HEALTH RESEARCH

FOR

DEVELOPMENT:

A MANUAL

ad. Joyce L. Pickering

Canadian University Consortium for Health In Development HEALTH RESEARCH

for

DEVELOPMENT: A MANUAL

ed. Joyce L. Pickering

Canadian University Consortium for Health In Development

(CUCHID) Printed by:

McGill Printing Services McGill University Montreal, Quebec, Canada April 1997

ISBN 0-7717-0502-6

Any part of this manual, including illustrations, may be copied, reproduced, or adapted to meet local needs, without permission from the authors or publisher, provided parts reproduced are distributed free or for the cost of reproduction (not for profit) and the user credits the source. The Canadian University Consortium for Health in Development would appreciate being sent a copy of any materials in which text or illustrations have been used. CUCHID, 170 Laurier Avenue West, Suite 902, Ottawa, Canada KIP SVS Authors

Frances Aboud PhD Nazilla Khanlou RN MSc Department of Psychology Clinical Health Sciences McGill University (Nursing) Programme 1205 Doctor Penfield Avenue PO Box 5 Montreal, Quebec H3A lBl McMaster University Canada Hamilton, Ontario L8S lCO Canada Peter Barss MD MPH ScD Injury Prevention Module Barthelemy Kuate Defo PhD Direction de le Sante Publique de Montreal Professor of Demography 4835 avenue Christophe-Colomb Department of Demography Montreal, Quebec H2J 3G8 University of Montreal Canada C.P. 6128, Succ. A Montreal, Quebec H3C 317 Will Boyce MSc Canada School of Rehabilitation Therapy - ICACBR Queen's University Charles Larson MD MSc Kingston, Ontario K7L 3N6 Dept. of Epidemiology and Biostatistics Canada McGill University 1020 Pine Avenue West David Fletcher B Art Sci M Ad Ed Montreal, Quebec H3A 1A2 Consultant Canada Holistic Community Pursuits 2528 Sherwood Street Deborah Lehmann MBBS MSc Halifax, Nova Scotia B3L 3G8 Papua New Guinea Institute of Medical Research Canada PO Box 60 441 Goroka, EHP Catherine Hagen MD MSc Papua New Guinea Clinical Instructor University of British Columbia Cathy Lysack MSc Dept. of Family Practice Assistant Professor Prince George Site Department of Occupational Therapy I Suite 3, 1230 Alward Street Shapero Hall I Prince George, B.C. V2M 7Bl Wayne State University ! Canada Detroit, MI 48202 USA James Hanley PhD Dept. of Epidemiology and Biostatistics Marilyn McDowell PhD McGill University Chair, Human Ecology I 1020 Pine Avenue West Mount Saint Vincent University ! Montreal, Quebec H3A 1A2 Halifax, Nova Scotia B3M 4H6 i Canada Canada I I l t I l • • Catherine Merritt MSc Joyce Pickering MD MSc Lab Coordinator, Cell Biology Rm • A4.21, Royal Victoria Hospital Department of Biology 687 Pine Avenue West • Carleton University Montreal, Quebec H3A lAl • Ottawa, Ontario KlS 5B6 Canada Canada • Kay Wotton MD MPH • Shree Mulay PhD Northern Medical Unit • Director of McGill Centre for Research Dept. of Community Health Sciences and Teaching on Women University of Manitoba • M4. 76, Royal Victoria Hospital 5100 - 750 Bannatyne Avenue • 687 Pine Avenue West Winnipeg, Manitoba R3E OW3 • Montreal, Quebec H3A lAl Canada Canada • David Zakus PhD • Sergio Nishioka MD MSc Director, International Health av. teresina 1415 Mount Sinai Hospital • Uberlandia, MG 38405/384 600 University Avenue • Brazil Toronto, Ontario MSG l:XS • Canada Joan Allen-Peters MSc • School of Nutrition and Food Science • Acadia University • Wolfville, Nova Scotia BOP lXO Canada • Acknowledgements • I would like to thank the International Development Research Centre (Canada) and the Canadian • University Consoi:tium for Health in Development for their funding and support for the development and • running of the first two workshops. • The success of the first workshop was due to the contribution of the facilitators (most of whom • are authors of the chapters in this manual), and to all participants who contributed their time, experience, • and ideas to teach us all. I thank Nazilla Khanlou, my research assistant for the Montreal workshop, and the International Health Office at McGill University for their support. • The illustrations are taken from the following sources: Where There Is No Doctor, David Werner, Hesperian Foundation; Helping Health Workers Learn, David Werner and Bill Bower, Hesperian • Foundation; My Name is Today, David Morley and Hermione Lovel, MacMillan Publishers; Clip Art for • Development, Paul Mundy, Hendi Bachtiar, Iowa State University, Ames, Iowa, USA. I would like to • express my appreciation for their generosity in allowing their pictures to be freely reproduced for non­ profit purposes. • Antoinette Sevensma revised and copy edited this manual with me. I thank her for her many • hours of work, without which this manual would not have been possible. • Joyce Pickering • Montreal, 1997 • • • • • TABLE OF CONTENTS INTRODUCTION • J. Pickering • CHAPTER 1 The Why and How of Research ...... 8 • J. Pickering, C. Lysack • CHAPTER2 Introducing Participatory Action Research ...... 18 ,.• K. Wotton, J. Allen-Peters, D. Fletcher, M. McDowell • CHAPTER 3 • Epidem.iologic, Demographic, and Biostatistical Methods: Assessing Numbers, Rates, Causes and Determinants of • Mortality and Morbidity in Developing Areas ...... 48 • P. Barss, D. Lehmann, C. Hagen, J. Hanley, B. Kuate Defo, S. Nishioka , .• CHAPTER 4 • Methods from Social Sciences: Overview of • Quantitative and Qualitative Methods ...... 101 • F. Aboud CHAPTER 5 • Evaluating Community Participation ...... 132 • W. Boyce, N. Khanlou, C. Lysack, S. Mulay, D. Zakus • CHAPTER 6 Practical Issues in Health Research in Developing Countries ...... 163 • C. Larson • CHAPTER 7 • Laboratory Issues in Developing Countries ...... 191 ,.• C. Merritt CHAPTER 8 • Translating Research into Action 220 • C. Larson • CHAPTER 9 • Working with Groups 232 • J. Allen-Peters • WRAP-UP ...... 266 • J. Pickering • • • • INTRODUCTION • J. Pickering • This manual is intended to serve as a resource for running a five-day workshop on health • research for development. • The original idea for this workshop came from members of the Canadian University • Consortium for Health in Development (CUCHID). Funding was obtained from the • International Development Research Centre (Canada) and the first workshop was held in • Montreal, Canada in 1995. The second workshop is to be held in Uganda in 1997 . • The objectives of the workshop are to learn research techniques that are particularly • applicable to health research in developing country settings. It is aimed at people who already • have some research experience, but have not had the opportunity to learn some of these • techniques . • This workshop uses participatory and interactive methods of learning. Thus, the time • for "talks" and "presentations" by leaders should be limited. No session in this workshop should consist entirely of a lecture. Problem solving by individuals and groups is the preferred method • of learning here . • , .• • • • Talk with people •.•..•..•.••....•.•..•. Not at them A suggested schedule for running the workshop follows. Each chapter in this manual • to corresponds to a session and, in addition information, includes exercises and a suggested • schedule for each session. An exception is Chapter 9, Working with Groups. Although group • processes occur throughout the workshop, this chapter was included to consolidate these ideas • and may be used as a reference throughout the workshop. There is no chapter for the optional session on the software program Epilnfo. Consult the Epilnfo manual if it is decided to hold • this session . • One of the most rewarding aspects of the Montreal workshop was the wealth of • experience that the participants brought to the workshop. This, together with the participatory nature of the workshop, resulted in a rich learning time for everyone concerned. It is our hope • that the workshop will be used widely, with similarly positive results . • • • • SUGGESTED SCHEDULE • Mon Tues Wed Thurs • Fri • 8:30 -12:00 Intro. Epid. & Methods from Practical Issues Translating The Why and Demographic Social Sciences Research into • How of Techniques I Overview Action • Research • 12:00-1:30 Lunch Lunch Lunch Lunch Lunch • 1:30 - 5:00 Participatory Epid. & Measuring Laboratory Wrap up and Action Research Demographic Community Issues Evaluation • Techniques II Participation • 5:00 - 5:30 Feedback Feedback Feedback Feedback • session session session session • Evening Welcome Optional • Reception session on Epilnfo. • • SUGGESTIONS • 1. Hold a meeting for all leaders of the workshop on the day before the workshop to go • over objectives and methods of facilitating. Refer to Chapter 9 for group processes. • 2. Emphasize the participatory nature of the workshop. Lectures should be minimal, group • problem solving should be maximal. If those leading the workshop are less comfortable • with participatory and interactive teaching methods, have a training session for them • before the workshop. • 3. Be sensitive to the feedback from the participants, and modify on the spot as much as • possible to respond to their needs. • 4. Aim for 30 to 50 participants in the workshop to maximize interaction among them. • Participants should already have some basic health research training or experience. • 5. An evaluation session held at the end of each day can give useful information for • improving the subsequent sessions. Have all facilitators and several representatives from the participants attend these sessions. Make them brief and to the point however -- they • do occur at the end of a long day. • • • • • • • CHAPTER I • THE WHY AND HOW OF RESEARCH ,.• J. Pickering and C. Lysack WHY Do HEALTH REsEARCH? ' • Why do health research? For some who have been involved for many years in health or • health research, the answer is obvious. Research is done to find solutions to health problems • and is thereby one of the stepping stones to improving health and well-being around the • world . • Others are more cynical. They observe that some health research focuses on issues • that seem unimportant, leaving untouched questions that are both more important (such as :. questions of equity) and more immediate. Moreover, health research costs money, perhaps • taking resources away from services that are seriously underfunded. One can also be disillusioned by the "research industry." Some health research seems to be done largely to • obtain large grants and promote the careers of individual researchers and their institutions. • They apparently have lost sight of the long-term goal of improving health and well-being • around the world . • These questions and problems become even more pressing when we think about health • research in developing countries. First, if research helps find solutions to problems and thus • improves health, a lot more health research should be done in developing countries, which • have more health problems than many developed countries. In spite of this, much less money is spent on researching the health problems of developing countries than on the health ,.• problems of developed countries. Researching "unimportant" health issues is a waste of limited resources, all the more tragic in a developing country. Precious resources can also be wasted when important, high­ • priority health problems are studied, if the researchers lose sight of the ultimate goal of • application or problem solving. Many readers may be familiar with research projects that • have been completed and written up, only to gather dust on someone's shelf. Nothing seems to have changed or improved because of those research projects. , .• • How TO DO HEALTH REsEARCH • This manual focuses on health research for development -- that is, health research that leads to better health and well-being, particularly in developing countries. Much of the • material presented is not unique to developing countries, but is placed in the context of • developing countries . • 8 • • • • Basic epidemiologic and social science research principles apply across all countries. • In this manual, we are not implying that one type of health research exists for wealthy countries and communities, and another type for poor ones. But we do focus on the types of • research problems that are much more likely to arise in developing countries than in • developed countries. One example is research that looks at basic health indicators such as • infant mortality rates. In countries with birth and death registration (usually the wealthier countries), these indicators are easily available. In countries with no birth and death • registration, a creative method may need to be developed to measure such an indicator. • Subsequent chapters will deal with some of the more specific issues on gathering data to • solve health problems in developing countries. However, before specific techniques are • touched upon, some fundamental issues about health must be understood. • WHY ARE SOME PEoPLE HEALTHY AND OTHERS NOT! • Many people have written about health determinants, that is, the reasons why some people (or populations) are healthy, and others are not so healthy. In addition to reasons like • genes and specific microbes, it is well known that such things as the money, education, and • power people have are very important to their health. Power refers to· the ability to make • real choices, not only about health issues but other issues as well. Working to discover why • some people have more access to resources, education, and decision-making structures is therefore a necessary role of health research. This may be why the ideas· of participatory • action research (PAR) first evolved in poorer countries. (PAR has also been used • extensively in adult education in developed countries. Adult education often involves • working with poorer, less formally educated, and therefore less powerful people.) • • • • • • • • • 9 • • • • • When people or communities gain the power to make changes in their individual lives or communities, we say they have been empowered. The term empowerment is usually used • in a positive way, meaning to give power to those who lack it. When people who are • already quite powerful get more power, we do not use the term empowerment . • Why is the idea of power important in research? If both the solutions to health problems and power itself contribute to health, the effect that research has on power is worth • looking at. We need to ask whether a particular piece of research empowered the people • who needed it. If it didn't, or if it took power away from people, the good effect of the new ;. knowledge created by the research might even be cancelled out by the power that it took away! How can the research process itself empower people? When they are actively • involved in it. The process of investigation, education, and action on factors that influence • one's life can help develop the confidence that comes from realizing that one is not just an • object being acted upon by the world, but an active player in a creative process of • development. • Many researchers are not used to the idea of looking at whether their research • empowers people. They think it is enough only to get valid information (information that • reflects reality). Valid information is fundamental to problem solving, but it is not enough if • we are serious about improving the health and well-being of people anywhere in the world . • How CAN A REsEARCH PROJECT AFFECT EMPoWERMENT? • Here is an example of research that could take power away from a community . • A very important university researcher from the capital city comes to do research on ,. risk factors for trachoma in a village. The village is chosen because it has a very high rate of trachoma. The researcher gets permission from the community by telling people that their •• community has a big problem with trachoma, one of the highest rates in the area. Because • she is very well trained, the epidemiologic methods that the researcher uses are excellent . • The researcher brings her research assistants with her who go door to door checking peoples' eyes for trachoma and asking all kinds of questions about their income, their garbage • disposal, their face washing habits, etc., all things that the researcher thinks may be • connected with trachoma. After all the questionnaires are filled, the researcher, with her • assistants, returns to the capital city to analyze her results on her computer. She finds that ,. the most important risk factor for trachoma is infrequent face washing, that is, it is those i. who wash their faces least often who are most likely to get trachoma. She writes up the results and gets them published in an international journal. She then moves on to another • research project . ,. Although it is easy to see that this research project didn't help the community, we need to stop and think about how it could have actually hurt the community. There are • several ways . • 10 ••• • • • First, the community has been told that it is badly off -- it has one of the highest rates • of trachoma in the area. Telling individuals about their problems (they probably already know about them) without helping them find solutions may often make them feel more • discouraged. The same is true for communities. • Second, the community probably feels that research is something they could never do. They have been told that they are the poor, the sick. Better educated people from outside • have to be brought in just to describe their problems. They may have been made to feel • useless and incompetent. • Third, the members of the community may not have appreciated having to give • personal information to someone they didn't know, and having their time taken up in being • examined and questioned even though it was of no benefit to them. • So, although the research results may have been very valid, the research process may have led to the community feeling worse off than ever, and thus actually been bad for them! • Using a participatory action research perspective helps us to look at this particular • research project in another way. For example: • 1. Why was trachoma chosen as a problem for study? Because the researcher thought it • was important, or because the community thought it was important? If the • community was asked first, they might have said that malaria, or the lack of an all­ • weather road, were problems that they were much more interested in solving. • 2. Why were income, face washing, and garbage disposal studied as risk factors? The • community might be much more interested in discussing other things that they think • are important such as foods, local treatments for eye disease, the effect of smoke, etc. • 3. Why could local residents not gather the information? Does it require outside • expertise? • 4. Why not sit around and discuss trachoma as a problem, rather than going door go door? Perhaps some solutions would come from putting heads together on this, rather • than just asking people individually? • 5. Why does the analysis have to be done on a computer in a big city? Cannot local analysis by community members give useful results? • 6. Why should the results be published internationally? The community is the one who • needs the results. • 7. What is to be done about the problem? A description is fine, but action needs to be • taken to reduce the problem of trachoma. At the very least, the community members • need to know that face washing is important, so that they can take steps to encourage • more face washing. • 11 • • i • • PulTJNG VALIDITY AND EMPoWERMENT TOGEIBEK , .• Although these are very important questions, they are only a part of the research process. • The "excellent methods" used by the researchers are important. For example, if trachoma and Vitamin A deficiency were confused, it is unlikely that the results of the study would be ,.• helpful in solving a problem . • Chapter 2 is devoted to techniques and issues used in particpatory action research. If • applied appropriately, these techniques should increase the control of community members in • the context of a health research project and thus favour empowerment. Other chapters look at methods from epidemiology and social sciences that are designed to ensure the most valid • results, that is, results that reflect reality. Many research methods can increase control and • validity at the same time. The two are not mutually exclusive. Although many health • researchers are less familiar with participatory action research techniques, it is worth learning about them because the principle of letting people control their lives is so important to • health . • Life is full of "grays" rather than blacks and whites. Research is like that also. i. Validity is rarely complete. There is always some chance that research findings are wrong -­ that they do not represent reality. Empowerment is also rarely complete. Some health ,.• research requires an outside expert. This expert quite naturally has his own career goals and interests to consider. Communities are not homogeneous. A research project may empower • only certain people in a community -- perhaps those already educated and relatively powerful -- leaving the poor and disorganized no better off than before the research was undertaken . • These things happen in real life. But just as the recognition that validity is never complete • does not prevent us from working towards it, the recognition that empowerment is rarely 1• complete should not prevent us from working towards it. • The goal of this course and manual is to promote research that leads to better health, • particularly for people in developing countries. First, our aim is to increase the validity of • research carried out in developing countries. We introduce research tools that those trained • or working in developed countries may not have had an opportunity to use. We do not go into any detail on standard health research methods that apply to both developed and • developing countries. We assume that the reader already has a basic knowledge of these, or • can obtain this knowledge elsewhere . Second, we aim to increase the positive impact of health research by teaching how to • consider issues of empowerment in research projects. These issues are considered during the • initiation and design of a project, in selecting the tools used to gather information and • analyze data, and in using the results to improve health and well-being . • • • 12 • • • • • NOTES FOR FACILITATORS • The first session, "The Why and How of Research" is very important for a • number of reasons. • 1. The atmosphere and type of interaction that will take place during the • course will be established in the first session. It is important to convey the idea that each participant will have important contributions to make, and to • try and minimize the difference between the facilitators and the participants. • You need to get away from the teacher-student dichotomy and try to move • towards the attitude that all are colleagues. • 2. The underlying assumptions about research, the differences between • participatory action research and "traditional research," etc., need to be • clarified. This provides a framework for the remainder of the course. • Everything else in the course can be understood as either contributing to or increasing validity or empowerment. It is assumed that all participants will • be familiar with the basics of traditional research, whether from a • biomedical or social science background. Fewer will be familiar with • participatory action research. The text of the current chapter gives a simplified explanation of participatory action research, which should give • the participants basic understanding, and the next chapter gives more • details. • 3. Group processes are important because research often takes place in a group, and because moving towards empowerment usually requires an • understanding of groups. In this course, group processes are taught by • activities scattered throughout the week. Although Chapter 9 addresses • group processes, in this course group processes will be "learning by doing." Point this out to the participants at the beginning of the course so that they • get the maximum benefit from the exercises. • • • • • 13 • • • • • • SUGGESTED SCHEDULE • 15 min Start with a brief welcome and an introduction, outlining the purpose of • the course as well as the format. You can also explain logistics such as • lunch time, coffee breaks, etc . • 15 min • • 15 min While still in their groups, have each group review the specific objectives ,.• for the course. Distribute the objectives on page 15. Is this what the • participants want to learn? Are there any changes to be made? 15 min Have each group report back to the plenary their comments about the • objectives. If any topics or sessions should be modified, try to do so. It • is important to demonstrate a response to the participants' needs early on • in the course . • 30min Give a presentation on the Why and How of Research, as in the current • chapter. Use overheads or slides to emphasize key points. You might solicit discussion from the participants. For example, ask how the • trachoma research project could have hurt the community, rather than , .• simply giving the answers. 10 min Present the case study. Use the one provided on community-based • rehabilitation and have copies available for the groups. If you have a • "closer-to-home" case study, use that one instead . • 30 min Have the groups look at the strategies employed in the study and • comment on how they might have, or might not have, increased validity • and empowerment. You might ask each group to focus on one or two strategies. What other methods could have been used to achieve these • ends? • 40min Have each group report back to the plenary with their answers. Allow • the other groups time to comment on each report . • 10 min Wrap up session by briefly summarizing the key points learned. • • 14 • • • • Handout • OBJECTIVES OF THE WORKSHOP • • 1. To examine our motives, values, and assumptions about doing health research in • developing countries. • 2. To explain the role of participatory action research, and how to begin to do it. • 3. To define and describe specific problems and solutions regarding gathering • epidemiologic and demographic data in a developing country. • 4. To show what social science techniques can be useful in health research in developing • countries and demonstrate how to do them. • 5. To examine the importance of community participation in health research in • developing countries, demonstrate how to incorporate it, measure it and evaluate it. • 6. To recognize practical problems that often arise in doing health research in a • developing country and propose some solutions. • 7. To review laboratory problems that may specifically be an issue in health research in • developing countries and how to avoid or solve them. • 8. To examine actions that will most effectively help us to link research with action. • 9. To be convinced of the importance of group processes and to develop skills in • facilitating these. • 10. To "network" with others involved in health research locally, and to share • information with them. • • • • • 15 • • • • • • Handout • CASE STUDY: UNDERSTANDING PARTICIPATION IN COMMUNITY DISABILITY PROJECTS • Most community-based rehabilitation (CBR) programs in developing countries are dependent • on local volunteers to identify people with disabilities and to provide simple rehabilitation interventions. The vast majority of volunteers are women. Yet, we know very little about • them and their reasons for participation in CBR. A study was undertaken to investigate the • motivations of CBR volunteers and identify the barriers they confront in the course of their • duties. This study was conducted with CBR volunteers trained by medical rehabilitation • personnel at a non-governmental organization (NGO) serving people with disabilities in rural • Indonesia . • A qualitative, multi-methods approach was used in this study. Written questionnaires •• provided basic demographic information about the female volunteers as well as insight into their everyday lives. Time spent on CBR activities and a description of their routine • activities were also obtained. Focus group interviews provided an opportunity for CBR • volunteers to share experiences with each other and discuss solutions to frequently • encountered problems. For example, it was learned that negative community attitudes, and the cost of special equipment often prevent families from participating in rehabilitation •• programs. Focus groups were also occasions for experienced volunteers to provide advice to • newer volunteers. For example, techniques to increase the self-confidence of disabled people • were exchanged, as were ideas about low-cost disability equipment and income-generating • opportunities. Finally, individual interviews were conducted with CBR volunteers and the personnel who trained them. This provided an indepth and confidential opportunity to • discuss, among other things, particular problems with the CBR training process, personal • reasons for committing oneself to improving conditions for people with disabilities, and to • better understand the impact of cultural norms and political factors on the delivery of simple • rehabilitation services in the community. The questionnaires, focus group interviews, and individual interviews were conducted by a Canadian female occupational therapist in • conjunction with an indigenous Indonesian who was also female, a physiotherapist, and a • trainer of CBR volunteers at the same organization . • Participatory strategies employed: • Three participatory strategies were deliberately employed to enhance study validity and • enhance the empowerment of the research participants. In small groups, please discuss the • advantages and disadvantages associated with the following strategies in achieving these • goals . • • joint decision-making control over the research topic, design, and interview questions • • a team approach to data collection and data analysis • • the influence of researcher gender ,.\•• 16 , .• • • • RECOMMENDED READING • 1. Cornwall A., Jewkes R. What is Participatory Research? Soc. Sci. Med. Vol 41, No • 12, pp 1667-1676, 1995. • 2. The Commission on Health Research for Development, Health Research: Essential Link to Equity in Development, 1990. International Development Research Centre, • Ottawa, Canada. • 3. Holter I, Schwartz-Barcott D. Action research: What is it? How long has it been • used and how can it be used in nursing? Journal of Advanced Nursing, 18, 298-304. • 4. Yach D. The Use and Value of Qualitative Methods in Health Research in Developing • Countries. Soc. Sci. Med. Vol 35, No. 4, pp. 603-612, 1992. • 5. Mustard JF, Frank J. The Determinants of Health, 1991. Publication Series by the • Canadian Institute for Advanced Research, Toronto, Ontario, Canada. • • • • • • • • • • • • 17 • • • • • • CHAPTER 2 • INTRODUCING PARTICIPATORY ACTION RESEARCH • K. Wotton, J. Allen-Peters, D. Fletcher, M. McDowell • This chapter focuses on Participatory Action Research (PAR) and how it helps to link health • intervention and community development to health research. By participating in various • activities, you will gain skills or practice that can increase participation of community people in research activities and identify the elements of a research process that can help empower a • community to take control of its health issues . • THE PAR APPROACH • PAR is an approach to research that brings together investigation, education, and action at the • community level to allow people to look at health concerns in their lives and their communities . • Characteristics of the PAR approach are: 1 • • it involves the participation of ordinary and oppressed people in problem posing and • solving • • it is educational for both researchers and participants who analyze the causes of problems • through collective discussion and interaction • it includes collective action for radical social change • • it provides a direct link between research and action • • participants are creative actors in the process, not the subjects or objects of research • • the core issue of PAR is power • PAR is built on the belief that people can understand and solve their own problems and • plan actions to achieve their own vision of health. It brings the research process back to the • people to help them create new knowledge about the factors that influence their lives, develop ways to learn and share the new knowledge, and plan actions to produce a more equitable world. ,.• Triangulation is a term used in referring to PAR: a variety of sources of information are used; a multidisciplinary team is involved; and a mixture of tools and techniques are used, both • qualitative and quantitative. Triangulation can apply to research, action, and education as they • proceed in tandem. Also, core issues are triangulated in poverty, power, and pollution . • PAR involves community people in decisions about research issues at every step of the • process so that ultimately they have control of the research project, whereas traditional research • approaches do not often involve the people being studied, and many researchers do not have the skills to involve others equitably. It has been said that epidemiologists are guilty of three "sins" • in carrying out community research: they decontextualize data, they objectify their subjects, and • they ignore questions of meaning . • 18 • • • • In PAR, control of the research is shared. As well, in-depth, long-term interaction • between researchers and subjects helps avoid the "sins" mentioned above. PAR skills therefore are valuable for those doing community research, especially in the health field. • PAR shares much in common with community development. Both require long-term • involvement by outsiders. Ideas for participatory research and community development • initiatives, like seeds, often take root and flourish years after they have been planted. Both need to proceed at the pace of the community which may be in conflict with university research time • frames. • PAR and community development both require respect for community priorities. Community interest in any area may fluctuate over time. Events may galvanize communities • around specific concerns about which they want more information before proceeding to an • outsider's area of interest. PAR can be a powerful tool for advancing community development • that can lead to change. • When this course was held in Montreal in 1995, many misconceptions about PAR • surfaced. Some participants viewed research as either scientific or as participatory. Others saw • a split between "real science" and "warm fuzzies" or good feelings. The latter was taken by • some to include all qualitative research. Some participants argued that PAR had to contain all • elements of the approach right from onset or it was not PAR. • PAR has been identified with the high standards of commitment, control, and social • activism that had been set by popular movements in Latin America whose members have given • their lives in the struggle for their health and their land. Without taking away from the importance of such struggles and their crucial role in bringing attention to the power, • importance, and credibility of PAR, we cannot afford to live in such an either/or world. If • health research is going to lead to the action needed to improve health in the developing world, • that research will need to be more participatory. • Participation means input at all stages from design through data collection to • dissemination. It leads ultimately to control of the process by the community - not control over • or control under, but control shared with outside researchers. Initially, few communities, and • even fewer researchers, will be able to share this control equitably. But, if what is happening • in aboriginal communities across northern Canada is any indication, once the idea is introduced it catches on very quickly. Communities are demonstrating they can gain the skills needed. • Many have now started to exercise their prerogative to set research agendas when it comes to • their own health. • • planning starts with PEOPLE! • • 19 • • • • PAR TOOLS • A wide variety of tools can be used to apply PAR and involve people in every step of investigating health issues: choosing issues to investigate, data collection, analysis and synthesis, • feedback and action. The work of PAR is often done in groups so, planning and decision-: • making tools and group building tools are important. Also, skill in intercultural communication •• is essential for individuals working together in PAR. Culture strongly influences our • communication interactions, not only in international health programs, but whenever or wherever • members of different cultural backgrounds or subcultures try to work together effectively. We • look at various tools in the following activities . • • • • ••• • • : .• i • •• • Culture strongly influences our communication interactions • • , .• • • 20 • • • SUGGESI'ED SCHEDULE • 15-30 min Minitalk on PAR. Provide enough information to engage and stimulate • participants to learn about PAR. Further articles or books can be provided • as handouts or made available for those interested. If you wish to have • discussion about people's perception of PAR, this part may well take 30 minutes. • 30 min Proverb Pairs or Cultural Iceberg - See Activities 4 and 5 • 60 min Dot-mocracy. Have the group select the four tools they would most like to • learn during the workshop. List all the PAR tools on a large piece of paper • posted on the wall. Give each participant four stickers and ask them to vote for the techniques. The following tools are described in this chapter. Divide • into four smaller groups, one each for each tool selected. Each participant • will get the opportunity to do one of the four tools the group as a whole • selects. Four facilitators are needed, one for each group. They need to be prepared to do whichever of the four tools the group chooses. • Voting/Priority Setting • • Dot-mocracy • Ranking • Discussion Activities • Forced choice decisions • Data Collection • • Human continuum sociogram • Mapping and modelling • • Seasonal calendars • • Critical events • Analysis and Synthesis • •SWOT • •Card sort • Suggested Schedule continued on next page. • • • • • 21 • • , .• • SUGGESTED SCHEDULE (CONf'D) • 20 min Report to Plenary. Each group prepares a poster on their tool and • posts it on the wall for viewing at the break. One person from each • group makes a 5-minute presentation to the others on where their tool • could be used and its advantages . ,.• 30 min Case Studies. This summarizing activity will help to clarify and bring ,. together what the participants have learned about PAR. It also serves to get them to think about putting aspects of PAR into their work and • research . ,.• 10 min. To capture some common learnings and to close on a fun note, close with a PAR song! Have the participants form a circle. You can start • by singing: ,. "PAR is working together, PAR is linking research and action, PAR is empowering. That's what PAR is!" You can then sing, "PAR is - • _____ and someone else adds other important factors. Use • any simple tune . • , .• ,.• ! • ! • • ,.• , . • • • 22 • • • • • • • • • • • • • • • • • • • • • 23 • • : .• • Handout :.• •

