use, are very specialised and if they did get out the risks are likely to be small anyway. He stressed two fundamental questions. Does the GMO itself pose a risk and can the genetic material that is being implanted in the GMO be transferred with a consequent risk? Again, this transfer of the risk down a chain, if you like, is quite a fascinating topic. He emphasised the need to improve our monitoring methods and to strengthen our ability to predict longterm ecological impacts. I think we would all have to agree with that. Dr Payne brought a rather different focus on the same set of issues. I think in terms of what is done to look at the potential risks we don't really differ very much one country to another; but we have rather different frameworks for doing it and I think that was the main theme that came through. Both he and Dr Van der Meer pointed to the qualitative nature of the assessments which are carried out rather than the quantitative, but he did go on and suggest that perhaps it's not necessary to shift too far from that position but perhaps qualitative is what we should be doing in any case. He went on to point out the flexibility that comes into the environmental assessment way of doing things and the ability of that particular way of looking at the problems to produce something which can be used much more generally than just by the risk assessors and the regulated. I think that is a very valuable point because in the biotechnology field perception is going to be a problem. I think we all agree that. Thank you Mr Chairman. Comments from Bob Clare (Smith Kline Beecham, UK) Two things. First, it struck me all the way through today that we were perhaps in danger, through too good a focus on individual issues, of producing drivers driving in the wrong direction. I just quote some examples. The MDC issue which was raised by Patrick Murphy, and the question that he very rightly asked: what does the public do when it hasn't got a decent paint stripper? Well it probably buys one that's flammable; and the guy who's doing the paint stripping also has an electrical paint stripper and blows himself up. Nitrogen purging is well known I think in most industries as a method of getting over the problem of flammable atmospheres but the apocryphal tale is that we kill more people with nitrogen than we ever did through explosions of flammable atmospheres. The overweening urge, on the part of DG XI, is to get rid of landfill. Indeed it has put landfill below incineration without the generation of energy as an option for waste disposal, plain crazy on any logical evaluation of environmental impact; the argument of nuclear versus fossil fuel; the Peruvian chlorination story which I noticed in one of the papers - a nice story of mistaken risk assessment; and finally can I just recommend Daniel Defoe on plague year 1665 where the closure of houses with one plague victim multiplied the number of plague deaths by three or four times compared with allowing them all to escape (those that could walk). So what I'm saying is we really ought to have integrated risk assessment in the same way that we have integrated pollution control. And let us therefore perhaps go away thinking about the penalties of single focus blinkered assessment. Second point - I was coming this afternoon hoping that 239 ...... 1 I...... 9 9 2 somebody would give me chapter and verse and mathematical equation for Brenner version 2. I haven't found Brenner version 2 indeed the papers didn't mention Brenner at all. It may be that it is a set of mathematics that is so endowed with criticism that nobody dared mention it; but I do hasten to remind people that Brenner is still used for evaluating containment characteristics and criteria for contained uses of GMOs under the current ACGM system and OECD in the UK at any rate and that's the mathematical system, so I'd appreciate some comments about that. And finally Mr Chairman, if I may, to the lady from the Press who is about to ask another question. I might suggest that in answer to her question this morning about multiple small doses I might remind her that the result of all that is death, indeed the result of life is death; and it is perhaps our search for immortality which is leading us down these somewhat tortuous paths that we've been discussing. Dr Jim McQuaid (Chairman, Session 4) Does anyone wish to attempt to answer the middle part of that specific question about Brenner? I tink the rest could be taken as a declaration but if anybody does feel competent to say anything about Brenner can they please do so? Dr Van der Meer (Directorate for Chemicals and Waste Management, The Netherlands) A: Could I make a general statement? The problem I have with your question or statement - I don't know what it is - and I don't know either whether it's supporting what we said or attacking what we said ("take it as an attack" shouted by a member of the audience to laughter). I'll take it as an attack, yes. You made two points: one is that we should not have this single focus so I fully agree with that. One of my examples was that if we look on the risk of getting resistance against the BT toxin, we should also see the advantage of not using the pesticides and I fully agree with that. On the other hand, before you can go into this integrated attack or approach you have to get the elements first so there is not a real problem of having this single focus provided you do realise that at the end you have to integrate them. I fully agree with that and as far as the contained use is