I • • RANKING • EXAMPLE OF PREFERENCE RANKING Individuals : .• Constraints to A B c D E Total Rank* • Immunization Score • Lack of information 5 5 3 5 4 22 1 • Unsympathetic health 4 3 5 4 5 21 2 • workers • Lack of transportation 3 4 4 1 3 15 3 • Long waiting time 2 1 2 2 2 9 4 • Injection pain 1 2 1 3 1 8 5 j. *Rank - 1st is the most important constraint • ,.• EXAMPLE OF PAIRWISE RANKING MATRIX • Favourite Pastimes

~ .I II • TV Reading Sleep Music Sport Score • TV • Reading :• Sleep • Music • Sport , .• • • 24 • • • • • • • • • • • • • • • • • • • • • 25 • • • • Handout • • Proverbs • Proverbs often describe "should" and "should not" behavioral rules. They are • statements about the nature of the world we live in according to the perceptions we • share with other members of our culture. Ideally the proverbs used are generated by the individuals involved in the training and will come from their different cultures and • family backgrounds. If this is not possible, the facilitator may put together pairs of • proverbs. Some that have been used are provided . • Proverb Pairs • • To everything there is a season, and a time for everything under Heaven. (Hebrew) • Time and tide wait for no man. (British) • The early bird gets the worm. (American) • All things come to him who waits. (Middle East) Choose your neighbours before your house, and your companions before your • journey. (African) • A man's home is his castle. (British) ,.• You can't teach an old dog new tricks. (Canadian) ' What an old man can see sitting, a young man • cannot see standing. (Cameroon) ,. Health is wealth. (Nigerian) With money in your pocket, you are wise, you are handsome, and you can sing well • too! (Jewish) The tallest nail gets hit the hardest. (Japan) ,.• The squeaky wheel gets the grease. (American) • Take care of the pennies and the pounds will take care of themselves. (English) • Penny wise, pound foolish. (English) • Honesty is the best policy. • If you can't say something nice, don't say anything at all. • • • 26 • • Handout • SUMMARY OF UNDERLYING AsSUMPTIONS AND VALUES • Western Eastern • Western refers to nations whose origins of Eastern as used here includes Asian and • law and of reasoning stem from Greek and other indigenous cultures, such as Native Judeo-Christian tradition. American Indian cultures • universe created and controlled by divine universe unfolds itself, not due to outside power power • • universe is lifeless mass universe is one vast living organism, continually changing and impermanent • universe separates the knowing human from universe is one vast living organism of things to know many interrelated forces and parts; humans • are a part of life force • thinking leads to clear and distinct ideas in things are known holistically, not by categories analysis; thinking leads to imprecise • statements. • knowledge leads to awareness of specific purpose of knowledge is to see unity of all • facts things • knowledge comes from scientific method knowledge comes from intuition and analytic logic • growth can lead to social change growth can lead to oneness with the • universe • time is moving from past toward future; time is a continuous wheel; humans • humans synchronize their time with clocks synchronize their time with nature and machines • individual needs may come before group group conformity is necessary for unity • communication is direct and verbal communication is indirect and often silent; • understanding is often grasped by • observation • • Adapted from Yooung Yun Kim Intercultural Personhood: An Integration of Eastern and Western Perspectives, • in Intercultural Communication: A Reader, 7th ed., LA Samovar and RE Porter, eds., Belmont CA, Wadsworth Publishing Company, 1994, pp 415-425. • 27 • • • • • • , .• • • ••• ••• • ••• : •. •• ••• • ••• ••• • • •• • • 28 ~· • • • Handout • THE CULnJRAL ICEBERG • --:n • ..... ~ -.i: -Cj • =... • literature • .-5 :t= • =-I. OI . drama • classical music • • pop~l~· music • folk dand~g • • games · • cooking • dress ·. • notions of modesty • conception of beauty • • ideals governing childraising • rules of descent • • cosmology • ~lationship to-animals • patterns of • superior/subor~nate relations • definition of sin • • courtship practices • conception of justice .• incentives to • work • notions of leadership • tempo of work • patterns of • group decision-making • conception. of cleanliness • attitudes • toward the dependent • theory of disease • approaches to . • problem-solving • conception of status mobility • eye behavior • • roles in relation to status by age, sex. class, occupatio~ kinship, and · so forth • conversational patterns in various social contexts • • conception of past and future • .definition of insanity • nature of • -=0 ..... ·c- friendship• ordering of time • conceptioi:i of "self' • patterns of • = visual perception • preference for competition or cooperation • · • -E... body language • social interaction rate • notions of adolescence • • c. notions about logic. and validity • patterns of handling emotions • • iacial expressions .• arrangement of physical space • . ... AND ·MUCH~ .MUCH MORE:~.· • • • Just as nine-tenths of an iceberg is out of sight (below the water line), so nine-tenths of culture is out of • conscious awareness. The out-of-awareness part of culture has been termed "deep culture." • Source: AFS American Field Service, AFS Student Yearbook and the Arrival Orientation. (New York: AFS American Field • Service), p. 71. Reprinted by permission of AFS American Field Service lntcrcultural Programs/USA. • 29 • • • ••• ,••• ,.,I , . :.,.• i • ••• i.,.'.• , . ,.• ••1 • ;.,• ,.•• , . •• ••i , .• ,.• , . ••• ••• ••\ • 30 ••• • • • • • • • • • • • • • • • • • • • 31 • • • • • I • Handout • EXAMPLE OF Two MAPS OF THE SAME COMMUNITY • • • • • • • , .• • 1l\Jbl01llll¥al/l111J11JJJ1 ~1111'/JJ'v111V1JWlll1M111,, • ••-o•••O•O •o• ·-•• -·~ ,.l • • • • The Men's Map The Boys' Map • ,.:•• , . • ;.• 32 , . • • • • • • • ••• • • • • • • • • • • • • 33 • • • ,. .•, • • Handout • SEASONAL CALENDAR OF A COMMUNITY IN THE PHILIPPINE.5 • • • •

• ... ;/-. ..•.. . ~:'."'~:-a-:-:;.~ ,,...~ ~ 1~-~··;. »,:;._.:> •••• _,,t:;.. ~ ·-·~v.ft,, .;.{, ~~ 11f1£•"' ,....,..,...... _ ... ,.,_:,\...a .~· I~,,,__,, . •,.·~1',j: f..':t·.1J' ~ f\~ ..... -""'oa· ·-: ', ~ •/i' --~~ "'1V~Y ...... _, ·..or.:~;:-~ .- .=::=- ~/,'• '?."'f't"'/, .f,1,l::-:1.}..~ .. ·" ~;::- ·~-' • ...... ,,~...,._...... ~~'ll'·"-''·. .. . ·' ~...... ,,., ,// , ( f'/ It '4. ~ 1.. :-.-.1:£.l:r.~...... -- J d" ,, - ---- fl/ ... ----...... r, • .,.,,.,.,,"I,., ""' •••••••••' J •I( I /, •'I.• f(f "'"'"""'•' '"'"'"'"" ""'""' ,.. ... • c--- ~M£ : .._.,...... 5.,.. - V' T- ~· '~ ~ ~"-'-...-----r-(~-~-~~.--~-.~~-tt.----~------. -----,.~--=-"""T

• 11------~~ ~ J.j.• ~--+-.,/~ P LENT Y '...,__-+----+---.....p=> Sc.Ar.: GE" --+---+---+---+--- 1~d:..~~_.tr~~:i:t'J====::::t:====::::t:====::::t:====::t:==>:> 7 • lie"'-'~·'"?" ~~ ~ 7 • • • FAO (1993), Guidelines for Participatory Nutrition Projects, Food and Agriculture Organization of the United Nations • • 34 • • • • • • • • • • • • • • • • • • • • • 35 • • • j ,.,• , . . • , . ,.• i • ,.,.• , . • ,.i•• ,•' , . • i•• ,.• ,.;.1 ; . • ,•• i•• :.,. i • • 36 ••• • • • • • • • • • • • • • • • • • • • • 37 • • • 1,.••• ' ,.• , . • , .• • ,.1•• :• ·•l ,••. , .• • i••• ,•:• •• ,.i

l • ,.• '

• -~-- • ,•• , . •• i,. 38 , • ••. ••q • • PuITING PAR INTO PRACTICE • PAR is more than just participating in a process of investigation. It is closely linked to processes of education and action. Good PAR needs to be practical, useful, and timely to the • community where the investigation is carried out. It must generate actions that lead to better • health. There is no ideal way to do a PAR project. In the real world, it is often impractical to • design and implement a PAR project from scratch, but it is always possible to build participation • into research activities. We can attempt to integrate the PAR approach into initiatives by working with communities to investigate and learn about their health concerns, and to enable • them to have some control in the process, and the actions that result. • Certain criteria of PAR can be used to construct a continuum between limited PAR projects and rich PAR projects. All projects can be enriched by integrating as many as possible • of the following, to improve levels of participation so that communities have greater control, and • to maintain the components of investigation, education and action: • • those people generally excluded from research are empowered by participating in an equitable manner • • people participate in defining the issues to be researched, ideally having control over the • process • • people participate fully in design, data collection, analysis and synthesis, and new knowledge creation • • results of investigations are owned by participants and the community and are shared in • an appropriate and timely manner • • education and learning of those directly involved in the project are valued, as are • resulting educational interventions • action oriented outcomes are a direct result of the process • Limited PAR projects may include two or three of these characteristics. The richest PAR would • include all. • • • • • • 39 • • • • ,.1 • ••• ; . ••• ,.•

,.l

1 • ••• !j. • 1•:e1 ,•:e! :• , .• • 1•• ,••• ,.·•l • '•• •• I•• i• 40 ••• ••• • • • .• , • • • • • • • • • • • • • • • • 41 • • • l • • • Overhead • • PERSPECTIVES • • • FINISHED FILES • • ARE THE RESULT OF • • YEARS OF SCIENTIFIC • • STUDY COMBINED WITH 1•• • THE EXPERIENCE OF • • MANY PEOPLE • 1•• • • • • 42 • • • Handout • CASE 1: LABRADOR llEALm DATA • A health study on northern Labrador Inuit, Innu and Settlers was commissioned in the 1970s • by the Nascaupi-Montagnais Innu (NMIA) and the Labrador Inuit Association (LIA) to document the health system and health problems. • While the report was widely read and lead to much animated discussion about health • and health care in the communities, its impact on policy makers was less than anticipated • because few disease rates were included. Health data for aboriginal people in Newfoundland was unavailable because statistics were pooled with those from the rest of the province. • In the 1980s, a resident doctor was provided for the northern coast of Labrador. • During discussions with one of the hamlet councils, the doctor was asked to try to document their health status starting with how and why people on the Labrador coast died. The ways in • which data could be collected and the advantages and disadvantages of various types of • information were discussed as were their reliability and validity. Eventually it was decided • to start with a community-by-community review of deaths. • Initially, three years of data were collected. Using community, church and graveyard • records, a list was obtained of everyone from the communities who had died. The richness • and completeness of the information available in the community about the deaths of residents • surprised many. • The information obtained in this way was verified using hospital summaries, nursing • station charts, and autopsy reports. Likewise official sources were checked with friends and • family members who contributed valuable community information of events, outcomes, and circumstances that were not available from official documents. It was decided to extend the • study to twelve years to catch temporal trends as well as to obtain larger numbers. • The finding that one-third of deaths were by violent means mobilized the community. • The situation was first presented to the hamlet councils and then widely discussed in the communities and in the media. A short time later, one community was able to create a safe • house for women, while another started a program for parenting. Alcohol programs were • initiated in several communities. • When the constraints of collecting information about the role of alcohol became clear • to the people through their involvement in discussions about the health study, one community • went on to organize their own study based on family and friends' stories. Presentations on • self determination and land as part of health were given at the Circumpolar Health • Conference by community members. Civil disobedience in response to the health impacts of low level military flights in the area has been ongoing. • • 43 • • • • , .• • Handout ,. CASE 2: HEALTH STUDY FOR ORIENTED STRAND BOARD MILL !.• When a $90 million Oriented Strand Board (OSB) plant was proposed for a small farming ,• town, the community was split over whether the mill was advantageous or not. Those for the mill, welcomed the jobs and the stimulation to the local economy. • Those against were concerned with the poor environmental record of the company, which • had been successfully prosecuted in the USA for environmental infractions to the tune of $11 • million . • In public hearings, local individuals, and environmental groups expressed concern • about how air quality would be monitored and by whom, the sustainability of the forest • harvest needed to supply the plant and the remote but real possibility of accidents. They pressed for a health study to be done and as a result a health study was included in the • environmental licence . • Although health risks were rated low, if the plant operated as intended, several • chemicals used in the process, especially dimethyl diisocyanate (MDn, formaldehyde and • phenol, are known to be toxic to humans . • A three-part health study was required of the company which included socio­ • demographic data, disease rates, a pulmonary function study and a health risk perception • study . • A Community Liaison Committee (CLC) was set up to oversee the study and permit • community participation. The CLC was given an information sharing role with final • decisions made by the chairman. Consultants were hired by the company to undertake the • study . • The study plan was presented to the community at a public meeting before ,. going to the CLC. Attendance at the public meeting was high and the questions asked by the • public were perceptive and thoughtful. • Most CLC members felt satisfied with the process, as they had received little • information previously . ••• i • :••• • 44 • l • • • Handout CASE 3: COMMUNITY DIAGNOSIS IN NIGERIA • A Nigerian local government health department and an indigenous NGO decided to • conduct a community diagnosis in three villages. • First they met with key informants to construct a map of the area. Focus group • interviews were then conducted by the research team and a community assistant. Discussions • lasting two hours were carried out in groups of 10 people. Separate groups were conducted • with men, women and mixed groups of youth. • Open ended discussions explored the concept of health and well-being and factors • influencing them. Common sicknesses in the community were discussed, according to • criteria of perceived severity, prevalence, ease of treatment and prevention. A ranking activity was used to gather options about what they could do about the factors influencing • their well-being. • Following the data collection, team members did a thematic analysis of findings. This • resulted in a brief, but comprehensive look at the perceived health needs and opportunities in each village. Each team member then wrote a story about one person's health situation, to • encapsulate the complex web of factors influencing health in that village. • Feedback sessions were held at large outdoor village meetings. Short drama • presentations were used to share findings. Solutions proposed by various community members were fed back to generate further discussion and analysis. Local government health • department officials attended each meeting. • A small honorarium (US$100) was provided to each community. In one village the traditional leader announced the honorarium would go towards cement for bricks to continue • with construction of a health post. People agreed to get together for further work. • In another village the men and women wanted to split the honorarium. The men used their share to rehabilitate a hand pump in the village. Women used the money for fertilizer • to start a women's community garden as an income generating project. • In the third village policing and security were identified as major health issues. There • were some heated conflicts with a neighbouring village over farm land. The community decided to post police to the community. They also decided to integrate health into the • agenda of a local respected development organization and plan an appropriate project. • • • 45 • • • • • • SUMMARY • PAR approaches are extremely useful in enabling people to understand and address health j.• concerns within their lives and their communities. These approaches can lead to greater equity in research and health service processes, and, ultimately, lead to transformations that ' • promote greater sustainability and well-being for people and the planet. • PAR approaches are complex and do not provide an instant recipe for success. It is • important to promote these approaches, however, and learn some of the tools and skills that • are necessary in order to collectively create a healthier and more equitable world . • • • • WHAT WE KNO\J ABOUT • OUR COMMUNITY • Needs: ---- • Soc 1a I • factors: • • • \ ( .• • • • 1., . i•• • 46 • • • REF'ERENCF.s • 1. Maguire, P. (1987), Doing Participatory Research: A feminist approach. The Centre • for International Education, University of Massachusetts, Amherst, MD. • • RECOMMENDED READING • 1. Arnold, R., Burke, B., James, C., Martin, D., Thomas, B. (1991) Educating for a • Change. Toronto: Between the Lines. • 2. Barnsley J. & Ellis D. (1992) Research for Change: Participatory Action Research for Community Groups, Vancouver: Women's Research Centre. • 3. F.A.O. (1993) Guidelines for Participatory Nutrition Projects. Food and Agricultural • Organization of the United Nations. • 4. Feuerstein, M. (1986) Partners in Evaluation: Evaluating development and community • programmes with participants. London: Macmillan Publishers. • 5. Green L.W., George M.A., et al. (1995) Study of Participatory Research in Health Promotion, Royal Society of Canada: University of BC, Institute of Health Promotion • Research. • 6. International Institute for Environment and Development: Sustainable Agriculture Programme (1992) RRA Notes Number 16: Special Issue on Application for Health. • 7. Kirby, S. & McKenna, K. (1989) Experience Research Social Change: Methods from • the Margins. Toronto: Garamond Press. • 8. Thesis, J. & Grady, H.M. (1991) Participatory Rapid Appraisal for Community • Development: A training manual based on experiences in the Middle East and North • Africa. London: International Institute for Environment and Development. • • • • 47 • • • • • • CHAPTER 3 • EPIDEMIOLOGIC, DEMOGRAPHIC, AND BIOSTATISTICAL • METHODS: ASSESSING NUMBERS, RATES, CAUSES AND • DETERMINANTS OF MORTALITY AND MORBIDITY • IN DEVELOPING AREAS ,. P. Barss, D. Lehmann, C. Hagen, J. Hanley, B. Kuate Defo, S. Mshioka • Epidemiologic and demographic methods provide an approach for assessing risk factors for :• health conditions and for monitoring health trends. When combined with biostatistical sampling ,. methods, it is possible to use small samples to make inferences about much larger populations . These methods are well known and are covered in standard epidemiology and demographic • textbooks and courses . • This chapter focuses on less well-known methods of data collection used to assess risk factors for health conditions and to monitor health trends. The methods we will look at can be • used in developing areas where standard data sources are unavailable or their quality doubtful. • Although epidemiologists can sometimes be guilty of "decontextualizing, objectifying, and • ignoring questions of meaning," as mentioned in the previous chapter, these methods are • valuable in obtaining valid data . • These include: ,.• 1. Mortality surveys: (page 50) ,. • verbal autopsies (mortality interviews) to assess causes of death • verbal autopsies (mortality interviews) to assess circumstances of death •• • sisterhood method to assess rates of maternal mortality ,.• 2. Morbidity surveys: (page 69) • • health interviews as part of long-term demographic surveillance • • health surveys using hospital records • 3. Demographic methods: (page 75) • • use of de jure and defacto population estimates :.l • estimation of age in individuals of unknown age • 4. Statistical methods: (page 91) • • sample surveys • • cluster sampling • 48 j • • • • • SUGGESTED SCHEDULE • You will need at least a day to cover the topics and have practical exercises. • If you do not have a full day, select topics to present, or have an associate present • some of the topics simultaneously for part of the group. You will need more time • if you want to include specific exercises on other related topics that might include disability surveys, surveys in refugee populations, surveys of vaccine efficacy, • surveys of reproductive practices; special issues in surveys of political violence; • basic topics such as calculation of population-based rates, age-adjustment of rates, • or an overview of sampling theory and methods. • While you are planning the session, it would be useful to include a checklist • in the course application form to determine how many participants have had basic • courses in epidemiology, demography, and biostatistics. If most of the participants • have had basic training, you could omit the overview of the basics and use the time for discussion and practical examples and exercises of special techniques. If about • half of the group need an overview of the basics, the group could be divided, or an • overview could be offered to the participants who need it before this course begins. • • • • • • • • • 49 • • • • • • 1. MORTALITY SURVEYS • Special surveys of mortality and morbidity tend to be time-consuming and expensive; however, • they are often the only source of valid and representative information about populations in • developing areas . This section provides an overview of some of the limitations of official data sources . • Special survey techniques may have to be used to obtain valid data on the causes and • circumstances of mortality in a population. We will discuss mortality interviews, including • verbal autopsies, and look at some practical examples. We will also look at the more specific sisterhood method of estimating maternal mortality, since it can be included as part of a general • health survey. The sisterhood method is based upon interview questions about maternal deaths • of sisters of the respondent. • LIMITATIONS OF OFFICIAL DATA SOURCTS • The following problems with official data sources may indicate a need for the use of mortality • interviews or other alternatives to conventional mortality reporting and certification: • Reporting problems • • lack of certification of village deaths with incomplete reporting • • differential reporting bias when using hospital-based data due to varying probability of hospitalization for injuries and other acute and chronic health • conditions (e.g., drowning, suicide, falls, burns, acute lower respiratory • infections, diarrhea, chronic obstructive lung disease) • • inadequate documentation of circumstances of death • Coding problems • inaccurate or inadequate certification by hospital physicians , .• • • problems with inaccurate or inconsistent coding by coders ;. • differential degrees of misclassification bias by tabulators; for example, injuries should be coded with two codes, the nature of injury and external cause codes . • The external cause code refers to the how the injury occurred such as a motor • vehicle crash. The nature of injury describes the actual injury, such as a • fractured tibia. People who are unaware of this double coding system may give • only one code, usually the nature of injury code . • • 50 • • • • Alternative Data Sources for Studying Mortality in Remote Populations • One possible solution for data problems is to improve standard reporting and coding systems for • deaths, including death certification by physicians, reporting by police and coroners, and coding • by staff in hospital record departments and vital statistics. If this is not feasible, other data • sources need to be considered. While mortality interviews in the context of demographic surveillance can be invaluable for studying mortality, there are other options that may be less • costly. Thus, alternatives to conventional mortality reporting and certification include all of the • following: • • questions linked to periodic censuses or other surveys • • improved certification, coding, & tabulation of deaths that occur in health facilities • • special retrospective surveys (single-round mortality surveys) Population-based -- total population, random sampling, cluster sampling • Key informants • • village surveillance and verbal autopsies (multiple-round surveys of mortality and other vital events) • Lay reporters perform regular surveillance of vital events; • Mortality interviewers then conduct verbal autopsies on deaths recorded by lay • reporters • MORTALITY INTERVIEWS • Mortality interviews are interviews of family members or witnesses of a death. They are used to obtain information about causes of death (e.g., pneumonia, diarrhea, measles, injury) and • circumstances of death, including personal, equipment, and environmental risk factors and the • activity at the time of death. • Mortality interviews have been used in many developing countries, in remote indigenous populations in industrialized countries, and in the general population in industrialized countries. • Verbal autopsies are mortality interviews used for ascertaining the cause of death. This provides • data on basic causes of death in populations where vital statistics data are unavailable or invalid. • In populations where vital statistics are adequate to provide basic information about the causes of death, mortality interviews are sometimes used as a supplement to vital statistics in order to • obtain further details about the circumstances or contributing risk factors for certain causes of • death. Although the cause of a particular death might be described simply as "drowning" in • vital statistics, an interview of family members could provide information on the use of alcohol, whether a lifejacket was worn, weather, etc. Mortality interviews can also be used to validate • cause of death data from standard sources such as death certificates and coroners' reports. • • 51 • • • • • The most frequent use of mortality interviews has been to study causes of death in • children. They have also been used to study maternal mortality and other adult deaths associated • with injuries, alcohol, and other causes. So-called "psychological autopsies" have been used to • investigate determinants of suicide . • Validity of Lay Reporting of Vital Events & Verbal Autopsies • Completeness of reporting can be verified by carrying out a periodic census to serve as a check . • The validity of the diagnoses obtained through verbal autopsies has been studied in quite a few • situations. See references 1 to 4 at the end of this chapter. • Characteristic syndromes or collections of symptoms for common medical causes of death • of special interest are defined. During the interview, a list of questions are asked to provide • information to assess whether the death fits the criteria for specific causes of death. The sensitivity and specificity of different combinations of questions can be estimated by measuring • against a "gold standard," although if the standard is relatively poor quality clinical records or • coroners' reports, verbal autopsy might be more valid than the standard. The interview usually • starts with filter questions of high sensitivity and low specificity. Subsequent questions improve specificity. Injuries are usually straightforward, although intentional injuries can be problematic . • For example, certain cultural attributes may lead to under- or over-reporting of suicides or • homicides . • • • • • • e· - ·. . ~ . . ~ • ..... -. .. . '· ..I • ~;.. . .. • - "" • • • 52 • • • • EXAMPLES OF USE OF VERBAL AUfOPSIES (MORTALITY INTERVIEWS) • IN INDUSTRIALIZED AND DEVELOPING COUNTRIES • • United States: • National Followback Survey, National Center for Health Statistics (1993-95) • Includes a national sample of deaths based on death certificates Examples include investigation of association between substance abuse • (tobacco, alcohol, & other drugs) and unintentional & intentional injuries • • Canada: Special study of fatalities among Alberta aboriginals during 1976 • Investigated association of alcohol and social factors5 • • United Kingdom: • Special study of suicide victims during 1970 • Used coroners' records as a source of cases & informants 6 • Investigated social & clinical circumstances of victims • • Papua New Guinea: • Long term demographic surveillance with multiple round recurrent surveys of all causes of death • Special studies of respiratory infection, including vaccine trials 7 • • Bangladesh Long term demographic surveillance with recurrent surveys of all causes of • death • Special studies of diarrhea8 • • Other locations • British MRC unit in the Gambia, French ORSTOM unit in West Africa, 9 • PAHO in the Americas, and others • • • • • • 53 • • • ••• • • PRACTICAL lsSUES WHEN USING VERBAL AUfOPSIES (MORTALITY INTERVIEWS) • The following issues must be considered: • • Who will carry out the interview -- education, ethnicity,. and sex of the • interviewer ,• • Who will be interviewed -- i.e., men may be poor informants for maternal • deaths • • When will the interview be done -- i.e., how soon after death? ,.1 • Will an open-ended interview or closed questions be used, or both? ,.l • For medical conditions where a symptom complex can be diagnostic for a • particular health condition, for example measles, it is desirable to use an initial screening question that is highly sensitive but relatively nonspecific, such as • inquiring about a rash, and then improve specificity by further question(s) such • as symptom complexes that are more specific for the particular illness . • • Use of local terms • • An interviewer must be aware of local concepts of disease and injury, such as • sorcery, blaming others, female pollution of males, and have knowledge of • common diseases in the study area. For example, is malaria prevalent? • • • i •

,.; • ,.i • • • 54 • ,.' •• • • • ••• • • • • • • • • • • • • • • • 55 • • • • • ••• • •••• •• • ,.••• ! • • • • • • • • ••• •• • • I • • 56 ••• • • Handout • • ADDmONAL SYMPTOMS OR DESCRIPTION PROBABLE DIAGNOSIS ICD • CLASSIF1CATION • 05 Acute Abdominal Conditions Oess than 3 weeks duration) • 050 Diarrhoea and/or vomiting • (no blood) Gastroenteritis 009 • 051 Diarrhoea and vomiting • (no blood) with dehydration Gastroenteritis 009 • 052 Bloody diarrhoea, mild abdominal • pain Dysentery 004, 006.0 • 053 Bloody diarrhoea, severe abdominal • pain, vomiting of foul material, • abdominal distention, history of • eating pig, other animal protein or nuts Pig bel 005.2 • 054 Abdominal pain • 055 Vomiting blood Haematemesis, peptic • or gastric ulcer 456.0, 531-534 • 056 Bright bloody stools, no diarrhoea • or fever • 057 Dark bloody stools Melena • 058 Abdominal pain and swelling • 059 Other unspecified • • • 57 • • • i •• ,.• • Handout • AnnmONAL SYMPTOMS OR DESCRIPTION PROBABLE DIAGNOSIS ICD • CLASSIFICATION ,.• 01 Fever with skin manifestations ' • 010 Fever with red rash disappearing Measles 055 011 Fever with vesicular blisters, • healing clear Chickenpox 050 • 012 Fever with pustules, healing ,. Smallpox 050 with pitting • 019 Fever with other and unspecified • skin manifestations • 02 Other Fevers • 020 Fever with coma Encephalitis/ Cerebral • malaria 323/084.9 • 021 Fever with severe constitutional • disturbance, no loss of • consciousness Malaria, proven by :•• positive blood slide 084 022 Fever with neck rigidity • (bulging fontanelle) Meningitis 320-322 023 Fever with paralysis • (muscle pain) Poliomyelitis 045 • 024 Subacute Sclerosing • Panencephalitis - SSPE Confirmed by MO or laboratory • 026 Febrile convulsion :•,.,,• 027 Typhoid Proven by lab investigations 028 Fever with other manifestations All other febrile illnesses ' • including septicaemia 029 Fever with no other manifestations Pyrexia of unknown origin 780.6 • (P.U.O.) • 58 • • Handout • 08 Evidence of lower respiratory tract involvement of less than 4 weeks duration • 080 ALRI (mild) Cough, productive or loose, Acute bronchitis 466.0 • with one of the following: chest pain, Influenza with other 487.1 • breathlessness (sotwin), rapid respiratory respiratory manifestations • rate, crepitations, rhonchi, fever Acute bronchiolitis 466.1 • 081 ALRI (moderate) ALRI associated with Pneumonia 480-486 • distress (breathing difficulty), chest lnfluenzae • pneumonia 487.0 indrawing, nasal flaring, or pleuritic pain Croup 464.3 • interfering with breathing Acute epiglottitis • 082 ALRI (severe) ALRI with cyanosis, shock, • cardiac failure, not feeding • 083 Pertussis. Cough with whoop and/or Whooping cough 033 cough with paroxysmal vomiting • 089 Not fully specified and other • Definition of ALRI death used in pneumococcal vaccine trial in children (Riley et al, Lancet • 1986;2:8977-881): "ALRI was accepted as a cause of death if the child had had cough and breathlessness, • with or without fever, in the period immediately before death (082). If the child had had a cough and fever but no breathlessness, the death was coded simply as respiratory disease [089). A death certificate, • or other account of a terminal illness from the health services, was accepted as it stood unless there were • obvious inconsistencies." • 09 Chronic cough and/or breathlessness of more than 4 weeks duration • 090 Tuberculosis. Loss of weight, blood in Respiratory tuberculosis 010-012 • sputum, recurrent fever, family history • 091 Chronic non-specific lung disease • (6+ months) Chronic bronchitis 491 • Cough with breathlessness often with Emphysema 492 purulent sputum Bronchiectasis 494 • 092 Cor pulmonale. 091 complicated with ankle oedema Cor pulmonale 416 • 093 Asthma. Recurrent breathlessness with Asthma 493 • or without cough, often in a younger person. Symptom-free periods • 094 Persisting respiratory infections; n.b. suspect • tuberculosis, cough 1-5 months • 099 Not fully specified and other • 59 • • • • Handout PAPUA NEW GUINEA INSTITUTE OF MEDICAL RESEARCH • TARI RESEARCH UNIT ,• MORTALITY INVE3TIGATION FORM ·Printed 16/02/'34 , . . ·.. Na;-ne: r=·t.ir· ..!G • IDND 47360 CNJJO ,. Dai;•::? cf 01·rth 00/00.!'33 Pl es Aqe ai::. death .2 i•I • 3'0 • , .• • :'.: :; ::.·1.- ·-~-/: ·.. *; •.::~-.: ·.-- e:.:. t ~ \/E ·:s.:=~ i ~) . · • -~~~--~~}j~~--~--&L

~! I I I I I • I i :t I I I Sick SICK G' "' ? ~------: • ·-::· KOF 0 l'l :: ------~-±--~------~----- • 8L$Gf- CV •

I I I I I SOD N ? 1 I ::t I 1 ' B~eatf11 essne~= • 0 :------: • I y I • CH PEN N Chest pain 0 ------i • ·-;:· FEVER Q) N ~------~-~---~------~-----; Fever • VOMIT - CD ;------:' - • ·-::· H;::TEM T (!j) Vomitt~n5 with bloc~ • ------··------~-- I • Pw.;R.; \' gJ '? ,_ ;------: • y ·-::· ..I with bloc~ PBLUT @) Diarrhoea • :------~----:- BF'EN y ® ? Abdominal pain • : ------;;_~---~------.: .t y .. BSWELL @ "7-' I Abdominal swelling • :---~-----;~-~------~=---~:- '( • JAUNp ~ "? • I 'Ja1..mdice :------~----: • WLDSS '( @ ? ~- . ·: · · l ·:· •Weight l c:i!:S • :~------~------7----: ' •· l . ' I .. •I l • EDEMA y (0 ? · • • :· _,·Oedema • :------~---~----~----: MASH y ? Measles like rash • CJ ~------: y ·-::· • I :- : • PREG Pregnancy· ® i------: • TRAUMA v ·(!),. ·-::1 Trauma ;------;- • j(AI { Q : Hec:-.vy intake of meal:. • - (before illness) • l~S::-______• -----~--c;_·--···-···- • Madie.al. C:)fficer C de::i.t!1 r:ertifir:.:;;te:' [ J • Lab I X-ray diagnosis c J • HEO. nursing sister APO. student nurse, assistant APO • Gare ~~Ker / rEla~iv~ .. ... • Di a_lnoses • • . -. . la --··--· ··----··--·--·-· ' ___ "'; . D r"G:. D t~:~:=~-:: • 1b -··-- ··-· ---· ···--·. ·- ··- ·-·-----·--·------·------·-···-- D Al32 D l Al::·i • DlA&.-.' •

.~. L· -.=i"i\; • Ceder ·-~'--· : =·.:~ l >./: 00.'°. ~··· ·-~· ~ ~~·;;,,,:,; ian·~·1·· 'l: • vv /·ty·· 1' c;cr / • 61 • I • • Handout • TARI SEARCH UNIT • MORTALITY REPORT FORM • PNNO • Name IDNO ,.! Sex SEX • Date of birth DOB i•• Date of death DOD • Census Number CNNO • Mortality Clerk OBS • Date of history 0 0 03 '3 DATE • Information: -~ <.J~.R_/ Relationship: INFO • Story: (what relative said) • f>< t!V\S~V • • • • • ••• • • • i•• • 62 Handout • DYS WKS MTS YRS • ( Sick • Cou gh / ~:: • Cou gh with blood v • ::aOF Breathlessness / • st pain v EE CH PEN • Feve r t • Vom it I • Vom iUblood / ~= • Oiar rhoea I • Oiar rhoea/blood 1- • Abd ominal pain ·v ~= • Abd o $Welling I P¥-i:: • Jau dlice I/ LLIJAUNO • Wei ght loss 1. ~WLOSS • Oed em::i / H--iEOEMA • Mea i/ sles like rash ,.. 1-----1--~ MASH • Heavy protein intake .__--.....__--1 KAI • (when before illness) f!-:~ Place of death: ___...__:Ir ...... ~------1---J.~=-!.-~ MORT • Place of treatment: l'1-S C.. .----...--1 PLEST • Type of .treatment: ______• • • Certainity of death: VVU9~ ~. CERT • Cause of death: CAUSE • - - 1a ·- -· -- OIAG1 • ~ . 1b OIAG2 • 1c OIAG3 • 11a OIAG4 • 11b OIAGS... • • • • • • • THE SISTERHOOD METHOD OF EsTIMATING MATERNAL MORTALITY • The sisterhood method is an indirect method of estimating mortality in developing countries • where direct measures, which include accurate registration of births and deaths through vital • statistics, are not available. First described by Graham, Brass, and Snow in 1989,1 the sisterhood method of estimating maternal mortality has been used in many countries in Africa • 2 3 4 5 and Asia. • • • Because the method uses four simple questions, it can be appended to census • surveys or longer mother and child health surveys. Advantages of using this method are that 6 7 • it is adaptable, easy to use in the field, • and requires much smaller sample sizes than ;.l 8 1• conventional maternal mortality surveys. The sisterhood method is called an indirect measure of maternal mortality because it does • not directly collect data on the numerator and denominator of the conventional measure of • maternal mortality -- the maternal mortality ratio, which is defined as the number of maternal deaths per 100,000 live births. Rather, it collects data on the number of reported adult sisters • dying of maternal causes, and relates this to "sister units of risk," which are calculated from the • respondent's age and published adjustment factors. See references 3 and 7 on the derivation • of the method . • The four questions in a sisterhood survey are: • 1. How many sisters (born to the same mother) have you had who were ever married? • 2. How many of these ever-married sisters are alive now? • 3. How many of these ever-married sisters are dead? 4. How many of these dead sisters died while they were pregnant, during childbirth, or • during the six weeks after the end of pregnancy? • Questions 2 and 3 are designed to provide a cross-check; if the respondent answers correctly, the SUJl?. of responses to questions 2 and 3 should equal the total provided in question • 1. Interviewers find the method easy to grasp and the questions easy to introduce, as they avoid • the potentially sensitive topic of deaths within the home they are visiting . • In Table 1 that follows, the sisterhood method has been demonstrated, using an example of data collected during a household survey in Karachi, Pakistan. The responses of 2651 adults • (respondents may include men and women) are • recorded by respondent age group. Two • important numbers emerge from the table: first, ,. 27 sister deaths from maternal causes were reported; second, 3767 sister units of risk were • calculated by multiplying the number of sisters • reported by published adjustment factors, and • summing these over all respondent age groups . • The result, the life risk of dying of maternal causes, is simply deaths (27) divided by sister • units of risk (3767), or 0.00717 - a little under • 1 %. • • 64 • • TABLE 1: SISTERHOOD DATA AND CALCULATION OF LIFE RisK OF MATERNAL DEAm • I II m IV v VI VII • Respondent's # # maternal adjustment sister life risk Age Respondents sisters deaths factor units • risk • 15 - 19 44 68 0 0.107 7.3 0.00000 • 20 - 24 286 440 1 0.206 90.6 0.01103 • 25 - 29 521 984 7 0.343 337.5 0.02074 • • 30 - 34 533 1988 5 0.503 1000.0 0.00500 • 35 - 39 476 1107 3 0.664 735.0 0.00408 • 40 - 44 341 808 1 0.802 648.0 0.00154 • .. 45 - 49 291 679 4 0.900 611.1 0.00655 • 50 - 54 159 352 6 0.958 337.2 0.01779 • TOTAL 2651 6426 27 3766.8 0.00717 • Notes: • - Columns II, m, IV are results of survey data • - Column V shows adjustment factors provided by Graham and Brass - Column VI is the product of columns V • ill, - Column VII is ratio of column IVNI • *According to method of Graham and Brass, # sisters in Column ill bas been adjusted; # sisters shown in the two youngest age groups is # respondents x 1.54, the average anticipated number of sisters to women in these • categories. The actual numbers of respondents were 49 and 385. • • • 65 • • • • The final step of transforming the life risk of maternal death to an estimate of the • maternal mortality ratio, depends on the availability of an accurate local estimate of the total • fertility rate. The formulas which follow show that the derived MMR from this sample was 153 • maternal deaths per 100,000 live births. Conceptually, it is not difficult to understand that the ·.• life risk of maternal death is approximately the average number of pregnancies a woman might expect (TFR) multiplied by the obstetrical risk in a given pregnancy (the MMR). In this study • the total fertility rate, 4.69, was estimated by asking women questions concerning their recent • reproductive histories in the same survey which collected sisterhood data. Investigators may wish to calculate sampling uncertainty (confidence intervals) for their • estimates of maternal mortality. See page 67 and reference 8 . • LIFE RlsK OF DEAm FROM MATERNAL CAUSES • Life risk maternal deaths • sister units of risk • ...IL • 3767 • 0.00717 • EsTIMATING MATERNAL MORTALITY RATIO ,.• MMR - 1 - [(Probability of survival)1JTFR] "• 1 - [(1- .00717} 1 '4 ·~ 0 231 • 1 - (0.99283) • • 0.00153 • NOTE: • life risk of maternal death is the sum of risks over a number of pregnancies • events are rare, there • approximate MMR can be calculated as a sum of probabilities: • MMR = life risk • TFR • 0.00717 • 4.69 • 0.00153 • 66 • • • • CALCULATION OF THE MATERNAL MORTALITY RATIO (MMR) AND CONFIDENCE • INTERV~ • In step one, the MMR is calculated from the life risk of maternal death as follows: • r = number of maternal deaths (27) B = sister units of risk (3767) • Q = life risk of death from maternal causes = r/B = 0.00717 • P = probability of survival = 1-Q = 1-r/B = 0.99283 • 1 • MMR = 1 - [(Probability of survival) "FRJ • 0 2132 MMR = 1 - [(0.99283) • ] • MMR = 0.00153 • In step two, the standard error of the MMR is calculated using Han1ey's formula which takes • into account variance of P and variance of TFR: • SE(MMR) = 1-MMR 1-P + (log{P}) 2 Var(1FR) }o.s • 1FR { BP 1FR2 • SE(MMR) = 1-.00153 1-.99283 + (log{.99283}) 2 (.005883) }~ • 4.69 { 3767 •.99283 4.65)2 • SE(MMR)=0.000295 • .• , In step 3, upper and lower confidence limits are calculated using the formula: • MMRu = MMR + Z.[SE(MMR)] • = 0.00153 + 1.96(0.000295) = 0.00211 • MMRL = MMR - Z.[SE(MMR)] • = 0.00153 - 1.96(0.000295) • = 0.00095 • Maternal mortality is therefore estimated to be 153 deaths per 100,000 live births, 95% • confidence interval (95, 211). For derivation and any further detail, the reader is encouraged • to consult Hanley et al., 1996 (ref. 8). • • 67 • • • , . • • To conclude, emphasize strengths and weaknesses of the sisterhood method of estimating maternal mortality. While it is a quick, efficient, and easy to use method, the • estimates produced are retrospective. On average, the lag time from the period of data • collection to the time for which the estimate is calculated is 10 years. As a result, the • method is not effective for testing program impact and may not reflect very recent changes in ,. maternal mortality. Where researchers need a local estimate of maternal mortality to help steer assessment of local need, however, the sisterhood method of estimating maternal • mortality is economical and informative. Those who wish to conduct a sisterhood study • should read references 1, 2, and 8 . •

! • • , .• • , .• • • • • • • i • • • • 68 , .• • • • 2. MORBIDITY SURVEYS • A. llEALm INTERVIEWS AS PART OF LoNG-TERM DEMOGRAPIDC SURVEILLANCE • Field surveys to monitor the incidence of non-fatal health events generally require surveys at • much more frequent intervals than for deaths, especially if the incidents are minor illnesses that tend to be rapidly forgotten. 1 For this reason, morbidity surveillance is arduous and • costly. However, for certain chronic health conditions such as heart disease or addiction to • tobacco, interviews with a proxy family member can be used to obtain data on all members 1 • of the household. Many chronic disabilities and handicaps can also be studied in a single round survey. For acute health conditions that require interviews at frequent intervals, • morbidity surveillance may be most feasible in the context of long-term demographic • surveillance. • Study Population • The following questions need to be answered. Who is included in the study? What will be • the denominator? How will you define the person-time at risk? For example, do they need • to be present at the time of visit? Will you have an ongoing recording of migration? What diseases are to be investigated? • It is generally better to ask simple questions related to all common illnesses occurring • in the area (except minor ailments such as small sores and runny noses) rather than considering just one condition. • Monitoring • The frequency and location of monitoring need to be considered. How frequently you have to monitor depends on the incidence and duration of episodes, as well as financial • constraints. Often, recall of symptoms and signs is short lived. • Should monitoring be done at the village level, and/or at the primary health care level and/or at secondary and tertiary levels of health care? • More complete information on incidence will be available in village-based • surveillance, but the data will be crude. Health workers at health institutions will provide a • more accurate diagnosis and more detailed investigations might be possible (e.g., blood culture, urine samples, chest x-ray). If investigating a serious illness for which people are • likely to seek treatment, then surveillance at a health centre/hospital might be sufficient (e.g., • typhoid). • • 69 • • • • • • Case Definition ,.• The investigator should establish a clear definition of a case. The more specific the case definition, the smaller the size of sample that will be required. It is generally useful to have • several outcomes. Severity of conditions to be investigated and criteria for assigning cases needs to be decided upon, such as for pneumonia and severe pneumonia. Certainty of ,.• diagnosis can be improved with blood tests and x-rays . • • • EXAMPLE OF CASE DEFINITIONS FOR MORBIDITY SURVEY OF ACCJTE l..oWER • REsPIRATORY INFECTIONS (ALRI) :•• Case definition & severity are based mainly on symptoms & signs: • • mild ALRI - cough + evidence of fast breathing • moderate ALRI - cough + fast breathing + chest indrawing • • severe ALRI - as for moderate with evidence of heart failure or cyanosis or • difficulty feeding • + Test-based confinnation: • • blood-culture-proven pneumonia • • x-ray-proven pneumonia ,.• Confounding diagnosis confirmed by signs + blood test: malaria - axillary temperature > 37 .5oC + P. falciparum parasite density • > 20,000/µL2

! • • , .• • • • 70 • • • • • • • • • • • • • • • • • • • • • 71 • • • • • • B. llEALm SURVEYS USING HOSPITAL RECORDS , .• • • • In discussing this topic, you may want to provide samples of clinic registers, hospital l. in-patient admission and/or discharge registers and summaries, and anonymous copies of • good- and poor-quality complete records to familiarize students with medical records. Also, • this could give ideas to health care providers on how to improve their institutional records . • Data collected in hospitals and health centres can be useful in health surveys and are • inexpensive data sources, compared with the expense of doing special surveys. However, • the quality of information can be poor unless the hospital staff have a special interest in health records and data collection. Also, data might be relatively complete for an illness or ,.• injury, but limited for personal, activity, equipment and/or environmental risk factors that • contributed to the illness or injury . • In using these data to estimate the incidence or prevalence and relative importance of various health conditions, you must be able to identify possible problems that could lead to • biases, such as: • • incomplete or invalid recording of patient information • systematic coding errors • • near-immediate death in drownings or other unintentional injuries, or suicides where • the victims do not survive long enough to reach a health facility and are not included • in official statistics ,. • if certification of deaths in villages is limited, victims may not be counted in any data • source • You may find that data collected prospectively in collaboration with health • professionals who are interested in disease prevention provide the most useful source of • information. All pertinent details should be recorded. Final discharge diagnosis and contributory actors must be recorded on hospital discharge forms by the attending staff, or • presumptive diagnoses must be recorded on out-patient registers . • • • 72 ••• • • The following is an example of how a hospital physician used hospital records to study various types of snake bites in Brazil. • Health Surveys using Records from Hospitals and other Health Facilities • Analyses of case series are often viewed as second-class epidemiologic studies as sicker individuals are usually over-represented in them, there are no controls, and they do not • represent an unbiased sample of the disease frequency in the population. However, hospitals • and other heath facilities can provide a useful and unique source of data that allows the • investigator to rapidly and inexpensively get an idea of the dimension and characteristics of • specific diseases or other health events, particularly in developing countries. • We will look at some surveys carried out in a Brazilian teaching hospital on snake • bite, to study the advantages and limitations of using hospital records as a data source. • Snake bite has been a relatively unexplored field for • clinical and epidemiologic research, although it is an • important health problem in certain areas. Physicians at • the teaching hospital frequently saw victims of snake bite • in the emergency room. They knew that some of the patients had complications and were hospitalized for long • periods, or even died. However, they did not know how • many cases, complications, and deaths there were, who the patients were, or where they • came from. • A retrospective survey on snake bite seen in the hospital during the previous years • provided answers to the questions above and was also useful in: • • allowing identification of risk factors for complications • assessment of prehospital care (e.g., use of tourniquets) • • assessment of performance of the physicians (e.g., correctness of diagnoses, time • from arrival to hospital and antivenom administration) • • identification of "new" problems (e.g., bites by nonvenomous snakes, "dry bites") • Analysis of data collected in this survey clearly indicated that better training of physicians • and medical students in the management of snake bite was necessary, and useful information • was obtained on stocking of antivenom. Also, several papers were published that have • proved useful to physicians working elsewhere in Brazil, and further work on the same and • other subjects was initiated. • Advantages of Case Series • Generally speaking, case series raise hypotheses for further observational studies and clinical trials, but can provide valid information by themselves. One advantage of retrospective • 73 • • • • • • studies is that information on many years can be gathered in a short time. This is particularly useful for health events that are not notifiable or on which available data is of • dubious quality. And, as the data can be collected quickly, short-term results are available to • investigators for publication, which motivates the investigators to continue research. For • certain events like snake bite, data from particular areas may be more interesting than national figures because national figures tend to lump different things together. Data • abstracted from charts allow the study of morbidity, mortality, and costs. Through these • studies previously unsuspected findings may be recognized. For example, a large proportion • of snake bites were by nonvenomous species and some of them were being erroneously • treated with antivenom. Another advantage is the low cost of case series . • Limitations of Case Series • A major limitation of these studies is that one does not know directly how representative the data are of a health event in the community, as the data can be biased towards more severe • cases, to non-fatal cases, or to individuals who have easier access to the facility. To assess • whether the data are representative, additional information must be collected from other • sources, such as household surveys. The calculation of rates can be biased due to problems • in correctly identifying both numerators and denominators . • Identifying a health event is not always straightforward: in the data analysis of snake {.• bites, several mistakes in recording ICD codes were detected, and several additional cases were included from consultation of log books of the several clinics where such patients could have been seen or admitted. As the routine recording of information in the charts does not • follow a protocol, it is almost certain that data will be missing, and it is very likely that the • severe cases will have better recorded data, which may bias any analysis of risk factors . Although the catchment area of the hospital is usually used as a denominator for the • rate, often the size and boundaries of this catchment area are unknown or poorly defined. • Authoriz.ation for using the data may be an issue in certain settings, but often authorization • can be obtained on the spot just by talking to local physicians and health administrators . • Conclusion • Observational studies and clinical trials are methodologically superior to case series • studies, but are often too expensive and time consuming. Although the snake bite survey took place in a teaching hospital, which is supposed to have better than average recorded • data, the idea of using descriptive studies can be extended to other settings. When the • quality of recorded data is too poor to be useful, efforts must be made to educate physicians • and other health care providers in registering information so that it can be analyzed in the • future. Although case series have limitations, they should be encouraged rather than • discouraged, as a useful tool for collecting data in developing countries . • • 74 • • • • 3. DEMOGRAPIDC MEmons • Demography is an important component of • epidemiologic research and essential to give an overall • picture of the health status of a community. • Demographic surveillance (or at least one census) is necessary to determine denominators for incidence • rates. Demography is also needed when evaluating • interventions, such as the effect of vaccines on • mortality or family planning on fertility. Demographic • surveillance usually begins with a census. • Without incidence rates, it is impossible to • compare the severity of various health problems • between communities, regions, and countries with different population sizes. In order to calculate incidence rates per unit of population for deaths, hospitalizations, and various health • conditions, it is necessary to estimate the number of health events of interest (numerators) • during a specified period of time, as well as population denominators. • Suneys for &timating the Number of Health Events and the Population • at Risk of these Events • Demographic surveys include censuses and surveys to determine the size and composition of • populations, as well as the vital registration system to determine change. Vital events such as births and deaths are not registered with any degree of completeness in many developing • countries, hence the need for retrospective mortality surveys and in-depth interviews such as • verbal autopsies to assess the major causes of death. Surveys of births and deaths can be • done once in sin~le-round surveys or at periodic intervals in multiple-round surveys. • De Facto and De Jure Systems for enumerating populations • de facto system: a person is counted wherever he or she is found at the time of census enumeration. The main advantage of the de facto system is that there is less chance of • double counting or omission of persons. It is more commonly used than the de jure system • and is recommended by the United Nations Population Commission. • de jure system: people are enumerated at their place of usual residence, irrespective of where they were at the time of the census. The main advantage of the de jure system is that • it provides a profile of the permanent population. • • 75 • • • • • • ISSUES AND EXAMPLES OF DEMOGRAPIIlC METHODS IN EPIDEMIOLOGIC REsEARCH • • Demography is an important component of epidemiologic research and essential to • give an overall picture of the health status of a community . • Demographic surveillance (or at least one census) is necessary to determine • denominators for incidence rates . • • Demography is necessary when evaluating interventions: e.g., - effect of vaccines on mortality • - effect of family planning on fertility • • Demographic surveillance usually begins with a census: - Define population to be included: • (a) people residing permanently in the area, or • people claiming affiliation to the clan • (b) total population, or • only children, or • children + new births • - Obtain information from any previous census or other records (e.g., baptismal • records) that will assist in determining age as well as family groups . ,.• - Have to consider how you are going to link all information collected subsequently on individuals (e.g., an identity number used on all forms and remaining the same • throughout life) . • • What information should be collected on a regular basis? • Pregnancy? Births? • Marriage/Divorce? • Deaths? • Migration? Essential to determine person-time risk Date of entry into study? • Date of exit from study? • Residency - village, house? • Sample size and frequency of monitoring will be dependent on whether you are • investigating primarily mortality or morbidity, and the incidence/prevalence of the • disease(s) being investigated . • • • 76 • • • • Age Standardization of Population-Based Rates • Age standardization of incidence rates may be necessary for comparisons of rates between • two or more populations if one population is much younger or older than the other. This is • because a significantly greater proportion of one population may be at risk of certain health • conditions that are more frequent in certain age subgroups, such as infants, toddlers, young adults, or the elderly. Thus, standardization is particularly important when comparing • incidence rates between a developed and developing country, or between rural aboriginal and • urban non-aboriginal populations in an industrialized country. • Standardization refers to the use of a standard population structure as a basis for • comparison of rates between countries with different age structures. Thus, for example, the • incidence rates in one country can be applied to the population structure in another country, • or the rates in several countries can be applied to a standard population such as the World Standard Population. This allows comparison of the expected number of deaths in the • standard population as estimated using the age-specific rates in various countries. Use of a • standard population adjusts for the factor of age differences and makes it feasible to discern • whether there were variables other than age responsible for observed differences in rates of • various health conditions. • Standard Populations • A larger population can be used as a standard when comparing rates of health events in a smaller population with the larger one. There are also various standard populations that • represent an average of many or all countries in a region of the world, including the World • Standard Population and others that can be found in publications such as the World Health • Statistics Annuals, which is published by the World Health Organization. The World • Standard Population is heavily weighted by China and is older than the populations in many developing countries with less effective family planning programs. The Pan American • Health Organization has used a standard population based upon the entire population of Latin • America. • Methods of Age Standardization • The most frequent methods of age standardization include the direct and indirect methods. • Direct standardization can be used only if the number of deaths in different age groups is available so that age-specific death rates can be calculated for the study population. Indirect • standardization is used when only the total number of deaths in the study population is • available, rather than the numbers in specific age groups, together with census data on the • population in different age groups. Only the concept of standardization is discussed here, • and details for calculations and further practical exercises can be found in standard texts. • • 77 • • • • • • • , .• • • • , .• • • • • • • • • • • • 78 • • • • Handout • • DIRECT STANDARDIZATION OF DEAm RATES FOR AGE • Standard Maori Non-Maori population • Age (N.Z. population Age-specific Expected Age-specific Expected • 1966) death rates deaths death rates deaths 1966 cols (2)x(3) 1966 cols (2)x(S) • (1) (2) (3) (4) (5) (6) • 0- 4 306,643 7.45 2,284 3.92 1,202 • 5-14 565,756 0.91 515 0.39 221 • 15-24 436,019 1.53 667 0.98 427 • 25-44 640,711 3.52 2,255 1.63 1,044 • 45-64 504,697 20.92 10,558 10.04 5,067 • 65+ 223,093 93.52 20,864 68.50 15,282 • Total 2,676,919 37,143 23,243 • Standardized death rate 13.88 8.68 • Crude death rate (for comparison) 6.37 9.07 • Source: New Zealand Year Book, 1971 • • • • • • • 79 • • • • • • Handout • • INDIRECT STANDARDIZATION OF DEAm RATES FOR AGE • Standard Maori Non-Maori Age rates (N.Z. • death rates Population Expected Population Expected • 1966) 1966 Deaths 1966 Deaths • (2) x (3) (2) x (5) • (1) (2) (3) (4) (5) (6) • 0- 4 4.37 39,539 173 267,104 1,167 • 5-14 0.45 61,728 28 504,028 227 • 15-24 1.02 34,725 35 401,294 409 • 25-44 1.76 43,686 77 597,025 1,051 • 45-64 10.44 17,626 184 487,071 5,085 • 65+ 68.74 3,855 265 219,238 15,070 • Total (8.86) 201,159 762 2,475,760 23,009 Total deaths actually registered 1,291 22,487 • Standardized mortality ratio (SMR): Actual deaths 1.694 .977 • Expected deaths • Indirect standardized death rate • (SMR xCDR) 15.01 8.66 • *Crude death rate (CDR) • Source: New Zealand Year Book 1971 • • ••• • • • 80 • • NUMBER OF DEAms/PERSON-:YEARS AT RISK BY AGE, AGE-SPECIFIED MORTALITY RATES AND PROPORTION OF PERSON-YEARS IN EACH AGE GROUP IN GoROKA TOWN, ASARO, TOTAL SURVEll..LANCE AREA .