concerned, what I have said before is that there are excellent mechanisms for contained use of organisms in general and GMOs follow on with that category. I think the existing mechanisms and containment levels for organisms are very adequate and could be used, so there's nothing to be revised. I think it works. Dr Barry Thomas (Schering Agrochernicals Ltd, UK) A: I have a slight acquaintance with the Brenner system. I think in answer to your "attack" it is fine to criticise a system but if you have something better then all well and good but I think you need to consider how the system started and where it has brought us. It has brought us and the biotechnology industry in the UK in a safe way. It has been operated in such a way as to convince the union side of the HSE that biotechnology is not as hazardous as perhaps they thought it might be and certainly, in the contained area, I'm not aware that the operation of Brenner as actually caused any nasty incidents. So from that point of view it was a pragmatic system, it was introduced at the time when nobody really knew what the true risks were and it was a system - a pragmatic system - that gave you 240 MI ...... 1 9 9 2 a number. I tink anyone involved in the system would admit that there's no precision in the number but it was a starting place and it got us from A to and we're now at B. Now if you wish to progress in a more logical and precise way then, if you have a better system, let's hear it. Comment ftom Dr John Payne (US Department of Agriculture) I won't speak to Brenner specifically, I'll use my own experience and knowledge to perhaps get at it a slightly different way. I think many of these systems were put in place for good reason and it's time to re-examine them, in my opinion. When we have gotten to the point in our own programmes where we have more restrictions and containment for growing genetically modified animal vaccines than we do for field testing those same vaccines it's time to re-examine the containment protocols that we require. Now I've used my own example, you draw whatever conclusions are appropriate. Ms Dieta Heding (Science Environmental Journalist) Q: Well first of all I want to comment on the dying process, of course we all die but I don't want to be forced to push that process. But I have a few questions concerning GMOs. In Rio, which was a milestone, 151 nations signed a Biodiversity Treaty and the rest of the countries said they would sign except one, as we all know; and it's not necessary to hide that George Bush was the most unpopular State leader in Rio and that he crossed his delegation leader who really implored him that he should sign the Biodiversity. But Bush didn't want to, because he said "it's my job to protect US industry". Now my first question is, did the industry buy Mr Bush or did they press him or did they threaten him or what ever did they do because he is not himself a scientist, he could not judge whether he should sign the Biodiversity or not? So this is a very aggressive question but I would like to know who is behind Mr Bush's "no". So secondly, I am very surprised that none of you mentioned the Biodiversity Treaty. Now are you going along with the Biodiversity Treaty lines when you research with GMO? Are you taking it into consideration or are you simply putting it aside? I'm very surprised you didn't mention it. That's my questions. Dr McQuaid I think we will have to call on Dr Payne to defend Mr Bush. Dr Payne A: Mr Chairman, even before you indicated that question might be addressed to me I kind of got that idea. Well let me put it this way: I wasn't sitting at the table when the decisions were made, I wasn't a party to them so I can't really answer your question directly. The US industry across the spectrum of industry has made various statements on that and I don't think you can characterise what industry thinks any more than one could in a few sentences characterise what the press thinks about an issue so I won't try to characterise what industry thinks about not signing. I can only direct you back to the various statements they've made. In terms of the decision, though, it was not that the US Government didn't support biodiversity and the protection of biodiversity, it had to do with the specifics of 241 ...... 1 9 9 2 implementation in the document itself. Supporting and protecting biodiversity are part of each of the programmes of each of the agencies of the Federal Government and many of the issues that we are specifically asked to address in the National Environmental Policy Act have to do with biodiversity and it is part of each of the decisions I mentioned. Ms Heding Q: What makes the US industry so important that as the only industry in the world it is the only one to say no to the Biodiversity Treaty? Dr Payne A: Well I think the US industry is only particularly important to the US President who then made the decision based on advice from the US industry. Mr McQuaid I'll invite Dr Van der Meer to comment as well. Dr Van der Meer A: Thank you Mr Chairman, I'm very pleased with this question especially because I was very deeply involved in the whole UNCED process. There is a misunderstanding living now, and excuse me but this is the result of this misunderstanding. The Agenda 21 and the whole UNCED process add about or 10 priority themes, one of them being the environmentally sound management of biotechnology, one of them being the biodiversity and several other themes. Of course it's true there is one line in the Biodiversity Convention which