Age Goroka Town Asaro Census Division Total Surveillance Area

Died/ Died/ Died/ Person- Rate/1000 Pyrs% Person- Rate/1000 Pyrs% Person- Rate/1000 Pyrs% Years Years Years

0 13/411 31.6 4.3 721766 94.0 2.5 112/1745 64.2 2.9

1-4 13/1497 8.7 15.8 33/3000 11.0 9.6 59/6413 9.2 10.7 00 - 5-14 1/2719 0.4 28.6 14/9000 1.6 28.8 23/16000 1.4 26.7 15-29 1/2973 0.3 31.3 2317667 3.0 24.5 34/14800 2.3 24.7 30-44 511653 3.0 17.4 25/5833 4.3 18.7 56/11148 5.0 18.6

45-59 31255 11.8 2.7 117/3852 30.4 12.3 164/7722 21.2 12.9

60+ 1122 45.5 0.2 84/1125 74.7 3.6 134/2186 61.3 3.6

Total 3119530 3.9 100.0 368/31243 11.8 100.0 582/60000 9.7 100.0

Source: D. Lehmann. Papua New Guinea Institute of Medical Research, Ooroka

r.... ••••••••••••••••••••••••••••••••••••••••••• r ! • ! • i • I • AGE Es'I'IMATION t • It is essential to have an estimate of the age of individuals in a population before age-specific or age-adjusted incidence rates can be computed. The age of many persons in developing ,.• countries is unknown. Several methods for estimating age include: • • estimation by observation • • historical calendar of events • • method of matching • • age-event method . • Estimation by Observation ;.• This method is based on observed physical characteristics of growth and aging, as well as the history of certain life events such as pregnancies. Physical criteria include: (i) for young • children, the state of tooth eruption, the ability to walk, and the feeding method (whether the • child has been weaned); (ii) for adults, whether one has white hair. For women, the history of the number of children ever born can be used. The major limitations of the method are j.• that the use of physical criteria can be quite difficult in its application, even with medical investigation, and the method may not provide a precise age because the physical appearance • may not be a good proxy for age. See the example on page 90 . , .• • Historical Calender of Events • For this method, an inventory of major events in the community is prepared or obtained prior to undertaking the census or survey. The nature of the events can be quite varied, .,.• including political, cultural, economic, religious, administrative, ecological, and others. The events must have precise significance and must not lead to any confusion in people's minds . • Their use makes it possible to situate an individual's date of birth within an interval of time . This interval can be narrowed with the introduction of more and more information about • events that have occurred in the individual's community around the time of his/her birth. • Limitations of the method include that it applies only to residents of the study area and to • persons who can be interviewed or about whom detailed information can be obtained from • other members of the household. However, these limitations apply to most techniques of age ,.• estimation. • Use of Official Documents This method consists of using official documents to obtain the age of individuals who are ,.• either unable to read or communicate their age spontaneously, or who are absent from the household when the interviewer visits. The other household members must have access to • one or more of the individual's official documents, such as a national identity card, birth • 82 • • • • certificate, family card, health card, immunization card, political card, religious/baptism card, or maternity card. This method is also useful as a means of validating self-reported • age, since self-reported age can be over- or under-estimated because of rounding and other • errors. A limitation of the method is that can be used only for members of the population • who have the necessary documents, who represent a small proportion of the total in most • countries with problems of age reporting. • Method of Matching • This method is based on the fact that in most rural areas, almost all individuals know each other and can thus situate each person compared to the other in terms of who was born first. • This ordering of individuals in the community from the youngest of the group to the oldest • can then be translated in terms of age. The basis for ordering is the known date(s) of birth of • some individuals, which are considered accurate either on the basis of official documents • such as a birth certificate or because the individual is educated and knows his/her age precisely. In addition to problems associated with individuals who are not seen and non­ • residents of the locality, there are two other major problems with matching. One is • psychological and the other has to do with data collection. The process of gathering people to • order them in terms of age slows data collection and lengthens the time in the field. The delays often lead to refusals and undercounting, which is a particularly serious problem since • the quality of matching will be a function of the number of people gathered. The greater the • proportion of the population that can be gathered, the smaller will be the errors of • estimation. The method of matching is particularly suitable for small samples. • Age-Event Method • The age-event method begins with preparation of individual biographies, including various • life events such as first schooling, first migration, first use of health service, first marriage, first birth, first immunization against measles, etc. These events are then linked together to • evaluate their occurrence consistent with the age of the individual at that time. The age-event • method was developed in the 1980s within the framework of the Demographic and Health • Surveys (DHS) Program in order to improve the quality of age estimation in most developing countries. Poor age reporting had been experienced in the previous World Fertility Surveys • (WFS) Program which started in the 1970s and continued to the mid-80s. The age-event • method is now used extensively in most surveys for which age is a key variable. • • • • • 83 • • • ••••••••••••••••••••••••••••••••••••••••••• USING A llisToRIAL CALENDAR OF EVENTS

...... '. - \ ' ..

Two men from the Goroka valley who remember seeing the comets of 1910. • • • • • • • • • • • • • • • • • • • 85 • • • • • • Handout • CALENDAR OF EvENTS FOR ASARO CENSUS IN PAPUA NEW GUINEA 1932 Danny Leahy and Jim Taylor arrive • 1933 First plane flies over Goroka valley • 1935 The missionary Buko settles at Naminaloka. He was brought there by Georg Hofmann (also called Chuma) and Johannes Flierl ,.• 1937 Missionaries settle at Asaroka, Lapeigu and Gululumba 1939 First airstrip is built on the site of the current business centre of • Goroka and is called Seigu airstrip 1941 A new airstrip is built at Humilaveka, the site of the current Goroka • Teacher's College • 1943 Japanese bomb the Goroka patrol post and the Asaroka mission station • Present-day airstrip at Goroka is built with 1000 Chimbu labourers . 1944 Construction of main road from Goroka to Asoraka High School. • Old hospital was built at Asaluifa • 1947 Papa Kuso built SDA church at Kabiufa 1948 Money is first used in Goroka • Kabuifa high school was built ,.• Work at Asaro Coffee began 1951 The first mass baptism is held at Asaroka for mission helpers and their families . • 1953 Okiufa community School is opened • 1961 Gefamo Catholic Mission is established 1962 Total eclipse of the sun • 1964 First General Election • 1973 PNG self government • 1975 Independence • 1987 Jim Taylor's death • • INFoRMATION SOURCES TO ASSIST IN DOCUMENTING OR ESTIMATING AGE • It can be difficult to estimate age accurately, especially among children in areas with a high degree of • malnutrition. The following examples can be used as a guide for possible sources of information for • age estimation . • Written sources: Birth register, health book, baptismal records . • Other sources: Grave yards (for reference ages), key informants, photographs (teeth, eye-balling), ,. school attendance (year began, grade attained, and when completed), social stages (initiation, mothers • in birth hut together), mean age at menarche, mean age at marriage, mean age of first-born child . • • 86 • • • • Handout • CHARACTERISTICS OF POPULATION STUDIES JN TARI AND ASARO, PAPUA NEW GUINEA • TARI ASARO • Site Rural Urban, peri-urban, rural • People in study All who claim All who live in area affiliation • Migration out High Low • Total population 28,000 7000 • Population 3800 900 • < 5 years • Surveillance* Mortality (morbidity) Morbidity (mortality) • • I I TARI I ASARO I • TIME PERIOD 1981+ 1979-85 1986+ • Frequency monitoring monthly fortnightly weekly • (fortnightly for a • sample of • children) • No. reporters 27 16 28 • Sex reporters Male Male Mand F • No. supervisors 3 2 4 • No. reporters/ 9 8 7 • supervisor • Average popn/reporter 1037 400 216 • Average popn/ 61 65 54 • reporter/day • Average popn 8(16) 8 6 • < 5 yrs/reporter/day • *The Tari study is mainly mortality and the Asaro study is mainly morbidity. • • 87 • • • • • Handout • PNG INSTITUTE OF MEDICAL RESEARCH-ASARO (Ceo 92/3) • IDNO: Survey Date: • Area: • Village: ------• Hamlet: • House: • Christian name: • Name 2: • Other names: • Sex: Date of Birth: • Age: Source of age/birth info (D G A): • Present in village last night? (YIN): Asples or Settler? (AJS) ••• Is this the usual residence? (YIN) • If NOT usual residence, fill in information below. • Usual area: • Usual village: • Usual hamlet: • • Father Mother • Name 1: Name 1: • Name 2: Name 2: • Residence Residence • Area: Area: • Village: Village: • • Interviewer: ---Coded by: ---First entry: ---- 2nd: ------• • 88 • REPORT BOOK FOR DEMOGRAPmC SURVEil..LANCE

lctentllY • S•atu Ila .DOI ·A I c D ll p 0 ff I I re L M ,. 0 p Q R s T. u v w x y z

I .... ,. .... " -·-·"' .. • t ..

00 \0 ...... ' • '!I ...... _ "' y v ., ~ v " "' 11 ,,, v v v v v L' rA' "' k , , I\ , .' ' I r ...... -- ·---. --· - .. -:...... ·------_ • t.. v p p p p ·p t.:S v V v v .V ./ v f l 1 I ) p p 1.'E I t t ~ .....~..__ ...... ______-.. .+-1..M. _.... .~_.;~--J.-1. -~ .... 1-.-. '"'. _+::-_+. ~'1+,....i:;. .• +v~ .... ~ ....~r(~;.:,.+--...,7. f-"":"! rv+.v±.,,~v'-:t-.,,,-:t-vtt.:-r.r-r.-1-r.-.,-r.-, -r.-, -r-, r-, -r:-,I.iii I ~ - -··I o~I .. I \ • i . ~.~ ·Iii_·- • i ...... -·· ':.·•· .. -...... ,...... v .,. v v v ti ti ·v - - v. v v \/ v , ) J I I ) I I I - ' I .. AAAlf " , :...... v ., v t/ LI I/ V' ti v " . -· - v v' v v r .f I r I I \ ...... - ·" ~ I/ t/ ...... lo•••• V' v II v v t/ I/ v v v V' " ' J I . I' I ,' ...... -· I ' ( ) . \ v t/ .,., v \I ,, ,/ r/ ti vi "' I ··- ...... " v .I ..J I I t I r r Ubl 2;;r - I ' r I l6101o1 &, 'f• N tr--f\TA " '" •• 1 ... ~ ~K lof t.I f' :/ v 1), i Q " AT ~- • Cf l.. tT - r o..~' ~· 1-l-:f.' Uli 1 'l.I " l't\ rf·ld~ I G ~ ~II ~T.,~ I l \ l ...... rg ..... ••••••••••••••••••••••••••••••••••••••••••• • • • AGE ESTIMATION OF NEW GUINEAN • CHILDREN • AT the annual Symposium of the Medical given to the meeting that ages tend to be Society of Papua and New Guinea. held under-estimated in New Guinea, most members • at Madang in August, 1967, Dr. L. Malcolm guessed an age well below the known age of presented a paper on " Some Aspects of each subject (as recorded in the local Catholic • Growth and Development of Children in the Mission Baptismal Register). Territory of Papua and New Guinea". • Dr. Malcolm pointed out that Bundi people I To illustrate his paper, he presented to the may be relatively retarded in child growth and • meeting a group of five children and adoles­ development compared with other populations cents from the Bundi area of the New Guinea in Papua and New Guinea, but the evidence I • mainland (Plate I), and invited his medical which is available from other parts suggests • audience to estimate their ages to the nearest that underestimation of age is very common . year. The results are recorded in Table 1. In For example, in the Chimbu valley, the mean • only one case (that of Edward) was the mean age of a class, as estimated by its teacher, was I estimation of the observer group accurate. In four years young.er than the mean value • spite of the warning which had previously been • obtained from the Missions Register. The implications for both education and • health workers of a tendency to underestimate ag.e are obvious. We may hope that this • example may stimulate a general revaluation of personal criteria for the estimation of age • throughout Papua and New Guinea by all who • are called upon to make such estimations . Table 1.-Medical Audience Escimacions of Age foe • Five Bundi Subjeccs.

• d ~ E " ::: 3: .. = ..,,~ i~ - ., ., • <.;: - I ..., I -' r.i • 2 1 3 10 ... • 4 10 11 5 8 9 • 6 1 7 7 1 • 8 2 9 8 • 10 11 11 4 12 2 • 13 3 14 5 3. • 15 10 T 16 8 11 • 17 4 1 18 3 6 • 19 3 Ho. • participating . 31 35 34 35 37 • mean Front Row : Kriv.·a, Bruno, John. estimate: 4.0 4.4 10.4 15.2 16.4 • Rear Rou:: Ed";';::,rd, Dr~ 'line:;, Ludwig, adult • Bundi male . TRUE AGE: 7.0 7.4 15.3 11.3 16.7 • l22 PAPUA AND NEW GUINEA MEDICAL ]OUR.NIU • • 90 • • • 4. STATISTICAL METHODS • SAMPLE SURVEYS • "Statistical inference is about the individuals who were not in your sample" • - Dr Jim Hanley, McGill University, 1995 • Following is a brief review of sampling theory. We provide an example of cluster sampling, • which is often used in developing areas and not always covered in introductory courses in • statistics. • Reasons to use a sample survey rather than a total population survey such as a census are: • • reduced cost and time • • more attention can be paid to ascertainment and quality of measurements • • in probability samples, the reliability of the sample estimates can be measured from the sample itself • • the high precision of a census is not needed (however, a census may actually be less • accurate than a sample survey in some situations!) • Types of Samples • Samples can be probability or non-probability. In probability samples, each individual has a • known probability of being selected and inferences can be made from the sample to the entire • population being studied with an estimated degree of certainty. In non-probability samples, inferences cannot be made to individuals who were not in the sample, since it is not known • whether the individuals who were surveyed were representative of the population being • studied. • Types of probability samples include: • • simple random sample • • systematic random sample • stratified random sample • • ratio estimates from stratified random samples • • single-stage cluster sample • • multi-stage sample • Steps in a Sample Survey • When planning a sample survey, consider: • • choice of target population (and subpopulation) • choice of information needed • • sample design • - develop sampling frame • 91 • • • • • - select units and subunits - construct estimators for variables being studied • - project uncertainty of estimators - pilot study to gauge variability • - confidence intervals if descriptive • - confidence intervals/power if comparisons being made • pretest • • organize field work • • data collection and processing • • data analysis • • estimates and uncertainty projections • Efficient Size of Samples: • Even if one country is 10 times larger than another, surveys in both are often done with the same sample size of about 1000. This is confusing for many people, since intuitively it • would seem necessary to use a larger sample for a larger population. However, while the • absolute sampling error may be 10 times larger for the larger population (e.g., an error of • 1 million for the large population versus 100,000 for the smaller one), the standard error or relative uncertainty will be the same for a large or small population (e.g., 5 % of the • population). It would be wasteful of resources to take very large samples from large • populations. Analogously, blood samples are taken to measure concentrations and not • absolute amounts. Thus, bigger blood samples are not taken from bigger persons . Although the choice of an efficient sample size is relatively unaffected by the size of • the population being studied, optimal sample size is affected by the primary purpose of the • study, which may be one or all of the following: i • • to provide a precise estimate of a variable for an entire population ( 1 answer) ,.• • to provide a precise estimate of a variable for each of several subpopulations (1 answer per subgroup) • • to compare estimates of a variable among each of several subpopulations ( 1 answer ,•,. per comparison) ,• Limitations of Sample Surveys • If there is an in-built bias in the measurements or in a comparison, increasing the sample size • will not make it go away. Two examples include: • Q. "What colour are your stools?" • A. "Well, the one in the kitchen is red and the one in the living room is green." Q. "Are you bilingual in the two official languages?" - 15% said yes • Q. "Can you keep up a conversation in the other official language for 15 minutes?" - 11 % said yes. j • • • 92 • • • CLUSTER SURVEYS • Conducting Cluster Surveys in Developing Countries • Cluster sampling: how to get something for (almost) nothing • Random Sample Cluster Sample • x xx x x xxxxxx xxx xxx xxx • x xx xx x xxx xxx xxx • x x x x x x xxx x xx x xx x x x xxx xxx xxx xxx • x xx x x xxx xxx • • The efficiency of cluster sampling in population surveys - defined conceptually as clumping respondents geographically for the ease of access - depends on the heterogeneity of • the data within the clusters. In brief, if clusters are likely to be homogenous with regard to· • the characteristic being measured (lumpy data), cluster sampling can be of little benefit. If, • however, the characteristic of interest is spread heterogeneously throughout the population, designing a survey which incorporates cluster sampling will increase the effective sample size • for a given cost, and is therefore desirable. • • Lumpy Data Scattered Data • cluster sampling NOT recommended cluster sampling recommended • • 000- -o- 0000 --0 ~ --00 -0-0 oo- • 0000 0-00 o- oo- o-o- -00 0-0- 0000 o- -000 --0 -0-0 0-0 --0 • -oo- • --0 0-00 -o- -oo- 0-0- --0 o-o- • o- 0000 --00 oo- 00-0 -00 • 0- 000- -o- -000 -00 -0-0 • 0000 --0 0000 --0 o- 0-0 -000 • 0000 0-00 0-00 o-o- -00 -0-0 • 00-0 000- -000 -000 0-0 o-o- • • • 93 • • • • How is Cluster Sampling Done? 1 2 • In a simple random sample of a population, • a sampling frame which lists individuals, such • as a voters list or telephone directory, is sampled randomly. The sample may be scattered 1 2 • widely in geographic terms. In stratified sampling, • the population is divided into groups, • such as age groups, and a random selection of individuals from each strata are sampled . • Cluster sampling differs from random sampling in that the sampling target is larger • than the individual, and is ideal in situations where lists of individuals (sampling frames) are • unavailable. The World Health Organization has used cluster sampling extensively to conduct vaccination prevalence surveys in rural areas of developing countries. Instead of • attempting to visit 96 individuals all at different sites (the sample size necessary to estimate • vaccination prevalence with standard error of 5%), the typical cluster survey may visit 210 • individuals, consisting of seven individuals at each of 30 different cluster sites. 3 In this case, • the clusters are determined randomly, not the individuals . • Cost of Cluster Sampling • If the seven individuals in each cluster of the WHO survey described here are all very much i. alike, the survey has gained little from choosing seven instead of one at each cluster site. This is the example of "lumpy" or homogenous data, such as might be expected of rubella • immunity. The lumpiness of data is referred to as the design effect. In practical terms, we • can use the design effect to mean "how much larger does my sample size need to be in order to counterbalance the effect of a cluster design? The design effect can be assessed in the • data after collection using statistical methods. 3 Where data were lumpy, the precision of the survey estimate is lower, and this is reflected in wider confidence intervals. , .• ,•• Benefit of Cluster Sampling ,. In brief, cluster sampling is much easier to do in the field. It decreases the cost of organizing a field survey and collecting data. Moreover, if the data are not lumpy and the • design effect is low, the clustered sample will have almost the same descriptive power as a • random sample, and the study design will be highly cost-effective . • , .• , .• • ••• • • 94 • • • SAMPLING METHODOLOGY: AN URBAN SURVEY IN PAKISTAN • In a difficult maternal health survey in urban Pakistan, no sampling frame for the • population of interest was available. A strategy which involved sampling by strata and • by cluster was designed to collect information about the catchment population of three • health facilities. • 1. The addresses (or neighbourhood, where addresses were imprecise) of the 500 • most recent users of each of the three facilities were plotted on neighbourhood • and city maps. 2. A geographical study boundary was drawn to contain 75 % of the users. • 3. 75 survey sites were allocated among neighbourhoods, in proportion to the • number of users. • 4. These survey site starting points were plotted randomly on local maps (to the nearest road of any size) using a computer generated random number grid. • 5. From the starting points, a further random selection procedure was used to • avoid street-front bias in the households chosen. On foot, a clockwise circle • was paced using all navigable roads, alleys, and footpaths. A random number • less than the number of steps paced was used to define the index household for each survey site. • 6. From the index household, the 40 nearest households (as paced by foot in all • directions) were included in the survey sample. In the event that the starting • point contained an apartment dwelling, all flats on a single level chosen randomly were included. The 40 nearest households were used because of the • need to provide maximum geographical concentration and security for the • female survey workers. • After analyzing the survey data, the results showed that this combination of stratified and cluster design was effective in collecting "non-lumpy" data, and no • adjustment of confidence intervals was necessary. • • • SUMMARY • Any small or large scale health survey in a developing country will benefit from careful • design, including an early consultation with a biostatistician to determine an appropriate • sampling strategy. Cluster sampling may be appropriate and can be highly cost-effective. • • • 95 • • • , .• SAMPLING METHODOLOGY: RATIONALE BEIDND SAMPLING FOR DEMOGRAPIDC, ,.• MORBIDITY, AND MORTALITY SURVEILLANCE IN THE AsARO VALLEY OF PAPUA NEW GUINEA:

• Factors that might influence the incidence and severity of acute lower respiratory illness • (ALRI): • • altitude • • social and economic status • • nutritional status • distance from health services • • crowding • • type of housing • • recent migration • • size and isolation of clusters • Nearly all villages were situated between 1500 and 1900 metres altitude . Coffee, the major cash crop, was evenly distributed through the valley . • Analysis of clinic records did not reveal any spatial clustering of children with malnutrition. • Housing in rural areas appeared to be homogeneous . In order to take into account access to health services, socioeconomic status and crowding, • the population was stratified into urban (Goroka town) and rural areas . The urban area was subsequently divided in three distinct areas based on housing type (high, • medium, and low cost). ,. The rural area was further stratified by Census Division (Lowa or Asaro at varying distances • from Goroka Base Hospital) and by village population size . • It was logistically appropriate to sample by village or section of town, and also of interest to • look at morbidity in relation to size of cluster . • Sampling procedure in rural area: Within each of the two census divisions, villages were • stratified by total population size ( < 200, 200-400, and > 400 people into five strata. From 1978 census data, the number of children in each village was known. Each village was • considered as a cluster and the cluster sampled with a probability of being chosen proportional • to the number of children in the village. Clusters were sampled until there were 200 children • included in each of the five strata . • • Adapted from Coakley K, Lehmann D, Smith D, Riley I, Howard P, et al. The Asaro Valley Surveillance Unit of the Papua New Guinea Institute of Medical Research: Methodology, Demography and Mortality. Papua New • Guinea Institute of Medical Research, Goroka, September 1993 . • • 96 • • •

REcOMMENDATIONS FOR PLANNING HEALTH SURVEYS IN DEVEWPING COUNTRIES: • A CALENDAR OF PRIORITIES • Eighteen to six months prior: • • determine priority of research questions with host organizations • • meet with community organizations/leaders • clarify study question: is it answerable using quantitative methods? • • simplify quantitative goals to a very few simple questions • • specify target population • • translate research question into study design • do sample size and power calculations - consult a biostatistician • • write protocol • • obtain ethics committee approval • • obtain census/electoral registers if available • obtain copies of other questionnaires used in other survey work • • walk or drive through as much of the study area as possible • • plan sampling strategy • Six to four months prior: • • design together: - questionnaire • - data entry program • - database structure - data editing procedures • - plan for analysis • • consider measurement issues; socioeconomic status, literacy • • have questionnaire translated, back-translated • pilot test questionnaire including test re-test reliability • • hire local staff • • safety considerations, including documents and photo identification • determine sample • Date collection stage: • • adequate interviewer training - budget one full week • • daily supervision • • daily data entry and editing - immediate feedback on data quality to field teams • • checks to do yourself: - digit heaping • - interviewer bias • - comprehension • • • 97 • • • • • Further Information • For further information on verbal autopsies, P. Barss or D. Lehmann; on sisterhood method • and cluster sampling, C. Hagen or J. Hanley; for health interviews, D. Lehmann; hospital • surveys, S. Nishioka; for age estimations and population estimates, B. Kuate Defoor D . • Lehmann . • REFERENCES - Mortality Surveys: Verbal Autopsies • 1. Kalter H.D., Gray R.H., Black R.E., Gultiano S.A. Validation of postmortem • interviews to ascertain selected causes of death in children. Int J Epidemiol ,• 1990; 19:380-386 . • 2. Todd J.E., De Francisco A., O'Dempsey T.J.D., Greenwood B.M. The limitations of • verbal autopsy in a malaria-endemic region. Ann Trop Paediatr 1994;14:31-36 . 3. Snow B., Marsh K. How useful are verbal autopsies to estimate childhood causes of ,.• death? Health Policy Plann 1992;7:22-29 . • 4. Chandramohan D., Mande G.H., Rodrigues L.C., Hayes R.J. Verbal autopsies for • adult deaths: issues in the development and validation. Int J Epidemiol 1994;23:213- i. 222 . • 5. Jarvis and Boldt, Soc Sci Med 1982;16:1345-52 . • 6. Bagley et al., Psycho! Med 1976;6:429-38 . • 7. Riley I.D., Lehman D., Alpers M.P., Marshall T.F.D., Gratten H., Smith D. • Pneumococcal vaccine prevents death from acute lower-respiratory tract infections in ,• Papua New Guinean children. Lancet 1986;ii:877-881.

• 8. Zimicki S., Nahar L., Sander A.M., D'Souza S. Source book of cause-specific • mortality rates 1975-81. Demographic Surveillance System - Matlab, Vol 13. Dhaka: • International Centre for Diarrhoeal Disease Research, Scientific Report No. 63, ,. 1985: 1-103 .

• 9. Gray R.H., Smith G., Barss P. The Use of Verbal Autopsy Methods to Determine • Selected Causes of Death in Children. Liege (Belgium): International Union for the • Scientific Study of Population, IUSSP Paper no 30, 1990: 1-46 (Also published by: • The Johns Hopkins University School of Hygiene and Public Health, Institute for • International Programs, as Occasional Paper No. 10, 1990: 1-46) • 10. Ross D.A., Vaughan J.P. Health interview surveys in developing countries: a • methodological review. Studies in Family Planning 1986;17:78-94 . • • 98 • • • REFERENCES - Mortality Surveys: Sisterhood Method • 1. Graham W.J., Brass W., Snow R.W. Estimating maternal mortality: the Sisterhood • Method. Stud Fam Plan 1989;20(3):125-133. • 2. David P., Kawar S., Graham W.J. Estimating maternal mortality in Djibouti: an application of the Sisterhood Method. Int J. Epidemiol 1991;20:551-557. • 3. de Groof D., Bagnou A.S., Sekou H. Estimation de la mortalite maternelle en zone • rurale au Niger, par la methode des soeurs. Revue d 'Epidemiologie et de Sante Publique 1995;42(2): 185-186. • 4. Oosterhuis J.W.A. Estimating maternal mortality by sisterhood method in rural • Zimbabwe. Tropical Doctor 1993;23:67-68. • 5. Shiferaw T., Tessema F. Maternal mortality in rural communities of Ilubabor, • Southwestern Ethiopia: as estimated by the "Sisterhood Method." Ethiopian Medical • Journal 1993;31:239-249. • 6. Ullah S., Chakraborty N. The use of modern and traditional methods of fertility • control in Bangladesh: a multivariate analysis. Contraception 1994;50:363-372. • 7. Wirawan D.N., Linnan M. The Bali indirect maternal mortality study. Stud Fam Plan • 1994;25(5):304-309. • 8. Hanley J.A., Hagen C.A., Shiferaw T. Confidence intervals and sample size • calculations for the sisterhood method of estimating maternal mortality. Stud Fam • Plan 1996;27(4):220-227. • RECOMMENDED READING - Sisterhood Method • 1. Trussell J., Rodriguez G. A note on the Sisterhood Estimator of maternal mortality. • Stud Fam Plan 1990;21(6):344-346. • 2. Walraven G.E.L, Mkanje R.J.B, Van Roosmalen J., Van Dongen P.W.J., Dolmans • W.M.J. Assessment of maternal mortality in Tanzania. Br J Obstet Gynaecol • 1994; 101 :414-417. • REFERENCES - Morbidity Surveys: Health Interviews as Part of Long-Tenn • Demographic Surveillance • 1. Halabi S., Zurayk H., Awaida R., Darwish M., Saasb B. Reliability and validity of • self and proxy reporting of morbidity data: a case study from Beirut, Lebanon. • International J Epidemiology 1992;21:607-612. • 2. Alonso et al. Lancet 1994;344: 1178. • 99 • • ., .• • ,. REFERENCES - Morbidity Surveys: Health Surveys using Hospital Records

• 1. Nishioka S.A., Silveira P.V.P. Bacteriology of abscesses complicating bites of lance­ i.• headed vipers. Annals of Tropical Medicine and Parasitology 1992:86(1);89-91. 2. Silveira P.V.P., Nishioka S.A. South American rattlesnake bite in a Brazilian • teaching hospital: clinical and epidemiological study of 87 cases, with analysis of • factors predictive of renal failure. Transactions of the Royal Society of Tropical • Medicine and Hygiene 1992:86(5):562-564 . 3. Nishioka S.A., Silveira P.V.P. Clinical and epidemiologic study of 292 cases of • lance-headed viper bite in a Brazilian teaching hospital. American Journal of Tropical • Medicine and Hygiene 1992:47(6):805-810 . 4. Silveira P.V.P., Nishioka S.A. Non-venomous snake bite and snake bite without • envenoming in a Brazilian teaching hospital: analysis of 91 cases. Revista do Instituto • de Medicina Tropical de Sao Paulo 1992:34(6);499-503 . • 5. Nishioka S.A., Silveira P.V.P. Bauab F.A. Bite marks are useful for the differential diagnosis of snakebite in Brazil. Wilderness & Environmental Medicine 1995:6;183- • 188 . • 6. Silveira P.V.P., Nishioka S.A. Venomous snake bite without clinical envenoming ('dry-bite'): a neglected problem in Brazil. Tropical and Geographical medicine • 1995:47(2);82-85 . • REFERENCES - Demographic Methods • 1. Baea M., Garner P., Lai D. An algorithm to estimate age in women during their • child bearing years. Papua New Guinea Med J 1994;37:23-28 . • 2. Smith P.G., Morrow R.H. Methods for Field Trials of Interventions Against Tropical l. Diseases. Oxford, Oxford University Press, 1991. • 3. Malcolm L. Age estimation of New Guinean children. Papua New Guinea Med J • 1967;10:122 . • REFERENCES - Statistical Methods • 1. Schlesselman J.J. Case control studies: design, conduct, analysis, Oxford University • Press, New York, 1982, 80-85 . • 2. Kleinbaum D.G., Kupper L.L., Morgenstern H. Epidemiologic research, principles • and quantitative methods, Van Nostrand Reinhold, New York, 1982, Chapter 4 . 3. Lemeshow S., Robinson D. Surveys to measure programme coverage and impact: a • review of the methodology used by the expanded programme on immunization. • International Journal of Epidemiology 1985(Suppl):65-75 . • 100 • • CHAPTER 4 ; .• • METHODS FROM SOCIAL SCIENCES: OVERVIEW OF • QUALITATIVE AND QUANTITATIVE METHODS • F. Aboud • Social sciences, such as psychology and anthropology, can study human behaviour and other • complex issues that concern health. We will look at some qualitative and quantitative methods used by social scientists to study health-related variables. Health-related variables • can include: • • psychosocial variables as risk factors for disease such as • • education of mother • • socioeconomic status • • health behaviours such as • • bringing children for vaccination • drinking alcohol l • • • psychosocial antecedents and consequences such as • • knowledge about and attitudes toward smoking • • family approval for using contraception • • mental and social well-being variables such as

•~ • •••• • a social support network ,.• • mental disorders • Education of Mothers Measuring the education of mothers is not always simple and • straightforward. It may be that functional literacy is important, rather • than the number of years of school. You could measure literacy by j. having a mother read a health message which correlates to about a grade • 4 level and see if she understands . • • 101 , • .• • • Our Behaviour • Our behaviour causes positive and negative reactions from others, which can have a powerful influence on whether we repeat a behaviour. • Expecting disapproval from family and friends might be enough to stop • us from trying something in the first place. You could find out about • consequences by asking direct questions, e.g., what would your mother feel toward you if she knew you were using contraceptives: happy, • angry, sad, shame, nothing special? • • Five methods used by social scientists are covered in this chapter: • 1. Focus Group Discussion • 2. Systematic Observation 3. Key Informant Interviews • 4. Problem-Solving Groups • 5. Structured Self-Report Measures • • SUGGESTED SCHEDULE • If you have only a half-day for this session, we suggest you choose two methods to • explain and practice in groups. If you have a full day, you could cover four or five methods. • • 30 min Give an overview of methods from social sciences • 10 min Depending on the time you have, decide which methods will be • covered during the half-day or full-day session. • 1 hour One method • 1 hour One method • If you have a full clay, have 1 hour per method. • • 102 • • • • ,.• Social scientists use both quantitative and qualitative methods. For example, if you were studying the use of contraception in a population, the quantitative method could include ,.• the prevalence survey of contraceptive use, attitudes toward contraceptives, and the value of children. A focus group discussion on attitudes toward contraception uses a qualitative • method to study the topic. i • Assume a group of professionals or a community wants to change family planning • practices in an area. They might first conduct a survey with quantifiable answers to find out • the contraceptive prevalence rate in the community, the age and education level of those • using contraceptives, and the ideal family size among married couples. However, to start a campaign to inform people about their contraceptive options, you must understand • psychosocial factors, such as the value of children, if and how reproductive decisions are • made by couples, reasons for using or not using contraception, and whose advice is sought • and respected on these matters. You must understand and address their values, reasons, and decision-making processes if you want to inform couples about their options. The following • is a brief explanation of the five social science methods that can be used to develop a • campaign to change family planning practices in a community. They are more fully • developed in the next section . 1. Focus Group Discussion: Groups of men and women meet separately according to • age (premarital, newly married, long-term married, elders). The group might be • asked to talk with each other about the issues raised above. Researchers would keep • a written record of the relevant psychosocial factors, e.g., how couples discuss and • decide on their desired family size . • 2. Systematic Observation: As a non-participant observer, you would record • interactions between men or women of childbearing age and family guidance • counsellors. You might record everything that is said in sequence by the two people, as well as nonverbal reactions that convey emotions such as resentment, fear, • reluctance, enthusiasm. The family guidance counsellor might later record i• observations of the interaction. Records of these interactions could be used to see • how receptive people are to being approached with information about contraception, • and what kinds of promotion strategies are not acceptable . • 3. Key Informant: Men and women key informants could be interviewed on sensitive • topics such as sexual relations between married and unmarried couples, occasions when couples talk about having or spacing births, and why people have unwanted • children. This kind of information might be useful in identifying subgroups with • special problems, such as unmarried girls, and the source of pressure to have many • children . 4. Problems Solving Groups: To start using contraception requires that a decision be • made and problems overcome. To study the process, one would ask pairs of women, • for example two married friends, to come up with solutions to the following problem: ,.• 103 ' • • • • How would you tell your husband that you do not want more children even though he • has told you he wants a large family? Another problem to discuss is: What would you do if you thought you might want to use a contraceptive but other people told you • it was dangerous? From these discussions, one could identify the information women • need and use to help them solve problems and make decisions that they are • comfortable with. • S. Structured Self-Report Questionnaires: These can be used to assess the value people • place on having children, their attitudes toward contraception, and their knowledge • about different methods of contraception. You could modify these questionnaires based on information gathered using the above methods, while retaining the structure • of the questions and the quantifiable nature of the responses. • The characteristics of qualitative and quantitative methods overlap in a number of • ways. Qualitative methods provide rich descriptive information about a community or about individuals, usually accompanied by information about the context in which they function. • This information is useful when the researcher lacks enough understanding of a phenomenon • to formulate propositions to be tested with statistics. Quantitative methods are used to • establish the quantity of a phenomenon such as the prevalence, the severity, the distribution across a population, the frequency of a behavior, the intensity of an attitude, and the norm of • a group. • Both types of methods are required to answer many research questions. The two • methods can be used sequentially to build on each other, or simultaneously to support and cross-validate each other's findings. For example, often questionnaires are given to people • immediately before they participate in a focus group discussion; this jogs the memory of • participants about their personal experiences of the topic being discussed, and can be used to • cross-validate the measures, while the focus group allows a broader range of response • options than were tapped by the questionnaire. • Distinctions between the two methods can merge or blur. For example, using a • coding scheme, you could assign numbers to the output from qualitative research and • conduct statistical analyses. Thus, researchers are learning how to incorporate the strengths of one method into the other. • One example of using both research methods is the Healthcom' s Health Education 1 • Project. This project used a diverse set of quantitative methods to collect community data on childhood diarrhea and mothers' care of sick children. Qualitative methods were used to • find out how easily mothers understood the concept of "germ" and how they reacted to new • ways of coping with childhood diarrhea. They used the data to develop new strategies for • increasing mothers' use of ORS. • Another example is the AIDS Rapid Assessment Procedures. 2 The Social and • Behavioural Research Unit of the WHO's Global Programme on AIDS has prepared • 104 • • • • • ,• materials to collect data on patterns of behaviour related to HIV and AIDS. They include • most of the methods described in this chapter, and something new -- personal diaries! • You could distribute Table 1 as a handout or put it on an overhead to clarify the • characteristics of qualitative and quantitative methods . • , .• • ,•• • • • i.,.• , . • • , .• • • • , .• • • 105 • • • TABLE 1: QUALITATIVE AND QUANTITATIVE METHODS • Qualitative Quantitative • Purpose: Relevant behaviours Prediction • Intelligibility Prevalence Provides context info Control extraneous info • Describe process Evaluate outcome • Synthesis vs Analytic: Holistic understanding Separate components • Respondents' perspective Statistical relation • Context: Context-specific Context-free • Driven by Whom: Respondent driven Researcher driven • Measures & Questions: Determined by respondent Predetermined • Response Categories: Determined by respondent Predetermined • Semi- or unstructured Structured • Sampling: Small sample Large sample • Not representative Representative • Information-rich Unbiased • Interpretation of responses: Biased by researcher Biased by respondent • who writes report who interprets question • Location of Researcher Bias: Interpreting responses Selecting questions • Selecting response options • Location of Respondent Bias: Recall limitations Interpreting question • Biased causal judgments Social Desirability Respond to presence of Res. • Reliable (test-retest) Must be demonstrated Usually very good • Validity Usually very good Must be demonstrated • Representative Only if proper sampling Only if proper sampling • Compare with Other Data: Difficult unless comparable Yes response categories used • Quantify for Stats Tests: Experts can scale data Yes Researcher codes responses • and tallies frequency • Cost, Effort, Time: At the end to code and describe At the start to develop, pretest, • responses revise measure • • 106 • • : . • i.• 1. Focus GROUP DISCUSSION A focus group discussion can be organized to provide information about perceptions of a • specific topic of interest. The researcher, or someone hired by the researcher, leads the • group. The leader or moderator's job is to facilitate discussion among the participants, while • keeping them on the topic. The moderator also encourages participants to express their ,• perceptions and opinions using two strategies: by posing open-ended questions, and by letting people respond to ideas expressed by other group members. The goal is to get as many • people as possible to express their different perceptions, not to have them come to an • agreement. ;.• A focus group discussion is carried out for one of the following reasons: • (a) identify issues that are important to people regarding the research question . • Once you know the important issues, you could write a structured i. questionnaire with closed-ended questions to include these issues . • (b) check that the results of a survey or another research method are accurate . • For example, you might question your survey results showing that women like and value modern contraception, if in your focus group they say it is valuable • only for other women who want to stop having babies but is dangerous for • women like themselves who want more children . • (c) get more detail from people if you find something unexpected in a survey . • Strengths: A focus group provides information in the language of the people, rather than in • the language of the researcher. It saves time, because you are interviewing about 10 people • at one time. People often say more and are more honest when talking with their peers, than • if directly interviewed by a researcher . • Limitations: The opinions expressed in a focus group are not necessarily representative of • the community; people who speak up loudly and clearly will be heard more often than people • who are shy. You cannot draw conclusions about the prevalence of certain opinions, e.g., about the number of women who would be interested in using contraception, because you do • not systematically ask each participant to state her opinion . • • When setting up a focus group discussion, pay attention to the following: • A. WHO WILL YOU INVITE TO BE IN THE DISCUSSION? i. Choose six to ten people to be in the group. The more people, the more opinions you will get. The fewer people, the more chance you have to get detailed descriptions from • everyone. Invite people who are not shy or embarrassed to talk about themselves . , .• 107 • • • • They should come from similar backgrounds so that they feel comfortable talking about themselves. For example, uneducated women might not feel comfortable talking in a • group of highly educated women. In this case, invite women who have a similar level of • education to participate in the group discussion. They should be different enough that they • might have different opinions and/or practices. If everyone agrees on everything, you won't • have much discussion. • B. WHO IS THE FOCUS GROUP MODERATOR? • The moderator must be someone fluent in the local language, who could be the researcher, or someone hired from the community. You must be careful not to hire someone whose • opinions are known in the community because group members might not feel free to express • their own opinions. The moderator must be able to encourage people to speak openly and • keep the discussion on track. • C. WHERE DO YOU HA VE THE DISCUSSION? • Have the discussion take place in a neutral location. Examples of locations that are not • neutral are: • • a clinic for a discussion about female circumcision • • a church for a discussion about abortion • Have everyone sit so they can see each other. A circle is good to make everyone feel as equal as possible. • D. WHAT QUESTIONS SHOULD YOU ASK? • Plan about eight questions ahead of time so that the discussion will provide the information • you need. You will have to remain flexible because you may think of new questions to • follow up on from comments during the discussion. • Start by welcoming people and introducing yourself (name and organization). Tell the • participants why you have invited them: • "We have invited you to come because we want your opinions about family • planning. We want to find out about as many different opinions as possible; we are not trying to have everyone agree. There are no right or wrong • answers. We want your honest thoughts and feelings." • The first question should be open-ended. For example: • "How many children do you think is a good number to have?" • • 108 • • • ,•• , .• • Follow this up with a sequence of more specific questions: • "Do you ever feel you want to stop having children?" • "Do you ever feel you want to wait before having another child?" "What would you do if you wanted to stop getting pregnant?" • "How do you people feel about pills (or condoms)?" • Repeat the question if the discussion goes off track. Ask a new question whenever the • discussion becomes repetitive . • E. How DO YOU ENCOURAGE DISCUSSION IN THE GROUP? • • Pay close attention to each person when they speak, just as a host or hostess would • when receiving guests into their home . • • Encourage a quiet person to talk by looking directly at them and using their name, • "Susan, I'd be interested to hear what you think of that." • • Don't appear to approve of certain answers and disapprove of others . • • Don't ask people to defend their opinions . • • Don't dominate the discussion as a leader, and don't express your own opinions. -·• • If everyone seems to agree, say, "Who feels differently?" • If this does not bring any different opinions, try stating a new, but reasonable, opinion: • "I know some women who are afraid of what their husbands might say if they • expressed an interest in pills. What do you think of that?" • F. How DO YOU RECORD THE DISCUSSION? • It is important to explain to the group at the beginning that the discussion will be taped • and/or notes will be taken. Make sure they understand that individuals will not be identified . Have an assistant record the discussion, not the leader or a member of the group. The best • situation is to have the discussion taped and have an assistant taking notes. Facial • expressions and body language cannot be picked up on a tape recorder but can be noted by • an observant assistant. • • • 109 ,.• • • G. How DO YOU ANALYZE THE DISCUSSION? • • Lists can be useful. For example, list all the different reasons people give for using • or not using contraception. • • Estimate how important an issue seemed by noting how much time was spent in • discussing it and how many people expressed an opinion. • • Classify responses to a question and count the number of people who gave each • response: e.g., three of the group members said they had discussed desired number • of children with their husband, four said they did not, and one did not comment. • • Have another person listen to the tape or read the transcript of the discussion and • have them classify answers and check that you agree. • • • • ·- ---·· --­ - ·- .:.- • • ==-.:- -""";":::'- • - ---- • • • • • • • • 110 • • • •• II • f • I • I , .• • ,•• ; . •• -·• t

,.; • • ,.••• , . • ••• '-· • ; .• ; .• ,.j.• r.'.. , . I • I !' • 111 t ; .• • • • • • • • • • • • • • • • • • • • • 112 • • • • • • 2. PARTICIPANT AND NON-PARTICIPANT SYSTEMATIC OBSERVATION Systematic observation is a planned, methodical, objective observation of events in their • natural sequence and context. The observations are recorded in more or less detail, • depending on how specific the research question. In an exploratory study, you may write • down everything you observe so you don't miss potentially important variables. If prior information leads you to suspect the importance of certain variables, the record may be • selective. A detailed record would include a complete description of everything you • observed: the physical space, the actors, the general activity, the objects, and the specific • actions of people. Or, a record could include only selective observations of how often • specific actions were performed by specific people in specific places. For example, how l. often do mothers wash their hands with soap while doing household chores, or how often do • children use the latrine or the bushes to urinate/ defecate during school hours . • Participant Observation • The researcher participates in the daily life or ongoing activities of the people being studied • over a period of time. A close relationship develops between the researcher and the subjects • as they go about their normal activities. The researcher observes what is happening, listens to what people talk about, and asks questions to clarify the meaning of events. The context • of events, the sequence and connection of events, people's reactions to events, the words • people use to communicate to each other, and their explanations to the researcher provide the meaning of events. , .• The participant researcher's record of events includes a daily diary or work schedule, • jottings of key words taken during an event, and expanded field notes of details and personal • impressions written at the end of the day. The expanded field notes are the core of the data . One of the most sensitive issues for participant observers is how to become accepted • as a trusted, unobtrusive participant so as not to alter the activity with one's presence. To • become trusted, the participant must be honest, unassuming, non-judgmental, and accept • whatever role he/she is allowed to play within the community. To be unobtrusive, the • participant must not be too obvious about taking notes in public . • Advantages of participant observation are: ,• • The researcher can observe an activity that is usually hidden from strangers, such as !.• circumcision or wedding rituals. To observe the activity one must be a participant. • The actions are described within their natural context and sequence of events. This • helps to determine natural antecedents and consequences of events . • The meanings and words people attach to events can be described. This helps to • design a sensitive intervention that will be meaningful to the community . • 113 • • • • Disadvantages are: • • The method of recording is time consuming and its credibility depends on the • expertise of the observer. • • The researcher's biases can potentially play a large role in what is observed and how it is described. A researcher who is aware of biases and takes pains to eliminate them • can achieve greater objectivity. • • Sequences of events do not address the reliability of a connection between two events. Many events may precede a health problem but only some of these will be reliably • connected with the problem. • Non-Participant Observation • The non-participant researcher is an observer who records events as they happen and does • not participate in the activities of the community. The observer does not try to become an • accepted member of the community and does not develop close relationships with the people • being observed. The detachment of the observer is expected to reduce bias and allow the observer to provide an accurate and complete description of events. Because the observer is • responsible only for recording events, and does not make judgments in the field about which • events to record, training is more specific. Community members can therefore be trained to • observe and record events. • This method is useful when a specific event is of interest, e.g., hand washing, beer • drinking, nurse-mother interaction. The researcher may even select beforehand the physical • setting and the actors of interest. The method is also useful for describing events that people • do not report clearly, either because they do it habitually without much thought, or because they do not often put it into words. It is more accurate than a self-report as a record of • complex behaviours and is uncontaminated by the respondent's memory. It provides more • detail about a smaller set of events than participant observation and can be statistically • analyzed more easily. • The observer keeps a low profile in the setting while recording events. A narrative • record includes details of events as they unfold, in their natural sequence. Things that • happen before and after are also recorded. When the narrative is being recorded • continuously, times can be written beside the event to indicate the duration and interval between events. • Sometimes the record includes information about a limited number of specific events • rather than a narrative. Specific events or behaviours are defined and listed on the record; the observer may check them off whenever they occur, note their duration, and note the • consequence. For example, observations of breast feeding may require information on time • of day, whether instigated by mother or child, duration on ea.ch breast, interruptions. • 114 • • • • , .• Observations of worker-patient interaction at a health clinic may focus on the following • behaviours: greets mother, asks mother about her child, asks mother open-ended question, • praises mother, asks mother closed-ended question, demonstrates mixing ORS, and has ,. 3 mother demonstrate mixing ORS. • The advantages of non-participant observation are: • • Health behaviours and other observable events can be described in great detail, • maintaining the natural sequence of events as well as the contextual antecedents and • consequences of the behaviour . • • The data are more amenable to statistical analysis; for example, by coding each unit • of observation under a category and analyzing the frequency counts for each category . • This means that communities can be compared and statistical relations between events • can be inferred . • • It is less time consuming and so a researcher can observe a larger sampling of people • and events . • • Because little memory or emotion is involved in the recording of events, the record is • less likely to be influenced by observer biases . • • Inter-observer reliability and reliability-over-time can be examined to enhance i.• credibility of the data . • I • • • the silen~ • observer • Disadvantages are: • • The observer must keep a low profile so as not to interfere with the ongoing activity . • People might not be used to having strangers record their actions and may feel uncomfortable at first. However, if the observer is present for a long time, people • will usually ignore him or her and carry on as usual; familiar community members I.• should always be trained to do the observing . 115 • • • • • Because the non-participant observer does not question people about their actions or their understanding of events, only the observed events themselves are clues to the • meaning of events. • Describing the Data • Data collected through observation must be reduced in order to be summarized, presented to • others, and used for the benefit of the community. Participant observation is usually • organized into themes, and the results applicable to each theme are described. Non­ • participant observations are usually summarized to allow quantification. For example, the domain is broken down into well-defined categories, e.g., washes with water only, washes • with soap and water, does not wash. Each observation unit is then coded as one of the • categories. The number of units fitting each category are counted and then statistically • analyzed. Complex categories that describe sequences of events can also be used. An • example of a sequence is the health worker questions, the mother responds, and the health worker acknowledges the response with speech or a gesture. New techniques are being • developed to reduce and organize narratives. 4 • Frequencies can be examined for statistical association. For example, when the number of breast feeding episodes instigated by mother vs child are available, the frequency • of these events can be cross-tabulated by demographic variables such as age and sex of child, • age and education of mother, and child malnutrition. • • • • • • • • • • 116 • • • • • • • • , .• • ••• • • !.i. :• , . ,.• ••, . ••• ! • ,.• , . ; .• • • 117 ••• • • • 3. KEY INFORMANT INTERVIEWS • THE KEY INFoRMANT • A key informant is someone from the community who has the following characteristics: • • wide contacts with people in the community and therefore is in a position to inform • the researcher about community events. The informant must be thoroughly involved • and active in the community. • • good communication skills, articulate, willing to talk without reservation and to volunteer information without direct questions • • reflective and therefore able to expand, modify, and clarify the researcher's • interpretations of events • • trustworthy • Different ways of selecting key informants are: • • to ask people in the community who is best informed about events in the community, • and who they trust to describe their community accurately. This amounts to a • popularity contest. • • the snowball method; people who are good key informants are asked to name others • who fit the criteria • The WHO suggests selecting one key informant for every 1000 population. • THE INTERVIEW • How does the interview proceed? A relatively unstructured interview might begin by asking • the key informant some open-ended broad questions. This leaves the informant to direct his/her answer in whatever way it goes. The researcher could then ask some specific • questions, again letting the informant provide as much detail as possible. Specific • interpretations may be offered by the researcher to obtain the informant's confirmation or • disconfirmation. The interview stops when no new information is being provided. The researcher can return to the informant after more information is collected from other sources • in order to clarify. • Some key informant interviews are more structured, with specific questions asked in a • specific order. For example, the WHO's key informant interview on attitudes toward mental illness follows a specific structure. 5 The interviewer first asks some specific questions to • determine the informant's range of contacts and power of perception and recall: • 118 • • • • • "How many blind people living in (name community) do you know by name? • How many people who cannot walk or can walk only with crutches living in • (name community) do you know by name?" • Then you ask about what kinds of healers work in the area, how many healers there are, and what type of patient they treat. Continuing with the topic of healers, you ask what kind of • help most people would first seek for 10 different problems, such as fever, convulsions, • sleeplessness, and excitement. Finally, read seven vignettes on common mental health • problems, asking the informants if they know anyone with such a problem, and then asking whether such a person would have trouble being accepted by their family, by a marriage • partner, and by a job employer. , .• • • • • • , .• • • • ,•• • • • • 119 • • • • • • • • • • • • • • • • • • • .• , • 120 • • • , .• , .• • , .• • • • :•• , .• ,.• : .• • ,•• . • • • : .• • , .• • 121 l • • • • 4. PROBLEM SOLVING AND DECISION MAKING • Most health development at the community level involves identifying and solving problems. • The process of solving problems to arrive at a solution can be recorded and each community • member's acceptance of the group's solutions can be assessed. • The problem to be solved can be taken from the focus group discussions that took place earlier. Have the class divide into groups of two to eight people. The group is given • the problem and discusses how they would solve it. A complete record is kept of the • discussion, preferably on a tape·recorder. The discussion is then transcribed and each • statement coded. The frequency of each code for each person is tallied for analysis, along with the number of turns taken or statements made. We will look at how to record, code, • and quantify process variables such as: • elaboration of the problem new solution offered • repetition of solution • elaboration of own solution, or other's solution • evaluation of solution • agreement or disagreement information sought • Individuals can be asked privately to recall the solution(s) they consider best. Or • each solution mentioned could be read to members of the group, who are asked how strongly they agree or disagree with the solutions. • Here is an example from Cynthia Ticao's6 study of pairs of mothers discussing a • child-feeding problem: • "What do you do if your child frets and will not eat during meals?" • Mother 1: I wait until the child asks for food. If he doesn't ask, I wait. He might be • tired. Mother 2: I allow them to rest for a while, then I give them food. I persuade them to eat • by offering something they like, like a sweet. • Mother 1: You can't compel them to eat. Sometimes one grandmother comes and gives • them food. They lose their appetite and my food is wasted. • Mother 2: I want to see what they're eating. There are some children who will not eat even if they haven't had food. • Mother 1: When they're hungry they ask for food. • Mother 2: And when they're sick, I really have to push them to eat. You have to • persuade them to eat. Mother 1: You really have to persuade them to eat. With children, even if you push them • and they're hungry, sometimes they don't want to eat. • 122 • • • • • A small discussion group can foster active problem solving with new solutions. In • the above example, a number of solutions are offered, such as waiting until the child asks, • waiting until the child rests, persuading, compelling and not compelling. Mother 2 • elaborates on the persuasion solution by saying she would entice the child with a sweet first. • The mothers also elaborate on the problem of a non-eater: grandmothers feed them, sick • children won't eat, food is wasted . • When the discussion was over, Ticao asked the mothers privately to say what they • thought were the best solutions. The solutions they mentioned came not from their own ,. solutions before the discussion, nor from what they heard from the other mother during the discussion. The solutions were new solutions generated by the mother herself during the • discussion . • • , .• • • ; . • • • • • • • • • 123 • • • • • • • • • • • • • • • • • • • • 124 • • • • • ••• • ••• • • , .• • • • ••• ••• • , . • • • • • •• 125 • • • • S. STRUCTURED SELF-REPORT MEASURES • Structured self-report measures are quantitative measures of the way a person perceives their • own behaviour, someone else's behaviour, the environment, or a person's knowledge and • attitudes about a target. Multiple items are used to cover all aspects of a question. The • items are worded so that all respondents will interpret the items the same way. Items are • usually pretested for clarity and interpretation, and can be open- or closed-ended. The person's responses are scored. Usually a lower score is negative and a higher score is • positive. For example, a high score on an answer to a question on risk of sexual behaviour • would indicate a low-risk sexual behaviour. The scores for individual items are added up to • give a total score. • Obviously, it saves time and effort if you can find an existing measure that suits your • needs. Some measures have been used a lot in developing countries. Other measures have • been used only in developed countries and would have to be modified for use in developing countries. Using an existing measure has the advantage of allowing for comparability. • However, if a measure does not exist, you must develop your own. • Existing Measures • The WHO has developed and validated measures for mental health in adults and children. • The Self-Reporting Questionnaire has 24 items which are easy to administer and score, and • require simple Yes - No answers from the respondent.5 For example: • Do you often have headaches? Yes No • Is your appetite poor? Yes No • Are you easily frightened? Yes No • Do you cry more than usual? Yes No • Other measures have been developed by researchers to assess stress, health locus of • control, subjective experience of pain, values, and health belief systems. The journal, Social • Science and Medicine, publishes research that sometimes uses standardized tests. It would • be wise to review these tests before deciding to modify or develop a new measure. • Developing a New Measure • When developing a new measure such as a KAP questionnaire, the following issues must be • considered: • • How will it be administered? Interview is best for people with little education, but • self-administration is possible for highly literate people. • • 126 • • • • • • • What is the best question format? Open-ended questions requiring short answers might be desirable when assessing knowledge, e.g., How can a person get AIDS? • Closed-ended questions that require only a Yes-No-Don't Know answer or Agree­ • Disagree answer require less verbal and memory skills from the respondent; e.g., Can • a person get AIDS from kissing? It may be best to have some open-ended followed • by some closed-ended questions . • • What is the best response format and how does one score the responses? The open­ • ended question requires a verbal response that must be written down verbatim and • then scored, perhaps one point for each correct idea. For example, in response to the question on how a person gets AIDS, you might give one point for each correct mode • of transmission for a maximum of three (how many do you expect from a lay • person?). The closed-ended question should allow for three possible responses: yes, • no, and I don't know. Saying one doesn't know is as bad as the incorrect answer and • so both should be scored zero; the correct answer is scored one point. • • How many items should there be? To enhance the reliability of the measure, it • should have many items, perhaps 10 or more. Scores on the items can then be • summed. On a KAP questionnaire, you would obtain three separate scores -- one for Knowledge, one for Attitude, and one for Practice. But administer the subscales in • the reverse order, with Practice items always first . • Generating items is a good exercise for a group of people; then show the items to several colleagues to get their opinions on the clarity of the wording and items to be added . • Pilot testing should be used to identify irrelevant items, poorly understood items, reliability, • and validity . • • • • • • • • 127 • • • • KAP ITEMS • Here are some KAP items we developed concerning mothers' care of infants over 4 • months, where K means knowledge, A means attitude, and P means practice. • Pl. What did you feed your baby immediately after delivery? • breastmilk = 1 • butter, water & sugar, cow milk = 0 • P2. What do you now feed your baby? • breastmilk + solids = 2 only breastmilk = 1 • else= 0 • A. Which of the following are good for your baby and which are bad? immunization shots Good= 1 • sunshine Good = 1 • uvulectomy Bad = 1 • Kl. What is a germ? • little = 1 • causes disease = 1 (for max 2) • K2. What is ORS? • for diarrhea child = 1 prevents dehydration = 1 • (for max 2) • K3. Does eating carrots prevent blindness? • Yes • • • • • • • • 128 • l • t