allows us to have, on a later date if necessary, a protocol added to it dealing with the safety in biotechnology but there is something infinitely more important and that is that in Agenda 21 which was signed by all Member States, including the US, there is this very important section, environmentally sound management on biotechnology, and half of that document - and I had it on one of my slides here because it is the starting point of the UNCED conference on that area - is that in order for us to benefit maximally we should have the preventive judicious approach towards biotechnology and there is a whole scheme and a whole blueprint in there: how to come to internationally agreed principles on safety in biotechnology and all of that is in there, in the document, and we are looking towards implementing that as soon as possible. So your claim that the Biodiversity Treaty was not signed because of this problem with safety in biotechnology, I think that is falsified by the simple fact that President Bush also signed that part of the Agenda 21 in which he said yes, we can agree with this. I think more of the problem with the Biodiversity Convention can be found in issues like patenting and stuff. So I have full sympathy with what you said but I'm afraid it's part of a misunderstanding and in Agenda 21 there is this very important chapter on safety in biotechnology in which the things I just described - the general issues there, the preventive approach - are worked out in detail and what is more important, there is a whole schedule worked out in detail: how to assist and help developing countries to make available to them the existing safety procedures, to adapt, together with them, the safety procedures, finally with a view to coming to an international agreement on safety in 242 ...... 1 9 9 2 biotechnology. So I hope that takes care of your problem, but I have a lot of sympathy with the point. Rudolf Frei (Chemical Safety Inspectorate, Basel, Switzerland) Q: I have a question to our guest, Dr Van der Meer. Can you perhaps give a definition of an accident in your field or perhaps just an unwanted outcome and, depending on your answer, can you also perhaps tell us how to measure that accident? Dr Van der Meer A: Mr Chairman, I thought I really did my best explaining how complex and difficult all this is. What I attempted to do with the example simply of the potato with 12 or 10 different genes in it is to give examples on what may be an adverse effect in this case and that case and that case. It is extremely difficult, if not impossible, to define an accident in this area. It could be called an incident or an accident if the containment is broken and things escape from a laboratory, but still to define the kind of adverse effects is extremely difficult because these may vary from the replacement of one species by another simply by competition or the toxic compounds or the change in geochemical processes or an increased predation or an increased pathogenicity. So, just as John Payne said earlier, the mechanisms themselves, the effects themselves are not new because it is all limited to what is available in nature: toxicity, predation, pathogenicity, that's the area we're looking in and it is almost impossible to describe, it would be very dangerous to describe it, and therefore, as I said before, we are still in the area of a case by case basis. We are describing the kind of effects varying from something which is almost immeasurable, like the competition between two plant species, to something which can be measured quite easily, for instance the occurrence of a new plant virus. So if you ask me: could you give a definition? I can't, even if you take me out for two hours' drink! I can never come up with it. Dr Payne A: I think that's a good question, there are expected effects and then there are unexpected effects. Unexpected effects may have to do with something that one did not expect that caused the organism not to function correctly or that the genetic engineering in fact changed its trait so it was less effective. You can predict that it will happen in a certain number of cases but never which case because one has inserted genetic material somewhere into the genome of the organism and inevitably probability would suggest that you would a certain number of times interrupt essential processes. Those you expect, though, to see in the field and what you have is sick organisms and it affects only those organisms. There are unexpected effects like breaking or loss of effectiveness of confinement procedures. If you've grown a barrier of plants around the experimental plants with the hope that any pollen will be caught by those plants, if you then monitor further out and you find that in fact a compatible plant further away has been successfully pollinated you have an unexpected effect that's more serious because it suggests that the protocols you established were inadequate for containment. I wouldn't use the term accident but that more approximates the way that term is often used. Then there are the more fundamental and serious effects that we are attempting to avoid, namely the unexpected effects of real changes in the ecosystem function. 243 E ...... 1 9 9 2 DrMcQuaid'8closingremark8 Thank you, and I think I'll draw the discussion to a close there. In winding up the session I would express our appreciation to all the presenters of the papers, to the rapporteurs and to the contributors to the discussion and I would ask you to show your appreciation in the usual way. 244 1