• I[ • f • I • • I

• I • I

• I! • I ,.• I I I I ' • I I • I I I I • I I I • I • !' ••• • • -• • • • • • • 129 • • • • REFERENCES • 1. Graeff, J.A., Elder, J.P., & Booth, E.M. (1993). Communication for health and • behavior change: a developing country perspective. San Francisco: Jossey-Bass. • 2. Scrimshaw, S.C.M., Carballo, M., Ramos, L. & Blair, B.A. (1991). The AIDS rapid anthropological assessment procedures: a tool for health education planning and • evaluation. Health Education Quarterly, 18, 111-123. • 3. Rasmuson, M.R, Seidel, R.E., Smith W.A., & Booth, E.M. (1988). Communication • for child survival. Washington DC: USAID. • 4. Bogdewic, S.P. (1992). Participant observation. In B.F. Crabtree & Miller, W.L. • (eds.), Doing qualitative research: multiple strategies. Newbury Park, CA: Sage, • pp.45-69. • 5. Harding, TW., Climent, C.E., Giel, R., Ibrahim, H.H.A., Murthy, R.S., Suleiman, • M.A., & Wig, N.E (1983). The WHO collaborative study on strategies for extending • mental health care Il: The development of new research methods. American Journal • of Psychiatry, 140, 1474-1480. • 6. Ticao, C.J. (1994). Mothers' problem solving in relation to child nutrition in the • Philippines. McGill University PhD thesis. • 7. Holmes, T.H. & Rahe, R.H. (1967). The social readjustment rating scale. Journal of • Psychosomatic Research, 11, 213-218. • 8. Saroson, I.G., Johnson, J.H. and Siegel, J.M. (1978). Assessing the impact of life changes: development of the life experiences survey. J oumal of Consulting and • Clinical Psychology, 46: 932-946. • • RECOMMENDED READING • 1. Morgan, D. L. (1992). Doctor-caregiver relationships: an exploration using focus • groups. In B.F. Crabtree B.F. & Miller W.L. (eds.), Doing qualitative research: • multiple strategies. Newbury Park, CA: Sage. pp 205-227. • 2. Moulton, J. and Roberts A.H. (1993). Adapting the tools to the field: training in the • use of focus groups. In RE.Seidel (ed.), Notes from the field in Communication for • Child Survival. Washington: USAID, pp.31-37. • • 130 • • • • • 3. Steckler, A., McLeroy, K.R., Goodman, R.M., Bird, S.T., and McCormick, L. ,.• (1992). Toward integrating qualitative and quantitative methods: an introduction . • Health Education Quarterly, 19: 1-8, 1992 . • 4 . Gilchrist, V.J. (1992). Key informant interviews. In B.F. Crabtree & W.L. Miller • (eds.). Doing qualitative research: multiple strategies. Newbury Park, CA: Sage, • pp.70-89 . • 5 . Weick, K.E. (1968). Systematic observational methods. In G.Lindzey & E.Aronson • (eds.), The handbook of social psychology. Volume 2. Reading, MA: Addison­ • Wesley, pp.357-451. • • , .• • • • • • , .• • • , .• • • • 131 • • • • • CHAPTER 5 • EVALUATING COMMUNITY PARTICIPATION • W. Boyce, N. Khanlou, C. Lysack, S. Mulay, D. Zakus • This session focuses on the evaluation of community participation and should take from three • to four hours. The term "participation" is common in development vocabulary. Everyone has their own definitions, their own understanding of it, and their own standards for success. • Community members and groups participate in primary health care, community-based • rehabilitation, health planning, and policy development. How do we evaluate this • participation? Is it effective? Can it improve? Training researchers to evaluate participation • is a challenging task! • In this session we shall: • • clarify the concepts and terminology of community participation • discuss our experiences of participation j • • • learn some practical methods of analyzing and evaluating participation • We base the session on the following assumptions: • • participation of people in the decisions that impact on their health and their lives is a • basic human right which ought to be supported and enhanced ,.• • community participation is complex, has many dimensions, and is difficult to • understand • • we learn whether participation is effective by studying, analyzing, and evaluating it. :. • participation can always be improved and improvement should be the primary goal of • an evaluation • Use brief presentations, case studies, group discussions, and practical evaluation activities . •• The session should be highly participatory, of course! ,. • • • 132 • • • • SUGGESTED SCHEDULE • This session on evaluation of community participation should take from 3 to 4 • hours. • 5 min Give a brief overview of topics covered. • 15 min Give an overview of the concepts and terminology of community, • participation, and community participation. • 40 min Activity 1 on community participation • 5 min Review the context and stages of community participation, roles of • participants and problems of community participation. • 30 min Activity 2 with case studies • 20 min Review the topics of evaluating participation. • 45 min Activity 3 on evaluation of community participation • • • •

HOW MUC~ AllE WE • ACCOMPLIS~ING IN PROBLEMS WE • TERMS OF WHAT 'WE MAVE RUN \Mlo HOPED FOR ANO z. • PLANNED? 3. t.'· ,.---::::::;.-~=- 4. • 3.... • ~~'S IWlPL15H.,, .. ""e o ~R: ... D ~'Y5 WE 1'1116)\l' • '· '·• .,,Rove-. .... • 2.. s. • .... • • 133 • • • ,.• ,. 1. CONCEPTS AND TERMINOLOGY OF COMMUNITY, PARTICIPATION, AND • COMMUNITY PARTICIPATION '. •-- This part of the session should take about 15 minutes and will: • • clarify the concepts of community, participation, and community participation • • look at the importance of participation in health projects and programs • • consider participation as a means or an end • • include a group exercise • A. CLARIFYING THE CONCEPTS (AND TERMINOWGY) OF COMMUNITY, • PARTICIPATION, AND COMMUNITY PARTICIPATION • Make sure that group members understand the importance of these concepts . • Many definitions of community and participation exist. Community ; .• • A community has been defined as: • "a group of people living in the same area and sharing the same basic values i•• and organization" • which combines concepts of geography and values such as freedom, prosperity, family; • "a group of people sharing the same basic interests at any given time" • such as economic, disability, women's interests; • "a group of people targeted for interventions" by a service provider, indicating a common need for things such as maternal health, accident ;.• prevention, or immunization. 1 • • Labonte's (1988) perspective on community is that: • "not all neighbours have the same interests, and affinity (shared values, • concerns) is an equally important aspect of community. Community cannot be • defined demographically - a common error committed by health planners . 134 , .• • • • • There is no Black, Asian, welfare, or women's community as such. Rather, community exists when individuals with a shared affinity, and perhaps a • shared geography, organize. "2 • Participation • The United Nations Economic and Social Council resolution 1929 (LVIll) defines • participation as: • the voluntary and democratic involvement of people in: • (a) contributing to the development effort • (b) sharing equitably in the benefits derived therefrom • (c) decision making in respect of setting goals, formulating policies and planning • and implementing economic and social development programmes3 • Various definitions of participation share three common characteristics: • (a) participation is an active process: • "the mere receiving of services is not participation" • (b) participation has the element of choice: "participation implies the right and responsibility of people to • make choices and therefore, explicitly or implicitly, to have • power over decisions which affect their lives" • (c) for participation to take place, the potential for the choice to be effective must • exist: • "this suggests that mechanisms are in place or can be created to 4 • allow the choice to be implemented " • Community Participation • Community participation must take place for two closely related ideas, community competency and community empowerment, to evolve. • "Community competency" refers to a group's ability to achieve a specific task. • Community competency is based on a consensus model, i.e., group members agree on group • goals. • "Community empowerment" occurs in the context of a disadvantaged group, so a degree • 135 • • • • • • of moral legitimacy is needed for empowerment to emerge from community participation . Community empowerment requires an in-depth understanding of a health issue and its • context, a collective self-awareness of the ability to achieve social action and improve social • conditions, and a fundamental struggle for power . , .• • B. PARTICIPATION IN INTERNATIONAL HEALTH DEVELOPMENT • Use the overhead on page 137 of the following points to generate discussion of the origins :• and history of community participation in international health development. Identify • purposes of participation in health . • Remind the class that when speaking of community participation, we assume that • participation of people in the decisions that impact on their health and their lives is a basic • human right which ought to be supported and enhanced . • Alma Ata Declaration (1978) 5 • "The people have the right and duty to participate individually and collectively • in the planning and implementation of their health care" • Bamako Initiative (1992) 6 "Goal: to improve the quality of basic public health services in sub-Saharan • African countries through the community's involvement in the management • and financing of local health services, and through strengthening of district • management" ,.• From Alma Ata to the Year 2000: A Mid-Point Perspective (1988)7 "the empowerment of people through the provision of information, decision ,. making opportunities, and technical support, so as to enable them to share in the opportunities and responsibilities for action in the interest of their own • health" UNICEF Health Policies and Strategies: Sustainability, Integration and National • 8 • Capacity Building (1992) • "strengthening of community structures and decentralization of administration; • setting of overall goals and targets at the national level; the community's • participation in financing and management of the health services at the local • level" • • 136 • • • • Overhead • • COMMUNITY PARTICIPATION IN INTERNATIONAL HEALTH DEVELOPMENT • Alma Ata Declaration (1978) 5 • "The people have the right and duty to participate individually and • collectively in the planning and implementation of their health care" • Bamako Initiative (1992) 6 • "Goal: to improve the quality of basic public health services in sub­ Saharan African countries through the community's involvement in • the management and financing of local health services, and through • strengthening of district management" • 7 • From Alma Ata to the Year 2000: A Mid-Point Perspective (1988) • "the empowerment of people through the provision of information, • decision making opportunities, and technical support, so as to enable • them to share in the opportunities and responsibilities for action in the interest of their own health" • UNICEF Health Policies and Strategies: Sustainability, Integration and 8 • National Capacity Building (1992) • "strengthening of community structures and decentralization of • administration; setting of overall goals and targets at the national • level; the community's participation in financing and management • of the health services at the local level" • • • • • 137 • • • , .• •••• • Note the recurring themes of community participation in the above: • • as promoting efficiency in service development and delivery This participation includes community contributions of money, labour, materials to achieve • the instrumental objectives of the project . • • as promoting social involvement and community building • This participation includes elements of community learning and self-responsibility and self­ • improvement. ,. • as promoting equity in decision making • This participation includes aspects of community control, social relationships and power • dynamics . • C. COMMUNITY PARTICIPATION AS A MEANS AND/OR AN END • The Instrumental Approach - World Health Organization • participation as a strategic means towards a goal • The Democratic Approach - European/American • participation as a right and an end in itself The United Nations Approach ; .• "The creation of opportunities to enable all members of a community and the • larger society to actively contribute to and influence the development process • and to share equitably in the fruits of development" participation as both a means and an end • ! • Approaches to community participation in health development projects vary between • two opposites. 9 This cartoon illustrates this well . I • I • I • f • 138 • I I • I • I • • Overhead • • • • Participation Partici pa ti on • as a way as a way for • to control people people to gain control • HOW CAN WE GET THEM WHY NOT .JOIN • TO DO WHAT WE WANT? HOW CAN TOGETHER, RENT A TRUCK, ANO • WE GET A TRANSPORT IT WHY NOT THROUGH FAIR PRICE • TO MAR.KET •coMMUN \TY FOR OUR GRAIN? OURSELVES? • PARTICIPATION"? we CAN GET INTERNATIONAL • FUNDING! • NOTKlNG­ • IF WE • STICK TOGETHER~ • • • . . ' . • .· • • • • 139 • • • • • • From participation as a way to control people: • Manager: "How can we get them to do what we want?" • Advisor: "Why not through community participation? We can • get international funding!" • To participation as a way for people to gain control: • 1st Farmer: "How can we get a fair price for our grain?" 2nd Farmer: "Why not join together, rent a truck, and transport ,.• it to market ourselves?" 3rd Farmer: "But what will the landlord say?" • 4th Farmer: "Nothing, if we stick together!" !.• Between these two opposites, there are many intermediate stages. .These vary • according to: ,.• • who really does the participating • • the function, or purpose of the participation • the centre of power : .• • It is important to remember that community participation can occur: • • without focused, intentional efforts at changing a social or health issue ,.• • without a collective self-awareness of increasing group cohesion and competency • • without the attainment of any social improvements • Sherry Amstein notes: • "There is a critical difference between going through the empty ritual of • participation and having the real power to affect the outcome of the process. "10 ••• • • • 140 • • • • • ••• • ••• • • • • • • • • • • • • • 141 •• • • • ••• • 2. THE CONTEXT OF PARTICIPATION • We shall take five minutes to review the context and stages of community participation, roles i. of participants, and problems of community participation. We shall then divide into groups and discuss projects, programs and policies using two case studies in Mexico and India. , .• Community participation in health occurs at numerous levels and in different settings • such as: implementing local health projects involving sanitation, immunization, health • education and community financing; organizing regional health programs involving research, training of personnel and development of infrastructures; promoting health policies which !.• affect a particular community . • Cohen/Uphoff Framework for Analysis of Community Participation • This descriptive framework views community participation as it happens at different stages of • a project, such as planning, implementation, receiving benefits, and evaluation. 11 • It was developed in the 1970s and 1980s to evaluate community participation in water , .• management programs in Sri Lanka and covers the following topics: • "WHAT kind of participation?" At what stage(s) of the project does participation occur? • a. During Planning • • initial decisions: should a project start? Where? Who should organize it? • • ongoing decisions - should the project change directions? •• • operational decisions - setting the rules for participation, membership, • meetings, selection of leaders • b. During Project Implementation • • resource contributions - labour, cash, materials, information • administration/ coordination - as employees, advisory committee members, • board members • • enlistment: volunteers, cooperating in program . • c. In Receiving Benefits • • material goods: not usually applicable in health projects • • social goods: publicly available goods, health clinic • • personal benefits: self-esteem, sense of efficacy, political influence • • • 142 • • • • d. During Evaluation • • formal evaluation of project • • informal support for project and lobbying • "WHO is involved?" Who is included and excluded? • a. it is important to assess the relative roles of participants: how are community • members, local leaders, professionals, bureaucrats involved • b. how many are involved • c. need to know their characteristics which may give them influence (gender, age, • ethnicity, SES, disability or health status) • "HOW are people involved?" How is participation organized? • a. basis of participation: motivations, voluntary or coerced • b. form of participation: formal or informal project organization, membership or open, • organizational rules, direct or representative involvement • c. extent of participation: time involved, duration/frequency, range of activities for community members vis-a-vis staff • d. effect of participation: degree of community power/influence • The following contextual factors may have an effect on participatory evaluation: • • past experience with participatory projects • • geographic factors of distance, weather, terrain • • socio-cultural-political traditions • Project design factors may also have an effect: • • resources to support participation (transportation) • • project flexibility to allow changes in direction • • number of staff members to facilitate participation • • 143 • • • , .• ,.• ,. , . • ~· ; . ;.• , . ,.• , . • ,.• , . • • • , .• • • • ,.• • 144 ,.• • • Handout • CASE A: COMMUNITY PARTICIPATION IN PRIMARY llEALm CARE, OAXACA, MEXICO • Oaxaca, located in the southwest of Mexico, is one of the poorest of Mexico's 31 states. Its • population of nearly three million is about 50 % indigenous and speaks several languages, though Spanish is prevalent almost everywhere. About 10 % of the population is • concentrated in the capital city, Oaxaca, and the other 90 % is scattered throughout some 570 • often very isolated towns and villages in the valleys and rugged mountains. Infant mortality • is about 40 per 1000 and illiteracy runs about 30%. Health services are delivered mainly by the Ministry of Health (MOH) to the poor uninsured population (25%), by the Mexican • Social Insurance Institute to those families who are covered through some form of • employment or by its program for marginal people (50 %) , and by the government workers • insurance program (25 %) . • The health system can be characterized by the typical pyramid with hospitals, large • urban health centres, town/village based health centres led by a recently graduated doctor, • and health posts that are serviced by volunteer health auxiliaries. Traditional healing • practices are also very popular and costly. Throughout the state, 55 modules, which are part of a MOH community participation program, each involve a health centre and about nine • surrounding health posts. A nurse supervisor from the health centre provides ongoing • support to the health post auxiliaries and the community health committees associated with • almost every health post. • The data from this case were derived from a study carried out in 1988 which utilized • questionnaire surveys, key informant interviews, observation, and literature searches. Forty • health centres throughout the state were visited, 20 with a module and 20 without. In all, 75 • towns and villages were visited and 39 doctors, 37 nurses, 37 auxiliaries, and 30 committee members were surveyed. • Auxiliaries were mostly female community members selected by the community • through various mechanisms. They had an average age of 28 years and an average education of six years. To become auxiliaries they received very little training: 25 % had none; 25 % • had some on location; 35% spent some time in a health centre; and only 10% had attended • the preparatory course. Seventy percent had received no training in the last year. They • were equipped with very little, if any, equipment -- a basic drug kit that was often depleted, • and very few educational materials. They saw about 40 patients per month. • The health committee members were on average 42 years of age, with an average • education of four years. Almost all were men and were selected mostly by their village. • Sixty-six percent had received no information about their duties and responsibilities. Both the auxiliaries and the committee members showed a marked lack of knowledge of primary • health care when contrasted with the nurses and doctors. • • 145 • • • •

1 • • Examples of community participation health activities included the health auxiliary herself and the health committee, construction of health posts and their maintenance, latrine • construction, street cleaning, building chimneys in homes, announcing vaccination • campaigns, transporting patients and medicines, and raising funds through dances, raffles or • a village family tax. The principal community participants included the auxiliary, health • committee members, municipal authorities, members of other local committees, teachers, and students. In reality, though, very little was being accomplished in the communities. Forty­ • five percent of the committees had done nothing in the past year . • It was also found that the Oaxaca MOH organizational structure included the module program, and its employees were charged with implementing community participation. The j.• MOH provided very basic material support, supervision (the checklist type), training, and a small monetary incentives to the auxiliaries ($4/month). Generally, though, the level of • support was very low, and there was little local input into the program and the determination • of its goals and targets . • At the local level there was no variation in health committee structure and only the • existence of a weak organization. Also, there was little variation in the operational processes • between the module and non-module health centres, although the module ones did have a little more supervision, participatory decision making, locally determined objectives, and • local contacts. Ultimately, it was found that the local participants were very dependent upon • the MOH and carried out very few activities. Interestingly, though, the perceived • effectiveness of their participatory efforts, when they happened, was high . • \ .• • • • ,.• : .• • 146 , .• • • • Handout • CASE B: CAMPAIGN AGAINST UNSAFE CONTRACEPTIVES, ANDRHA PRADESH, INDIA • A key health issue in developing countries has been and continues to be family planning. • Most family planning programmes are demographically driven. Many developing countries have attempted to curb population growth by using overtly coercive measures without making • any serious efforts to improve the delivery of primary health-care. Women in developing • countries have been the primary targets of family-planning programmes. Clinical trials for • new contraceptive research have also targeted women from developing countries as subjects. Distribution of unsafe contraceptives has been commonplace in family planning programmes. • The case described here is meant to illustrate the efforts of grass-roots women's • organi7.ations to mobilize their communities against the distribution of contraceptives that adversely affect the health and lives of poor women. From 1981-1982, the Indian Council • for Medical Research (ICMR) conducted phase ill and IV clinical trials with an injectable • contraceptive with the trade name Net-En. The initial study included over 3,000 women. • Phase IV trials begun in August 1984, were aimed to cover 45 primary health care centres • attached to 15 medical colleges and intended to recruit 2,250 women. • Another group, the Stree Shakti Sanghatana (Women Power Organization), a • grass-roots organization in Andhra Pradesh, was formed in 1985 and took up the issue of • using poor women as "volunteers" for clinical trials. Andrha Pradesh is a very poor province and most of the population in these villages are landless peasants who work as • waged labour. The problem of drinking was particularly serious among labourers. Many of • the labourers spent their entire wages on country liquor, depriving their families of bare • necessities and resulting in violence against women in the family. So the first campaign • launched by the Stree Shakti Sanghatana was against drunkenness. The women activists involved in this work were successful in winning the trust of the rural women because they • worked and lived in the villages. •.:; One such activist, a physician, found out through her contacts in the family planning department that Net-En was being distributed in one of the rural health care centres. The • Stree Shakti Sanghatana collected information about Net-En and produced materials about • Net-En in the local Telegu language so that women could better understand the pros and cons • of the contraceptive. In April 1985, women from the Stree Shakti Sanghatana decided to • take the information on Net-En directly to women being recruited for clinical trials. They went to a village medical camp at Patancheru, which was about to inaugurate a 12-month • trial with Net-En. When requests to stop the trial were refused, they requested permission to • speak to the 50 women who had agreed to be in the trial. They explained the nature of the • trials in Telegu. They found out that these women had only been told that if they took the injections they would not become pregnant. However, they were neither told that they were • participating in an experiment nor about the potential side-effects of Net-En. After their • session, only five women agreed to go through with the clinical trial programme. The news • 147 • • • • i • •• of the campaign spread to other cities among women's organizations and similar campaigns \.• were organized in other cities and districts. Stree Shakti Sanghatana, along with two other national women's organizations, five • physicians and a journalist, filed a writ petition in the Supreme court against the ICMR, • Ministry of Health, and Andrha Pradesh government. The petition, the first of its kind to focus on preventive action in women's health in India, popularized issues that were • considered to be taboo, such as questioning the family planning programmes. They also • tackled a subject previously considered to be too technical for women to handle. At the • same time, the court action impacted negatively on community participation because technicalities of presenting the case in a legal form restricted the petitioners to one type of ,.• contraceptive and its potential danger, and other similar contraceptives were excluded from the debate. The technical nature of the arguments made the case incomprehensible to • ordinary people. It also diverted attention from the need of men to participate in birth • control. While the court decision went in favour of the petitioners, the court expenses were • a burden on the few women's groups which launched the petition . • Subsequently, national and international links have been made and the campaigns • extended to other contraceptives such as anti-pregnancy vaccines. Several topics came up in • the mobiliz.ation by the community such as: • • sustaining the momentum and extending the campaigns • • influencing the policies of the international or multinational organizations that support • the contraceptive research to be women-centred, • • changing the attitudes of key players who make national policies, and • influencing the international women's movement to focus on empowerment of women • and reproductive rights and reproductive health in the context of primary health care , .• • • • • • • • 148 • • • 3. EVALUATING PARTICIPATION • This section will focus on evaluating participation. We shall: • A. evaluate the process of participation • B. evaluate the outcome of participation C. consider quantitative and qualitative challenges in measuring participation: • indicators or descriptions • D. learn about participatory evaluation of participation • E. look at methods for evaluating participation EVALUATION: • F. have a group evaluation exercise A TOOL OR A • Presentation (20 minutes) BURDEN • 7 • • A. EVALUATING THE PROCESS OF PARTICIPATION • To evaluate participation, we must begin by examining the process, or operationalization, of • community participation: • Context • • What is the socio-cultural tradition and leadership in the community'? • • What can we expect will come easily to community members? • • What are the social and structural barriers to participation'? • What resources are available to support the organization of participatory • mechanisms'? • Strategies • • How does community participation happen? • • How is it organized, including recruitment and selection processes'? • Breadth • • Who is participating, including the numbers of persons actively involved'? • • What are they doing, including the activities being done? • Depth • • How much control do community members have? • • What is the degree of influence of community members? • • 149 • • i

\ .• • • B. THE OurcoMES OF PARTICIPATION • We also need to assess the effects of participation . • Impact • • What is the effect of community participation on health status or health determinants? • • What are other benefits to participants such as jobs or reduced health costs? • Social Equity ,.• • Have participants achieved a change in social status due to their participation? }. • Have local power dynamics changed positively or negatively? • Has group empowerment occurred? • Sustainability ,.• • Is the program more sustainable due to participation? • C. INDICATORS VERSUS D~CRIPTORS OF COMMUNITY PARTICIPATION • Community participation is an exhaustive concept! Are there key things that an evaluator ,.• should look for? Community participation can either be empirically observed and quantified or it can • be qualitatively described. In many cases, the best solution is to carefully select a range of • quantitative indicators and qualitative descriptors . • Quantitative indicators of community participation are objectively observable events. These . I • indicators can be measured at both the level of the individual and of groups. Some examples ,•• of quantitative health indicators are as follows: the number of organizations represented at a health meeting • the percentage of community members aware of a health initiative I • the number of news stories broadcast per month on a new health project I the percentage of target group members who benefitted from the health project • the proportion of resources contributed by public and private health agencies • the number of activities that originated with the community • the number of Board representatives from the community I • Qualitative descriptors of community participation describe the nature of participation. For • example, was there a lot of involvement by people, were they influential in decision making, l ,.• and were they satisfied in the end? Other examples are: 150 I • • • • How resilient, committed, or influential are specific groups of participants? • Is interest in the health issue growing? Are recommendations for improvement proposed by community members adopted as • often as those suggested by health professionals? • Have financial resources been dedicated to enhance community participation? • Do participants in the project feel satisfied with their levels of involvement? Has the health project extended beyond its original target population? • Is there a collective sense of improved understanding about the health issue? • The table on pages 158, 159, and 160 at the end of this chapter illustrate a grid for the • evaluation of participation in a community health project. • D. PARTICIPATORY EVALUATION OF PARTICIPATION • Evaluation of community participation can be undertaken in many different ways, one of which is a participatory evaluation. Participatory evaluation is a form of participatory action • research and thus has two strict criteria: • it must combine research, education, and action it is fundamentally about who has power • In a participatory evaluation of community participation: • the overall goal is political action local people control definitions of health problems, information gathering, and all • decisions about action • local people and professional researchers are equals (learners) • Participatory evaluation uses methods derived from a Brazilian educator, Paulo • Freire, who suggested that exposing the contradictions in community processes was the best • way to learn from experiences in order to develop further.12 Michael Bopp discusses many ·~ of the issues in conducting a participatory evaluation of participation itself! 13 • E. METHODS OF EVALUATION • A number of discrete methodologies are used to evaluate community participation, such as descriptive analyses and measurement scales. This session will focus on two scales for • assessing community participation. The methods are generic and, although developed in the • areas of urban renewal and health, can be applied in other areas as well. • Review the two scales, clarifying methods as necessary. Then have session participants form groups to apply the evaluation methods of community participation to the • case studies handed out previously. Make sure both cases are evaluated using both methods. • 151 • • • •

• 4 • Rifldn Scale to Assess Community Participation in PHC Programmes • This scale was developed in the 1980s and applied in Latin America, Philippines, Thailand'. • It has been adapted by Shrimpton, 14 Taal, 15 and Boyce. 16 The original Rifkin scale assesses five aspects of direct participation, which is • instrumental involvement in project activities. It does not assess social participation, which • includes influence, control, impact, and social equity . The scale poses a five-point continuum of indicators from "wide" participation in ,.• which community members plan, implement, and evaluate the program using professionals as resources, to "narrow" participation, in which professionals make all decisions with no ! • involvement of community members other than as beneficiaries, clients, or patients. This :. scale primarily focuses on the breadth of participation, on the question of who was involved • and controlled the program . • Groups assign an indicator to five direct participation factors. Then one can compare • ratings at different times or by different groups to suggest areas for improvement. Rifkin • suggests that different projects cannot be compared to each other due to their individual • natures . ,. Direct Participation Factors

• a. needs assessment: e.g., roles of professionals, leaders, health committees, community • members in assessing needs • b. community organiz.ation: e.g., health committee imposed from outside or organized from within • c. management: e.g., community health worker supervised by professionals or the health • committee • d. resource mobilization: e.g., funds/resources from outside or raised locally • e. leadership: e.g., project dominated by traditional leaders or a local committee • representing the entire community • Shrimpton expanded the scale to include other factors. 14 • f. training: e.g., location and extent of training for health workers • g. orientation: e.g., curative or preventative focus • h. monitoring, evaluation, information exchange: e.g., availability and use of feedback ••• system • • 152 • • • • Taal has added factors related to external supports for participation in regional/national 15 • programs. • 1. institutional support: e.g., support and coordination from the state • J. decentralization of decision making: e.g., from centralized to local decision making • Boyce has added factors related to external supports and social participation, or • empowerment. 16 • k. formal participatory mechanisms: e.g., presence of regulations and legislation supporting participation • 1. role of target group members: e.g., inclusion of disadvantaged groups • m. influence of community group: e.g., relative control of community and project • organizers • n. impact on local conditions: e.g., range of impact on health, knowledge, behaviours, environment • o. beneficiaries: e.g., degree of benefits to target group • • • • • • • • • • • • 153 • • • • l • Overhead ; .• • RIFKIN SCALE • • ,.• ,. Management ,• 5 4 3 ,.j • • , .• • • • • 5 • • • • • • 154 • • • Overhead • Arnstein Ladder of Participation • This analytic tool was developed in North American urban renewal programs in the 1960s. 10 • It is primarily a ranking method used to describe programs on an eight level scale. It assesses the level of influence and control by community members. • • LADDER OF CITIZEN PARTICIPATION • • Citizen control • 8 ., Degrees of Delegated power • 7 cl tlzen power .,• Partnerah Ip 6 • Placation • 5 • Conaultatlon Degrees of • 4 • token lam • Informing 3 • Therapy • 2 • Manlpulatlon Nonpartlclpatlon • 1 • • • Source: See Arnstein, Ref. 10 • 155 • • • • • After reviewing participation in a project through documents, interviews, and observation, 1. the group selects the position on the ladder which best reflects the community's level of • influence. The ladder is intended to allow comparisons between projects on this one • dimension of participation . • a. Citizen Power • citizen control: disadvantaged groups obtain majority of decision-making positions, or • full managerial power • • delegated power: groups select representatives as delegates • • partnership - groups negotiate and engage in trade-offs with traditional powerholders • b. Tokenism • • placation: groups can advise, but powerholders decide • • consultation: groups are requested to give opinions •• • informing: groups are informed of decisions • c. Non-participation • therapy: groups are encouraged to change themselves • • manipulation: groups are told what to do • • • • • • • • • 156 • • • • • • • • ••• . ; • • • • • • • • Feedback from Ses.sion Participants • Have the session participants answer the following questions: • • Did the session clarify terminology and concepts of community participation? • • Did the session provide you with tools to analyze community participation? • Did the session provide you with tools to assess the value of different evaluation • methods for participation? • • 157 • • • • • • COMMUNITY PARTICIPATION SCALE • • Indicator Nothing/Narrow Restricted/Small Mean/Fair Open/Good Wide/Excellent EMPOWER no target group target group 1-2 target target group majority of CHC • MENT members on CHC members involved group members are target group • Role of target or involved in on CHC or in members involved on members who group implementing implementing involved on CHC as a plan and • membert1 program; program CHCin minority in implement the • benejidaries only planning and planning, program implementing implementing • program and evaluating • program Influence of program program program program program • community controlled by controlled by controlled by jointly initiated, grou~ professi.onals or outsider NGO; local NGO; planned and planned and • bureaucrats; consultation only information implemented implemented by • paternalistic with selected sharing with by NGO and CHC with community CHC CHC; support of NGO; • members options CHC determines • presented to options for CHCjor action • dedsion • making • Impact ofCP no impact on limited short tenn impact on impact on long tenn impact on local causes/effects of impact on effects knowledge and behaviours on structural • conditioni1 poor health and ofpoor health attitude contributing conditions; • disability in and disability; components of to poor legislative and community curative focus poor health health and regulatory • and disability disability changes • PLANNING imposed; no imposed; limited CHCis existing local local groups • AND active local local imposed, then organizations initiate and MANAGE organization organization links becomes very are used control • MENT support active organization • OrganirJltion° • Leadership" organization CHW acts CHW consults active CHC CHC represents • dominated by independently with CHC on support; joint full spectrum of local elites actions initiatives local interests, • with CHW which direct • CHW actions • • • 158 • • • • Resource no resources no fundraising; occasional occasional CHC regularly • mobilizationa given by service fees; fundraising fundraising raises funds, community,· no CHC has no from part of from all collects fees, • fees for services control over community; community; controls funds • funds CHC collects CHC fees but does controls • not control funds but • funds does not collect fees • Management' professional staff CHW CHC CHC CHW • manage program independently manages; no manages,· responsible to • and supervise manages,· supervision of some and actively CHW professional staff CHW activity involvement supervised by • supervise CHW in CHW CHCin • supervision partnership with external support • Orientation of 1IO clear goals,· process-oriented impact- impact- impact-oriented • progromb mainly curative goals,· curative oriented goals,· oriented goals,· balance • in nature heavily curative goals,· between curative emphasized over emphasized curative and preventive • preventive over partially • preventive emphasized over • preventive • IMPLEMEN No needs Needs assessment done by community community • TAT/ON assessment done by outsiders outsiders,· assessment,· assessment, without CHC discussed with outsider analysis and • Needs involvement CHC, whose helps in choice of action interests are analysis and • assessment' considered choice of • action • Trainin/ little or no CHW CHW training in CHW training short local short local CHW training; training in local CHW training,· regular • remote in unfamiliar institutions,· no institutions,· training,· in-service • language in-service occasional in- regular in- training by local training service service trainers • training by training by • outsiders outsiders • Monitoring, no information outsiders are information CHWhas CHCandCHW ttialuation and system or aware ofproblem system used in access to have access to • information exchange,· no dimensions and daily CHW information information • exchangeb awareness of program activities,· system,· CHC system,· CHC problem progress,· 1IO CHWaware of promotes promotes • dimensions or feedback to CHW problem community community • program progress dimensions awareness of awareness of and program problem problem • process dimensions dimensions and • and program program process process • 159 • • • • • • Beneficiaries benefits to non- benefits to benefits to benefits to benefits to most of programd target group preferred target 71Unority of majority of needy target • members (elites, group members needy target needy target group members; • professionals, and non-target group group benefits NGO staff, group members members members generated for • bureaucrats) broader • community • EXTERNAL little or no centralized training and support from coordination of SUPPORTS institutional development support at district, program with • FOR support from support district level NGO's, and district, NGO's, • REGIONAL district/state accountable and local AND local organizations • NATIONAL organizations • PROGRAMS • Institutional • support • decision making Decentrali no coordinated loose top-down clear top-down joint decision • zation of decision making links; unclear links; making and by accountable decision between links with district decisions on pooling of local • makinT community, and state allocations are resources organizations • NGO's, district do71Unated by with support and state the centre from • district/state • Fonnal no regulations or occasional accepted regulations legislation • participatory accepted consultation with practices of regarding mandating mechanism! practices community community community community • groups before consultation representatio representation; • decision making and n on decision representatives participation making elected by • in decision bodies; communities • making representativ bodies; es appointed • representatives by • appointed by community central leaders • authorities • CHC = Community Health Committee • CHW = Community Health Worker

• 8 = Rifkin (1988) • b = Shrimpton (1989) • c = Taal (1993) • d = Boyce (1995) • • 160 • • • • REFERENCES • 1. Bichman W., Rifkin S.B., Shrestha M. (1989) Toward measurement of community • participation. World Health Forum, Vol 10 (3/4): 467-472. • 2. Labonte R. (1988) Health promotion: from concepts to strategies. Healthcare Management Forum 1:24-30. • 3. Midgely J., Hall A., Hardiman M., Dhanpaul N. (1986) Community Participation, • Social Development and the State. London & New York: Methuen & Co. • 4. Rifkin S.B., Muller F., Bichmann W. (1988) Primary health care: on measuring • participation. Soc Sci Med 26:931-940. • 5. World Health Organization. (1978) Primary Health Care. Report of the International • Conference on Primary Health Care, Alma-Ata, USSR, 6-12 September 1978, • Geneva, p. 3. • 6. UNICEF. (1992) The Bamako Initiative: Progress Report. (E/ICEF/1992/L.6). New York, NY: United Nations. • 7. World Health Organization. (1988) Alma-Ata Reaffirmed At Riga. A Statement of • Renewed and Strengthened Commitment to Health for All by the Year 2000 and Beyond adopted at a WHO meeting: From Alma-Ata To The Year 2000: A midpoint • perspective. Riga, USSR, 22-25 March 1988; p. 12. WHO. • 8. UNICEF. (1992) UNICEF Health Policies And Strategies: Sustainability, Integration • And National Capacity-Building. (E/ICEF/1992/L.7). New York, NY: United Nations. • 9. Werner D., Bower B. (1982) Helping Health Workers Learn: A Book of Methods, • Aids, and Ideas for Instructors at the Village Level. Palo Alto, CA: Hesperian • Foundation. • 10. Amstein S. (1969) A ladder of citizen participation. J American Institute Planners • 35:216-224. .,

11. Cohen J.M., Uphoff N.T. (1980) Participation's place in rural development: seeking • clarity through specificity. World Development 8:213-235. • 12. Freire P. (1970) Pedagogy of the Oppressed. New York, NY: Seabury Press. • 13. Bopp M. (1994) The illusive essential: evaluating participation in non-formal • education and community development processes. Convergence 27 (1):23-45. .,• 161 • • • • •• 14. Shrimpton R. (1989) Community Participation in Food and Nutrition Programmes: • An Analysis of Recent Government Experiences. Ithaca, NY:Cornell University . • 15. Taal H. (1993) Decentralization and Community Participation for Improving Access • to Basic Services: An Empirical Approach. Florence, Italy:UNICEF Child • Development Centre . • 16. Boyce W. (1995) Evaluation Instruments for Community Based Rehabilitation . • Kingston, ON: ICACBR, Queen's University . ,•• • RECOMMENDED READING • 1. Peter Oakley et al. (1991) Projects With People: The Practice of Participation in • Rural Development. Geneva, International Labor Organization . i.• :•• • , .• , .• • • • • • • • 162 • • • • • CHAPTER 6 • PRACTICAL ISSUES IN HEAL TH RESEARCH • IN DEVELOPING COUNTRIES • C. Larson • OBJECTIVES ,• In planning and conducting field research, key issues significantly impact on the success of a • research project. These include: • • project feasibility ,. • establishing realistic time lines :. • preparing a budget • ethical issues • • selecting the research staff • • setting up the research team • maximizing participation • • anticipating project management problems • The objective of this session is to sensitize trainees to these and other practical issues, which can • affect in important ways the successful completion of an investigation . • TEACHING STRATEGIES • Handouts - to be prepared before the session • • Workshop notes: A brief set of notes outlining the topics covered and the case history • will reduce note taking during the session, providing more time for trainee participation • and interaction with the instructor. Sample workshop notes are provided in Section 1 of • this chapter. Modify them as necessary . • • Examples of field research documents: Either use those provided here in Section 4 or • bring your own from a research project you have been involved in. These might include: • sample size estimates • • sample questionnaire(s) • • coding manual ,. • data collection schedule , .• • 163 • • • • • Research protocol outline: Trainees will find an outline of a research protocol helpful • as a general guideline and "checklist" while planning an investigation. See Section 7 of this chapter. ••• • Readings/references: These are not required for the completion of the course but are· • useful and appreciated. They are usually read after the course or when a trainee embarks upon a research project. See the end of this chapter. • • • • • • • • • • • • • • • • 164 • • • • • • SUGGESTED SCHEDULE • This session is best taught by one instructor and one or more small group facilitators . • Facilitators will enhance the flow of information from the instructor to the small groups and • provide ongoing feedback to the instructor during the session . • 10-15 min Give a brief overview to orient the participants and to place the session in • perspective vis-a-vis an entire workshop, unless the session is to stand on its own. Start by asking the participants what they hope to achieve in the • session. Using your own experiences, point out that much more attention • should be given to managing and monitoring a project. Follow this with an outline of what is to be covered. The quality of participant contribution and • interaction with the instructor and facilitators largely determine the success • of this session . 10-15 min Present one or more case histories based on your personal experience. This • puts the focus on real problems of field research. Trainees must understand • that most problems can be anticipated, but that ideal solutions are not always ,• available. The case history provides the baseline information for small groups to deal with concrete issues. A sample case history is provided in • Section 2, or you may provide your own. If you provide your own, make ,. sure that you adjust the questions in Activity 1 so that they fit with your case history. You may hand out these case histories before the session and/or • have them available for each group . 30-45 min Small group problem solving. Depending on the number of trainees in the • session and time available, from four to six trainees should participate in a • group. The group should begin by designating a recorder and a • spokesperson, who can be the same individual. Each group is given an issue to deal with and must answer specific questions. Have these prepared • in advance and written on a sheet of paper that the group can refer to. See • Activity 1, Part I and II. They should be encouraged to examine the issue • generically, and bring up considerations based on their own experiences . It is helpful if the instructor and facilitator circulate among the groups to • clarify questions and provide additional information as necessary. Example • questions are found in Section 2 . • 20 min Small group presentations and feedback. Presentation by each group to the • whole class and feedback. Timing is important as no group should be cut • short or have to rush through topics. The instructor should encourage full interaction between the presenting group and the rest of the class to draw on • the wealth of experience of the participants. Recording key points on a flip • chart or blackboard for future reference is helpful. If you have time to do both Part I and II of Activity 1, it is useful to have each group present Part ,.• I to the whole class before returning to their small groups to do Part II. ' • 165 • • • • 1. WORKSHOP NOTES: PLANNING AND CONDUCTING A HEALTH ~EARCH • PROJECT • 1.1 PLANNING A HEALTH REsEARCH PROJECT • • , In planning a health research project, a researcher must consider the following points to assess . the project environment and the approaches available to the research team to best carry out the project. • A. Project feasibility: Initial considerations • • time required and time available • • access to a study population and sample size requirements • • human and physical resources required • study population's acceptance of research methods to be used .,• B. Local and institutional supports • Individuals or groups with a vested interest in a study and/or the outcome can make or break most population-based investigations. A research must consider: • • who they are: local leaders, health practitioners, regional or national program directors, • academics • what they will gain or lose • • the best way to deal with them: information, coordination, project modification, • participation • C. Realistic time lines .,• Consider the following to calculate a project time line: • • start-up: hiring, training, piloting • enrolment of study population • • conducting the study • • data analysis • • write-up and dissemination • D. Budget preparation • The following categories or line items must be included: • • personnel (title, hours, rate, cost per annum) • • equipment and supplies (include justification) • • travel (conducting the study and dissemination of results) • 166 .;• • • • • • • analysis (hardware, software, data entry, consultants) • communication (conducting the study, management, dissemination) • • shared costs • • other costs • E. Consent and ethical issues • • rights of individuals • • rights of communities/populations studied • • cultural influences • • international conventions • F. Selection and training of research staff • job descriptions and clear expectations • • who will select • • qualities vs qualifications ,• • reliability • blinding • • confidentiality • G. Creating a research project team ,.• • selection based on quality, relevance, what is expected (expertise) • beyond researchers, who to consider, e.g., • - health practitioners • - those being studied - planners, policy makers • • set clearly defined roles and expectations from the outset • • clarify authorship(s) • • 1.2 CHOOSING A S1UDY POPULATION • Population Characteristics The researcher must be aware of the characteristics of the study population to be able to make • generalizations from the data gathered. For example, urban and rural populations have different • characteristics and therefore, usually, should be studied separately. Institution-based populations • such as hospitals, clinics, or schools may be convenient sampling sources but are potentially biased and will restrict generalizability. Even randomized, population-based studies can be ,.• biased. Bias can occur because of: • • 167 • • • • • repetitive sampling within a single community • • limiting the sampling of a community, e.g., to those living close to a road • selecting communities that are chosen or volunteered by local regional authorities • Random vs Systematic Sampling • While the preferred choice, randomization is not always the most practical choice. A population • must be enumerated to do random selection, otherwise many potential subjects do not have an • equal opportunity of being selected. Systematic sampling can be quite effective, particularly at • the household level. In some cases this can be followed by random selection of household members. If a community is well-defined and not too large, a researcher could enumerate • households and follow this with a random selection of household numbers. • Whatever system is used, a researcher should anticipate the common problem of non­ • response, i.e., nobody home or household heads absent at the time of the visit. This can be minimized by visiting households during non-working hours, avoiding periods of the year when • people are the busiest, and working out a plan for repeat visits. • Sample Size Requirements • Generally, a researcher should add 10 to 15 % onto the calculation of the sample size. The cost • of adding 15 % is not great and it is better to err in having too many subjects rather than too • few. If selecting communities, it is advisable to have an "alternate" community as a back up, • because a selected community could be lost due to rain, lack of local support, or security problems. • • 1.3 PROJECT FEASIBll..ITY: REvlsmNG THE lsSUE • Once you have set clear objective(s) and methods for your project, ask, "Is the project still • feasible?" A researcher should return to this question at several times during the development • of a study protocol and during the early stages of a study. You should consider the following • points when assessing feasibility: •a 1. What do you know about the current status of exposure(s) and outcome(s) within the target population? • Let us take the example of a project on infant feeding, where the objective is to see the • effect of prolonged breast feeding on the nutritional status of children. If all mothers or no • mothers practice prolonged breast feeding, then you will not be able to answer the question.If 11 11 • you were asking, At what age is malnutrition most prevalent and in what form, but • malnutrition is rare in the source population, the study will not be relevant to that population, as well as not feasible. • • 168 • • • • • • 2. What do you know about important covariates (confounders or effect modifiers) in the • relation being studied? • For example, in a study of malnutrition in relation to prolonged breast feeding, covariates would include social and economic status of the family, level of education, health of mother or • child, food supplements (what and when), or other factors which influence a mother's decision • to stop breast feeding. If these key covariates were not monitored during the study, the study • would be vulnerable to misinterpretation of the results . ;.• 3. Will it be possible to measure the exposure, outcome, and key covariates? If not, are • there proxies which might be used? When looking at measurements, consider: • • time required to conduct measurements and the required length of follow-up • • availability of required physical and human resources • • acceptability among those studied • 4. Will it be feasible to obtain the required support? Consider support from the: • source population • • government agencies • • training facilities , .• • laboratory facilities • 5. Are the costs justifiable in an absolute and relative sense? Consider: • • direct costs of investigation • indirect costs to those participating , .• • 6. Are there time constraints which can be met? 7. Are changes occurring in the study environment or will they occur during a study? For ,.• example, changes in policies or services that will affect exposures and/or outcome in an • intervention study are essential to anticipate and monitor . • 8. Are you able to follow the study protocol? One typically puts this question to test during • a project's pilot phase . • • • • 169 • I • I • • 1.4 CONDUCTING A HEALTH REsEARCH PROJECT • In conducting a project, the following items must be considered carefully: • A. Roles of principal vs co-investigators: • • supervision of coordinators • • protocol adjustments • • financial accountability • B. Day-to-day coordination and staff supervision: • • reliability of data collectors • • travel and logistics • • staff-participant interaction • data management • C. Trouble shooting and crisis management: • A researcher can anticipate problems and should have a plan for any of the following: • • insufficient enrolment • • data collection errors • loss of confidentiality • • cost over-runs • • salary demands • D. Calculators, computers and commonly used software: • • user preparation and quality control • • for descriptive and bivariate statistics relatively less sophisticated statistical • software, such as EPI-INFO, is adequate • • advanced PC program: SAS-PC, SPSS-PC • • keep two back-up copies of everything, and in different places • E. Communication/Coordination: • • research team (e.g., principal investigator, manager, etc.) meetings: frequency and contel • research staff (e.g., interviewers, data collectors, etc.) meetings: frequency and content• • regular meetings with community leaders, other stakeholders e F. Cut-off rules: • During the study, it may become clear that the study has reached its objectives or that it cannot • do so. At this point, the research team must decide whether to continue, so that time and money • are not wasted. • • , 170 . • • • • 2. CASE HISTORY ,.• This is a brief overview of a project proposal to examine the relationship between prolonged • breastfeeding (PBF) and early childhood growth . • 2.1. BACKGROUND • Breast milk is universally accepted as a hygienic, nutritionally • balanced source of food in infancy. The WHO and UNICEF have • vigorously advocated breast-feeding, particularly in less developed countries, as it offers the best source of protein and micronutrients ,.• at low cost, plays an important role in the prevention of infection in early childhood, and is associated with a reduced risk of infant i. and under-five mortality. Over the past decade several investigators have reported an adverse relation between PBF (greater than 12 • months) and child growth. Among most maternal and child health • professionals such an association is surprising. Is this association • real or does it simply reflect that PBF is closely linked to poverty, ,. low education, and disadvantaged environments?

• It has been argued that the finding of an association of PBF with poor childhood growth • is the result of inadequate control for confounders and lack of the assessment of the role of effect • modifiers. Yet, several studies which have controlled for socioeconomic status, mothers' reproductive history, environmental sanitation and host factors (age, sex, and health) have found • repeatedly -- though not consistently -- that the duration (duration could be < 12 months) of ,.• breast-feeding is inversely related to nutritional status. More recently, co-factors such as weaning histories and supplemental energy intake have been assessed. Few reports have been published, but one study conducted in a rural Chinese population has found that when ,.• supplemental food intake is controlled for breast-feeding, it is positively associated with early ,. childhood growth . • An alternative explanation for the PBF-malnutrition assoc1at10n is that of reverse ,. causality. In this case, because a child is undernourished the mother decides to prolong breast­ feeding. Two cohort studies that looked at the relative risk of discontinuing breast-feeding • among severely malnourished infants reported contradictory findings. One study reported an • increased risk and the other a decreased risk of discontinuing breast-feeding . • There are several methodologic limitations in the published studies on the effects of PBF . Most studies are cross-sectional in design, e.g., community-based surveys that compared • exposure histories between those with and those without malnutrition at a single point in time. • These studies did not confirm the directionality of the relationship between nutritional status, the ,. decision to continue or discontinue breast-feeding, and the timing and/or quality of weaning ,. supplements. A second limitation is that, while most studies attempted to control for important potential confounders, they are incomplete and varied. Preferably, an investigation should ,.• prospectively include and monitor key covariates such as host factors, social and demographic characteristics, mothers' reproductive histories, environmental sanitation, nutritional histories • (including weaning, supplements) and food availability and security (food types, quality, • quantity, and reserves) . • 171 • • A third limitation to research in this field is the absence of growth curve standards for breastfed children. Breastfed infants have distinctly different growth curves over the first two • years of life when compared to the NCHS-WHO reference, which is based upon middle class, • non-breastfed, white infants. Breastfed infant growth curves are characterized by more rapid • growth in the first 2 months, less rapid growth from 3 to 12 months, and catch-up growth by • age 2 years. It has been demonstrated that the growth patterns of healthy breastfed infants are remarkably similar across populations. • A final limitation is that most observational studies are conducted in populations with • minimal primary health care coverage. It is far more relevant if participating communities are provided with the minimum of a safe water source, an expanded program of immunization • (EPI), a maternal-child health (MCH) program, and functional-trained community health • workers. This allows one to identify important predictors of malnutrition within an environment • that is more consistent with internationally accepted standards of basic health service. • It is proposed to carry out two similarly designed cohort studies, one urban and one rural, • to examine the relationship between the duration of breast-feeding and the nutritional status of • children from birth through 36 months of age. Relevant nutritional, social, health, reproductive, • and environmental co-factors will be prospectively monitored. Factors that influence the duration of breast-feeding and the decision to stop breast-feeding will be looked for. Weaning • and energy supplementation practices (quality, quantity, duration) over time will be monitored • and assessed for their impact on growth. • 2.2. REsEARCH QUESTION • "Does prolonged breast feeding (PBF) lead to an increased risk of malnutrition?" • Focus your objective(s) to solve the problem. In this case you would need to clarify the • following issues: • • What is the definition of PBF? • • Should you consider all types of malnutrition or a specific type only? • • Should you look at malnutrition at all ages? • How much difference do we expect in nutritional status in children who are breast fed • for a prolonged period? • Definitions: • Prolonged breast-feeding: breast-feeding beyond 12 months of age • Weaned: the point at which the mother stops breast-feeding her child for seven or more • days • • 172 • • • • • Malnutrition: any child with a z-score equal to or less than -2 standard deviations from the • mean, based upon the WHO breastfed reference curves to age 12 months (if ,.• breastfed), or the National Center for Health Statistics (NCHS) age-standardized norms (if not breastfed). If the WHO reference curves beyond 12 months for • breastfed children are available, they will also be applied . • Note: z-score = Expected - Observed • Standard Deviation ~· Wasted: wt/ht z-score equal or less than -2 • Stunted: ht/age z-score equal or less than -2 • Undernourished: wt/age z-score equal or less than -2 • 2.3. RE'sEARCH OBJE~ • To answer the research question, "Does PBF lead to an increased risk for malnutrition," • separate samples of urban and rural children followed from 0 to 36 months of age will be ••• studied. The specific objectives are: • determine the relative risk for undernutrition, stunting, and wasting in relation to breast­ • feeding beyond 12, 18, 24 and 36 months of age after adjustment for biologic, health, • environmental (physical and social), behavioral, and nutritional covariates • • identify predictors of breast-feeding duration • • describe weaning and other nutritional practices • • assess the relative impact of health (mother and child), reproductive, nutritional, environmental (physical and social), behavioral, economic and health service • characteristics on the occurrence of malnutrition in children through 36 months of age 2.4. POPULATION ••: . Source Population: Any child born to a resident (> 12 months permanent residence) of the • local town or one of the 10 selected rural villages (plus one alternate). • There are estimated to be 2,000 live births in the town and 200 live births/village per year. , .• Study Population: An attempt will be made to enroll all uncomplicated live births (census • sample) over a period of one year. Given an infant mortality rate of 100 • deaths/1000 live births and a 10% rate of non-entry or loss to follow-up, • an expected 1600 urban and 1600 rural children will be followed through 12 months of age. Given an additional 10% loss/year, it is expected 1440 • and 1330 will be followed through 24 and 36 months, respectively . • 173 • • • • • • • • • • • • • • • • • • • • • 174 • • • • l '! • ! I I • I • I • I • l • I ,. I I l ! :. l • I • I • I • I i • I I • ! • I •• I • I •• I • I • I ••• I • ' •• I • I • I • • I ! • 'I • I ~ • I' • I • l • I I • 175 !' • I • l l • ' • ! • • 3. SMALL GROUP PRESENTATIONS/FEEDBACK • If not mentioned by the groups during their presentations to the plenary, these points may be • raised. • PART I: GROUP 1: • A. Generalizability: It would be safe to assume similar rural, geographically close districts. • Widened scales of application might include more regional or developing populations in general. • The case becomes weaker as the scale enlarges. This matters, because such studies are difficult to carry out and one would hope to maximize the benefits obtained from the knowledge derived • from the one study. • B. Population selection: You can enter into a discussion of the potential benefits of meeting • with local health authorities and community or district representatives. • Other topics: • • random vs. systematic selection (pros and cons) • unit of selection - communities vs individuals • • start with pregnant women or newborns • • clinic or household based enrolment • • who will enroll (TBAs, CHWs, research assistant) • C. Exclusion criteria: considerations include: • • health of mother (physical and mental) • • birthweight and health of newborn • length of residence in community • • multifetal pregnancies • PART I: GROUP 2: • A. Research team: consider the following: • • expertise in study methods/analyses • • expertise in early childhood nutrition, anthropology • community representation • • local practitioners (consumers of information). • B. Advisory group: • • policy/program makers • • MOH, Institutional representation • • funders • • ombudsman • 176 • • • • • PART I: GROUP 3: • A. Vested interest: • • other researchers/experts, task forces • Ministry of Health: MCH or Nutrition Divisions • • policy makers • • local communities • • pregnant women, mothers of infants and young children • • health workers, including midwives • • artificial milk producers, merchants . • B. Dealing with vested interests: Options range from: • do nothing • • keep informed ,.• • ask for advice/ consultation ,. • actual participation in study \. Consider each of the above at separate stages in the study, i.e. planning, conduct, analysis, • write-up and dissemination of results, and application of results . • PART I: GROUP 4: ,.• Selecting an Interviewer A. Interviewer qualifications: consider the following:

• • female • available • • discrete • literate, languages spoken • • likeable, trustworthy • articulate • in good health • similar life events :.• • a mother • pastexperience • • not member of same community • B. How to select and who selects: • This should be done to protect equal opportunity, and avoid favouritism and nepotism. It is best • done as a group responsibility rather than by a single individual. Create a list of candidates first • and avoid premature promises of employment. Be clear about the job description, work • conditions, and remuneration . • As part of the job interview, you could ask candidates to carry out an interview and • another task that they would be expected to do in their job . • 177

, • .• • • C. Pay: • • be consistent with other local employers • • adjust for fact that no pension is provided • • it is generally best to pay by salary, but there will be exceptions • PART I: GROUP 5: • A. Ethical is.sues: • • what are the anticipated, direct benefits to participating communities? • • are basic health services in place (safe water, EPI program, antenatal care)? • • confidentiality: what and how? • • any potential harmful effects? • B. Consent procedure? • • who will provide consent: community leaders, husband, mother, others? • • what should participants be informed of? • explain fully what will occur if they agree to participate • • explain that they are free to consent or not without affecting services received • • verbal vs written consent • PART II: GROUP 1: • Pay increase: This is a common problem. It is best to clarify expectations and criteria for change from the outset. Hold firm! Study request carefully and discuss among research team • members and with other research groups. Unless there has been a change in external • circumstances warranting a salary change, stick with the original agreement. • PART II: GROUP 2: • Underenrolment: This is a common problem. If you can increase the enrolment numbers • without threatening the integrity of the study then do so. This can be done by expanding the source population or selecting additional communities. Enrolment may be low because of a • misunderstanding during the consent phase of the project. • There will be situations where nothing can be done, in which case you should consider • ending the study. • • 178 • • • ,.• l I • ! • PART II: GROUP 3: • Competing interventions: This represents a potential problem, but also a potential asset. Any I! • competing intervention must be well documented and accounted for when analyzing results. By • working together, it permits the additional assessment of food supplementation and its interaction • with other factors influencing breastfeeding and early childhood growth. I.• PART II: GROUP 4: Heavy rains: Because of delays in data collection, you could reach a point where it is no longer • possible to obtain reliable information. This frequently happens with individual participants, but • may also involve an entire community. One "insurance" manoeuvre is to add an alternate, back­ • up community to the study. Communities can also be lost because of loss of support by community leaders, loss of data due to theft or fire, and the application of improper selection ,.• procedures . • PART II: GROUP 5: ! • Unusual data: Data is either accurate or it is not. If not, this could be differential error • (pertaining only to specific subgroups, i.e., breast feeders are differentiated from non-breast • feeders), or non-differential error equally effecting all participants (such as one due to a faulty • scale, or misworded question). You must consider whether the discrepancies are deliberate; for • example, completion of questionnaires based upon false interviews, i.e., the interviewer never completed the interview. Where error can be confirmed, data is either corrected, treated as • missing, or, in the extreme, entire subjects dropped from the study . • This type of problem can be minimized by frequent monitoring of incoming data and periodic observation of data collection procedures, including interviews. Inter-observer 1.• reliability testing at the outset and mid-way into the study will also detect inconsistencies. j • • • • I • I • I • ! • 179 I • • I • • 4. EXAMPLES • Presentation of Sample Size Estimates • Sample size in the PBF proposal: The desired level of confidence was set at . 95 and the power • of the study at 0.80. • SAMPLE SIZE CALCULATIONS • Comparison of two means •

2 2 • Number I group = (Za + Z8l (SD) x 2 (d)2 • Comparison of two proportions • Number I group = 2 PCl-Pl x 2 (d)2 • • Za : desired level of confidence (usually set at 0.95; Z.95 = 1.96) • Z/3 : desired power (usually set at 0.80 or higher; Z.m = 0.84) • (1.96 + 0.84)2 = 7.84 • SD : standard deviation • P : overall prevalence • d : minimal detectable difference • • • Using the formula in the above box for the comparison of two means: to detect at any • point in time a mean difference of 0.5 SD (z-score) in anthropometric status between • breastfed and non-breastfed subjects will require 63 subjects/group (let SD= 1). If we • were to reduce the minimal detectable difference to 0.25, we would require 251 subjects/group. • • Using the formula in the above box for the comparison of two proportions: to detect at • any point in time a 15 % minimal detectable change in the prevalence (relative risk of 1.5 • or greater) of stunting based upon an overall prevalence of 30% would require 146 subjects/group. • • 180 • • • • • • Based upon a recent feasibility pilot survey, anticipate most urban mothers will • discontinue breast-feeding between 12 and 24 months, while the majority of rural • mothers will stop between 18 and 30 months. The estimated number of women breast • and non-breast feeding over time (including adjustment for losses) are summarized in • Table 3. From this table it can be seen sufficient numbers of children to detect a 0.5 SD group difference in mean z-scores are present from 0 to 30 months among urban and 18 •• to 36 months among rural children . • • Table 3. ES'r.IMATED NUMBER OF SUBJECTS BREAST AND NON-BREASTFED OVER TIME • I #Subjects I • Entry 12 mo 18 mo 24 mo 30 mo 36 mo • Urban: • breastfed 1800 1100 750 480 240 0 non-breastfed 200 500 750 960 1080 1240 l • Rural: ••• breastfed 2000 1550 1350 1200 660 240 • non-breastfed 0 100 150 240 660 1000 , .• • • • • , .• • • • • 181 • • • • SAMPLE QUESTIONNAIRE - JIMMA PROLONGED BREAST FEEDING STUDY BASELINE ANTENATAL INTAKE INTERVIEW • ID Number --- Date ____ Interviewer------• Failier ______• Family Names: Moilier ______Child------• Address: Kefitegna:_ Kebele: House Number: __ • CHILD • A. Antenatal History • 1. Date of Birili 2. Gender: male female • 3. Lengili _._ cm 4. Weight_._ kg • 5. Lengili of Pregnancy: short_ normal_ long_ L weeks) • explain...__ __ • 6. Birth of child: problems? no_ yes_, • Since ilie birili has your child been seen • 7. by traditional healer? no_ yes_, explain...__ ___ • 8. by healili worker? no_ yes_, explain...______• 9. Is your baby in good healili? yes_ no_, if not, explain.______• B. Nutritional History • 1. Breastfed? yes_ no_, if not explain~------• a. If yes, how long do you plan to breastfeed your child? • • b. and currently breast-feeding _ times a day • 2. Are you bottle feeding? no_ yes_, if yes • a. bottle feeds contain what? ______• 3. In your opinion is iliere enough food for ilie family? yes, always_ yes, most of ilie time_ • no, most of ilie time_ no, never_ • 182 • • • • Sample questionnaire - p.2 I! • I • MomER I I • 1. Age (last birth day) __years. 2. Civil status: never married married I • divorced widowed I A. Reproductive History I • i • 1. # of deliveries 2. # of live children I • 3. DOB of youngest sibling ____ • I • 4. Were you sick during this pregnancy? No_ Yes_, if yes, explain....._ __ • B. Mother's Health • 1. Height_._ cm • 2. Weight_._ kg ,. 3. Arm Circumference_._ cm • 4. Do you now have any physical health problems? ;. no_ yes_, explain.______• • • Mother's Health During Pregnancy • Symptom Yes No Duration • Daily cough • Diarrhea • Weakness • Fatigue • Pain • Other • • • 183 • • • SAMPLE CODING MANUAL • JIMMA PROLONGED BREAST-FEEDING STUDY • VARIABLE CODING INSTRUCTIONS FIELDS • 1. ID number enter 4 digit # 4 • 2. Interviewer 1. interviewer #1 1 • 2. interviewer #2 etc. • 3. District enter kefitegna 1 to 5 1 • 4. Sub-district enter kebele 01-22 2 • CHILD • 5. Age enter age in months 2 • 6. Gender 1. female • 2. male 1 • 7. Length enter to nearest 0.5 cm 3 • 8. Weight enter to nearest 0.1 kg 3 • 9. Length of 1. short 1 pregnancy 2.normal • 3. long • 10. Birth 1. no problem 2 • 2. yes, problems • Birth Problems: mother • 11. prolonged labour 1. no 2. yes 1 • 12. obstructed 1. no 2. yes 1 • 13. excess bleeding 1. no 2. yes 1 • 14. other 1. no 2. yes 1 • Birth Problems: child • 15. breathing problem 1. no 2. yes 1 • 16. limp/weak 1. no 2. yes 1 • 17. other 1. no 2. yes 1 • 184 • • • • lI • I f I • I VARIABLE CODING INSTRUCTIONS FIELDS I • ! ! • 18. Traditional healer 1. no 2. yes 1 • 19. Health worker 1. no 2. yes 1 I! • 20. Good health? 1. no 2. yes 1 I • Health Problems (to be detennined) I ! • 21. 1. no 2. yes 1 • 22 . 1. no 2. yes 1 • 23. 1. no 2. yes 1 • 24 . 1. no 2. yes 1 • 25. Breastfed 1. no 2. yes 1 • 26. Duration breastfed leave blank if not • enter duration in months 2 • 27. Currently breastfed 1. no 2. yes 1 • 28. Currently bottlefed 1. no 2. no 1 • Bottle Contents • 29. juice 1. no 2. yes 1 • 30. milk 1. no 2. yes 1 • 31. cereal 1. no 2. yes 1 • 32. other 1. no 2. yes 1 • 33. Enough food 1. yes, always 1 • 2. yes, most of time 3. no, most of time • 4. no, never • MOTHER • 34. Age enter age in years 2 • 35. Civil status 1. never married 1 • 2. married 3. divorced • 4. widowed ,. 36. # deliveries enter# 1

37. # live children enter# 1 'I • [ ! • ! • 185 I • I • • VARIABLE CODING INSTRUCTIONS FIELDS • 38. age, youngest sibling enter age in months 3 • 39. sick in pregnancy? 1. no 2. yes 1 • 40. Mother's height enter to nearest 0.5 cm 3 • 41. Mother's weight enter to nearest 0.1 kg 3 • 42. Mother's AC enter to nearest 0.5 cm 3 • 43. Health problem now? • Health Problems (to be detennined) • 44. 1. no 2. yes 1 • 45. 1. no 2. yes 1 • 46. 1. no 2.yes 1 • 47. 1. no 2.yes 1 • Health During Pregnancy • 48. Daily cough 0. no • if yes, enter duration in # weeks 2 • 49. Diarrhea 0. no • if yes, enter duration in # weeks 2 • 50. Weakness 0. no • if yes, enter duration in # weeks 2 • 51. Fatigue 0. no • if yes, enter duration in # weeks 2 • 52. Fevers 0. no if yes, enter duration in # weeks 2 • 53. Pain 0. no • if yes, enter duration in # weeks 2 • 54. Other O.no • if yes, enter duration in # weeks 2 • • • 186 • • • • • DATA COLLECTION SCHEDULE • Measurement Schedule (Months) • Variable Instrument Birth 1-6 7-12 13-36 1. Nutrition • a. breast-feeding and • supplement history • + food availability BI+ FUI x qlmo q2mo q3mo • b. anthropometry Ht, Wt x qlmo q2mo q3mo c. feeding practices, • food availability BI+ UDI lmo 6&12 24mo • mo 2. Health and Health Services • a. Child: prenatal, antenatal, • postpartum care and health BI+ FUI x qlmo q2mo q3mo • b. Mother • i. physical health BI+ UDI x 6mo 12mo 24mo ii. mental health WHO-SRQ 1&6 12mo 24mo • mo • iii. reproductive history BI+ UDI x 6mo 12mo 24mo • 3. Social • a. income, education, literacy BI lmo • b. family support BI 1&6 12mo 24mo • mo c. family size, I under-five, • 1&6 12mo 24mo dependency ratio BI+ UDI • mo • 4. Environment a. source of water BI+ UDI 1&6 12mo 24mo • mo • b. sanitation BI+ UDI 1&6 12mo 24mo • mo • • BI: Baseline Interviews (alllenatal and 1 month) FUI: Follow-up Interview UDI: Family-Child Updated Interviews • WHO-SRQ: World Health Organization-Self Reporting Questionnaire • qlmo: every month/ q2mo: every other month/ q3mo: every 3 months • • 187 • • • 5. RESEARCH PROTOCOL OUTLINE • Summary: Write this section last and briefly summarize the problem (phenomenon, object) to • be studied, its relevance, the question asked, method, and expected application of results. Write • in a clear, easily understood style avoiding jargon and acronyms. Maximum length is 350 · • words. • Definitions: To avoid ambiguity, provide clear definitions of the key terms found in the text • of the protocol. • Introduction: This is a brief (one page) explanation of the phenomenon (object/problem) or • activity under investigation, its health impact or relevance, and why you are proposing this • study. • State of Knowledge: Present current knowledge about the phenomenon, its occurrence, distribution, determinants, impact, interventions; or activity -- what, why, how, who, when, and • results. Discuss particular strengths or weaknesses in the literature, relevant methodologic • issues, and existing gaps in knowledge. Examine the literature collectively and limit the critique • of investigations to those that highlight a particular issue or are pivotal to current knowledge. Address the significance and originality of the proposed study. After reading this section a • reviewer should be aware of the current state of knowledge, what important deficiencies exist, • and recognize the need for the proposed study. • Objective(s): State each objective concisely and clearly. Where indicated, include related • hypotheses to be tested. The remainder of the protocol should be a logical extension of the • stated objective(s). • Methods: • A. Study Design: Include the study design to be employed, the setting of the investigation, and • expected dates or time-frame. • B. Population: • • Source population: Describe the population pool from which subjects will be selected, • followed by exclusion and inclusion criteria. • • Study population: How will study subjects be selected from the source population? • Fully describe the sampling methods (selection procedures) to be used. • • Sample Size: Outline the criteria used in estimating the required sample size and cite the appropriate reference for the formula or software applied. Include clear assumptions • about desired power, confidence, and the precision of estimate. The final sample size • should also factor in anticipated losses during the conduct of the study. • 188 • • • • • C. Intervention: If an intervention study is planned, then describe it in detail. Provide enough • information to fully explain the content of the intervention, who is providing the intervention, • its timing-duration-intensity, and what is required of the subject and the research team. If • subjects are to be randomly or systematically assigned to one or more interventions, describe • the procedure to be followed . • D. Measurement: Describe all measurement instruments and their purpose, content, method • of application, by whom, and when. Cite any literature in support of a measure's validity and • reliability. A table or figure may simplify the presentation. If "blinding" is to be applied, • describe the procedure to be used. Append questionnaires . • E. Analysis: Include a brief description of what will happen to the data once it is collected and • procedures followed prior to beginning the analysis. Describe all analytic procedures (tests) to be applied to the data. This is easiest to follow if presented in the planned sequence; for • example, data cleaned, reduced, and then analyzed in terms of descriptive, followed by bivariate • and multivariate techniques . • Ethical Issues and Consent: Review all anticipated ethical issues and justify the conduct of the study. Include a description of what potential study subjects are informed of, what will be • expected of them if they agree to participate in the study, assurances of confidentiality and lack • of any penalties if they should choose not to participate or subsequently drop out of the study . Conduct of Study: Describe how the study will be managed and the sequence of activities to ,.• be followed. Each activity should be matched to an expected time-frame. The chronology is easiest to follow with a figure or table. Indicate who has participated in the development of the • research protocol, what and how, and future roles they will have. Append letters of support ,. and/or approval from key institutions, agencies or individuals .

• Budget: Include the categories listed below. Provide enough information to explain and justify • the budget. This might require a narrative explanation appended to the protocol. Justify all equipment requested. Budget categories include: : .• a. Personnel: Monthly full-time rate Hours/wk Duration Total • b. Materials/Supplies/Communication: • c. Equipment: • d. Travel: (include anticipated costs related to presentation of results) • References: Relevance: In this section, make a convmcmg argument/justification for the proposed ;.• investigation. Relate the proposed study to the health phenomenon or activity under ,. investigation and explain how the information attained will apply to understanding or solving the • identified problem or gap in knowledge . • 189 • • • • RECOMMENDED READING • 1. Woelk G.B. Cultural and structural influences in the creation of and participation in • community health programs. Soc Sci Med 1992;35:419-424. • 2. Ramalinaswami V. Health research, a key to equity in health development. Soc Sci Med 1993;36:103-108. • 3. Freeman W.L. Model volunteer consent forms for the Indian Health Service. Indian • Health Service, Ottawa, Canada, 1993. • 4. Mbuyi-Muamba J.M. Drug trials in developing countries: the need for ethics. World • Health Forum 1993; 14:28-29. • • • • • • • • • • • • • 190 • • • • • • CHAPTER 7 • LABORATORY ISSUES IN DEVELOPING COUNTRIES • C. Merritt • Some people in this course will have little or no background in laboratory work. If so, you should aim for a middle ground so that experienced people are not bored and those new to ,.• laboratory work are not lost. The facilitator should be a laboratory technologist with experience in developing countries, or with significant laboratory experience of a generalist nature . • This workshop should be run on two levels. On the content level, you can use the laboratory experiences of the participants to give an overview of the issues relevant to working effectively • with lab facilities in developing countries. On the process level, a case study can then be • presented to develop a laboratory research strategy incorporating the issues. I suggest that small • group role play is used to complete the task. Group work and analysis will be important aspects • at this level. • OBJECTIVES • Using a participatory approach to problem solving, we shall introduce health researchers to the • complex issues of working effectively with lab facilities in developing countries. We shall • cover: , .• • • appropriate laboratory technology ,•• • quality assurance • • personal safety and laboratory safety • i • • • We shall also introduce some techniques for group analysis during this session . • 191 • • • • SUGGESTED SCHEDULE • 30 min Introduction to Session - Participant Input • Briefly introduce session by relating a practical lab issue from your experience. Ask participants to relate their own experiences (good & bad) • from working in labs overseas. Write issues on a flip chart under four • general headings - sample collection, testing, personnel, safety. • 30 min The Issues - Summarize • When the input from participants is exhausted (supplement with your experiences) use the overhead on page 199 to summarize the issues. Add • information not drawn from group. • 15 min Description of Group Work (six to eight people per group) • Using the overhead on page 200 (also handouts), explain the case study and • role play. • The case study can be any lab-based research problem, as in one of the two • case studies included here, or one taken from the participants. Be sure to give enough scientific background about the problem that participants with • little lab experience can understand. • Explain role play and give out the Role Play Guidelines to the observers. • 45 min Small group role play and problem solving • Float among groups to get a feel for how the process is working. Make yourself available to clarify questions. • 45 min Reports and Discussion • • Case Study - Have someone from each group report on the issues • discussed and relate these back to the four areas of discussion in the • introduction. Write up on the flip chart. • • Role Play - Have the Role Play observer report on the group • interaction surrounding the role play. Write up on the flip chart. Summarize role play emphasizing group dynamics. • Distribute handouts on group work on pages 204, 205, and 206. • • 192 • • • • • • REFERENCE FOR THE FACILITATOR • Details of Issues: • Before the role play exercise begins, the facilitator should go over the issues in detail with the participants. Recommend that they read in the chapter "Field Laboratory Methods" from the • book by Smith and Morrow (1991), included as a reading on pages 211 to 219. Other good • sources such as the lab manuals for tropical countries by Monica Cheesbrough and the World • Health Organization, are listed in the Recommended Reading. Inexperienced participants would • find these books very helpful. • A. SAMPLE COLLECTION • Sample collection is often complex. The ethics of sampling must be considered. Cultural ,. sensitivity could be encountered about giving blood, stool samples, or using needles or other foreign equipment. Health researchers must make sure the participants understand the intent • of the study and obtain informed consent from the participants. If necessary, the participants • should also consent to the use of secondary data before sample collection begins . For example, from a participant. Blood samples were collected from the Cree • for a particular study on hepatitis B and then frozen. Several years later, • researchers were interested in testing the same samples for HN, after stripping • the samples of identification. Is this ethical? ,.• Exact procedures for sampling must be worked out. Important questions to consider are: ,• (a) What volume of specimen is needed for the test? Samples for seroprevalence testing may be pooled (Babu et al, 1993; Ko et al, 1992; • WHO, 1992). Microtechniques allow the use of small sample volumes (Arimitsu et al, • 1994; Brogdon, 1987; Gajanana et al, 1995; Hudson-Thomas et al, 1994; Kenyon et al, • 1994; Renganathan et al, 1995). Blood samples may be stored dry on filter paper and • later eluted for testing (Lillo et al, 1992; Guimaraes et al, 1985; Zwicker et al, 1990) . • (b) What equipment will be needed? How many and is it sterile? • The safety of both the sample collector and the participant must be considered . For example, from a participant. CIDA gave no money for latex • examination gloves in their budget but after an autopsy on a water • buffalo in Nepal revealed rabies, CIDA agreed to cover the cost • of gloves . • • 193 • • • • (c) Consider the labelling, transport, storage and perhaps export of samples. Labels must • be legible, indelible and often anonymous but traceable. Transport and storage may involve refrigeration, often a problem in developing countries. • For example, from a participant. Stool samples collected in Africa • were shipped unrefrigerated by air freight back to Canada for testing. Needless to say they were useless and very high! Export • of samples will require government approvals at both ends which • may take some time to process. • Once the above procedures for sampling have been worked out, put together a pilot kit • and try it out in the field. • B. TEsrlNG • To ensure reliable results, samples must be properly tested. It is important to use in-country • capabilities and infrastructure as much as possible, as well as appropriate technologies. A • list of appropriate test procedures recommended by the World Health Organization, which take • into account availability of reagents, lack of refrigeration and simple instrumentation, is • appended. Pilot test all techniques for validation of methods. • If supplies and equipment are sent from outside the country, details of import procedures • must be investigated. • For example, from a participant. The project involved setting up • a laboratory in Angola. The technologist had been informed that • the shipping container had arrived from Canada. When she arrived • she found out that the container was still in the port of entry, 200 km away awaiting customs' clearance and the only transport truck • available had a broken axle! The technologist waited a month for • the lab equipment. • Instrumentation must be compatible with the local power supply. Electrical equipment is usually built to run on either llO or 220 volts and, exceptionally on both. Plugging a llOV • centrifuge into a wall plug supplying 220V will ruin your centrifuge. Transformers that convert • current from one voltage to another are available but expensive, can break, and are sometimes • forgotten to be used, even when available. In some countries, surges of power can be considerable, and surge protection is also advisable. Power interruptions or even blackouts also • occur regularly in some countries. • • • 194 • • • • • • Regular maintenance of equipment and availability of spare parts are also important • issues . • For example, from a participant. The NGO I worked with had shipped an old chemistry analyzer to China for use in a hospital lab. It was not ,.• working, no spare parts were available and it was the wrong voltage. It was useless! There are references for appropriate instrumentation • included in the Recommended Reading . Documentation of every part of the investigation is vital. Records of sampling • procedures, specimens, testing procedures, test results, equipment maintenance, batches, • suppliers and expiry dates of reagents, and quality control tests should be kept in duplicate with • one copy kept off-site . For example, from a participant. We had all our data stored in • duplicate on paper and on computer disc but when the office • caught on fire we lost everything . Quality control procedures ensure reliability and reproducibility of results. Technologists • should perform blind assays with both internal and external controls. If some techniques are • performed in the field they should be randomly confirmed in the laboratory . • For example, from a participant. The field lab in Kenya is partnered to a provincial lab but often the procedures aren't • transferable and don't agree. References for quality control • procedures are included in the Recommended Reading and the • laboratory manuals mentioned have good sections on quality • control. • The allocation of resources i.e., whether to spend money on personnel vs. equipment, • is another issue. Investigate the local area to determine if lab facilities are underused and could • be recruited into your study. One participant suggested the creation of a 'mega-lab', sharing • facilities by animal and human studies to maximize resources and technical know-how . • C. PERsONNEL • The selection and training of staff is crucial to the success of a study. Once again, locally • trained technologists will be more familiar with their own working circumstances and can be • taught new testing procedures as needed. This often gives them a sense of ownership of the • results and leads to increased performance, reliability, and sustainability. Clearly outline the procedures and tasks both verbally and in written form. Include some method of monitoring • performance and evaluation. Remember to be sensitive to their other duties and increased work • loads . • 195 • • • • For example, from a participant. We used the laboratory in the local medical clinic and trained the technologist to perform the tests we required with excellent • results. He was given extra money for his added duties. • D. SAFEI'Y ••• The safety of everyone involved in the study is very important. Personal safety must be • maintained as blood, sputum, and stool may all carry dangerous infections. People in contact • with these and their containers should always wear latex gloves. Reagents used in the lab also • have the potential of danger from bums and poisoning, and standard laboratory procedures such as wearing lab coats and no pipetting by mouth must be enforced. Safety guidelines for • personnel working with lilV are given in the Recommended Reading and the lab manuals cited • have excellent sections on lab safety. • The community must also be kept safe. Sample collection should not introduce new infection into the community. An adequate supply of clean and sterile collection equipment • should be provided. Disposal of wastes such as samples, reagents, gloves, needles and syringes • is best done by incineration or burial if incineration is impossible. • Benefits to the Community • Finally, the community will be more supportive of your investigation if they receive some • benefits. Examples include treatment of diagnosed illnesses or free insecticide-impregnated bed • nets in a malarial prevalence study. They should also be informed of the results obtained in the study. • • • • • • • • • 196 • • • • • • LIST OF LABORATORY INvES'TIGATIONS & TECHNIQUES • THAT SHOULD BE AVAILABLE IN HOSPITALS • (RECOMMENDED BY WHO) • lnvesti2ation Technique • • Haematology -Hb(colorimetric) Haemoglobin cyanide -Total white cell count Haemocytometer • -Differential cell count Thin film, Romanowsky Stain • -ESR Westergren Tubes • -Haematocrit Capillary centrifugation -Sickle cell screen Sodium Metasulphite Test • -Reticulocyte test Brilliant Cresyl Blue • -ABO grouping Tile • -Rh grouping Tile & tube :. -Cross matching Saline tube RT Saline tube 37°C • Albumin tube 37°C • Urinalysis -Specific Gravity Urinometer -pH Paper strips • -Deposits Concentration, Microscopy • -Glucose Dip sticks/Benedict's Reagent • -Protein Dip sticks/SSA -Bilirubin Dip sticks/Lugol's or Fouchet • -Blood Dip sticks • -Ketone Dip sticks • -Urobilinogen Dip sticks • -Pregnancy Test HCG detection • Cerebrospinal Fluid • -Leucocyte Count Haemocytometer -Differential White Cell Thin Film, Romanowsky Stain • -Glucose 0-Toluidine Test Kit • -Total Protein Sulphosalicylic Acid (SSA) • -Globulin Phenol (Pandy's) • • • 197 • • •~ • Parasitology Stool -Direct Microscopy-saline, Lugol' s Iodine • -Concentration Formol-saline-ether • Urine -Parasites, eggs Wet prep. microscopy • Genito-urinary-Direct Wet prep. microscopy Blood -Direct Thick & Thin films, Romanowsky • -Concentration Centrifugation, microscopy • Skin -Direct Wet prep, microscopy • Bacteriology • -Smears -Gram Stain • -Ziehl Neelson Stain • Serology -Syphilus -Rapid Plasma Reagin Test(RPR) • -HIV -HIV-Chek • Clinical Chemistry -Serum Urea (BUN) -Test Kit(Diacetyl Monoxime) • -Serum & Urine Creatinine -Test Kit (Jaffe Reaction) • -Serum Bilirubin (Total • & Conjugated) -Test Kit (Jendrassik-Grof) • -Serum Total Protein -Test Kit (Biuret Reaction) -Serum Albumin -Test Kit (Bromocresol Green) • -AST (SGOT) -Test Kit (Reitman/Frankel) • -ALT (SGPT) -Test Kit • -Blood Glucose -Test Kit (0-toluidine) • • • • • • • 198 • • • • • • Overhead • Lab Issues In Developing Countries • 1. Sample collection - consent • cultural sensitivity • transport, labelling, storage • equipment

• l • safety j • i • !' 2. Testing - in-country capabilities ! ; .• I • test kits ' • I instrumentation i • !' ! • reagents ! , .• documentation I' • I' !I • quality control I • ! • I Ir • 3. Personnel - training ! ,.• I • outline procedures and tasks I I I • ! I • I 4. Safety - Personal I • I • i Community i • I t ! • [ 199 i • i ,.• I I t • [ • I • • Overhead • Lab Issues - Small Group Role Play & Problem Solving • OBJECTIVE: to set up a lab-based survey for the prevalence of malaria. • • METHOD: role play exercise • • 1. Choose roles to be played among your group • Roles - Primary Investigator • Clinician or country counterpart • Laboratory personnel (1 or 2) • Sample collector • Community representative • Role play observer • 2. Questions to aid discussion: • Where will the study be done and who will it involve? • How will you obtain informed consent? • What clinical investigations are to be done? • How will the samples be collected, labelled and stored? • What additional training and resources will be needed? • What safety precautions will be taken for the subject/community? • What information will be given back to the community at the end of the • study? • • 200 • • • • • ROLE PLAY , .• • Why do it? • Role playing allows everyone to participate in the discussion even if they have little knowledge • or experience of the topic being discussed . • It allows us to artificially recreate a situation in which you might find yourself in the field . • It draws out all issues surrounding a topic. In our example, the community participation and communication problems came out during role play and might have been missed without the role • play. Success of any project depends on the achievement of tasks and on how well you achieve the tasks. Does the community accept your decisions? , .• During analysis of role play, bring up the following points about group process and stress that • health researchers must understand them and integrate them to run a successful project: • • interactions among group members • leadership ,.• • decision-making processes • problem solving • • conflict resolution • (See Chapter 9 by I. Allen Peters on working with groups) • Reference An excellent, inexpensive introduction to working with groups are the pamphlets prepared by • Hedley G. Dimock for the Centre for Human Resource Development. These are four • monographs in the Leadership & Group Development Series: Factors in Working with Groups • How to Observe your Group • How to Analyze & Evaluate your Group • Planning Group Development • They are available from: Office for Educational Practice • University of Guelph, • Guelph, ON NlG 2Wl Handouts - Guidelines for Role Play Observers • Checklist for evaluating your own performance in your group • Design for Action Work Sheet • 201 • • • • CASE STUDY 1: A COMMUNITY-BASED MALARIA PREv ALENCE SURVEY OF DRUG REslsTANI' STRAINS OF Pkzsmodium falciparum IN 1'HAILAND • Introduction • In Thailand, the two species of malaria that occur are Plasmodiwn falciparum and P. malariae. • Because drug-resistant strains of falciparum malaria are increasing and spreading, it is necessary • to determine the occurrence of P. falciparum within the population in order to prescribe effective • drug treatments. • Problem • Financial support has been given to Dr. Arnold Smith of the Liverpool School of Tropical • Medicine from the Malaria Unit, Division of Control of Tropical Diseases, World Health • Organization. The study will be carried out in conjunction with the Malaria Division of the Ministry of Public Health, Bangkok. • Your task is to set-up and conduct a community-based study of malaria infections in the • South West District, a malaria-endemic area of Thailand encompassing fifteen villages, in order to determine the percentage of the population infected with P. falciparum. Lab facilities are • available in the eight-bed District Health Centre. The results will be used by the Ministry of • Public Health to set guidelines for the routine first-line drug treatment for uncomplicated malaria • infections. • How would you go about setting up and conducting this study? What lab-related • problems are you likely to face? How would you solve them? • • Note: Laboratory procedure for the diagnosis of malaria • involves taking blood samples by fingerprick, preparing a thick • film blood microscope slide, staining and microscopic • observation. The slides are graded as positive (asexual malaria parasites seen) or negative (no asexual parasites seen), based on • the inspection of 200 fields of the microscope. The staining can • be done in batches and takes approximately half an hour. • Malaria species determination is done using a microscope slide • prepared with a thin blood film and the same staining procedure. • • • 202 • • • • • • CASE SnIDY 2: COST-SAVING THROUGH MICROSCOPY BASED VERSUS PREsuMPTlvE ,.• DIAGNOSIS OF MALARIA IN ADULT OUTPATIENTS IN MALAWI • Introduction A diagnosis of malaria in most outpatient clinics in malaria-endemic areas is made on the basis • of the presenting signs and symptoms, without microscopic confirmation of parasitaemia, and • treatment is based on this 'presumptive diagnosis'. Because the clinical presentation of malaria • (fever, headache and/or myalgia) is similar to that of other common illnesses, this presumptive diagnosis often results in unnecessary administration of antimalarial drugs. The extent of ;.• antimalarial drug overuse due to a wrong presumptive diagnosis will influence drug budgets, and could increase the development of drug resistance in the parasite population and the incidence • of adverse reactions to antimalarial drugs, as well as delay the determination of correct ,. diagnoses . • The alternate approach is for each malaria treatment to be based on a microscopically • confirmed diagnosis. The cost implications of such a policy have been discussed, but not • measured directly in an adult outpatient setting in a malaria-endemic area. • Problem • Funding has been obtained from the British Overseas Development Assistance Program and the principle researcher will carry out the study in the adult outpatients department of the 800-bed • Queen Elizabeth Central Hospital in Blantyre, Malawi during the main malaria transmission • season (January to April) . • Your task is to set up a research study that will determine the costs of malaria treatment in the adult outpatients department of a government hospital in Malawi, and examine the effect • on drug expenditure of changing from a policy of presumptive diagnosis to microscopy-based • diagnosis . How would you go about setting up and conducting such a study? What lab-related ,.• problems are you likely to face? How would you solve them? Note: Laboratory procedure for the diagnosis of malaria involves taking blood samples by • fingerprick, preparing a thick film blood microscope slide, staining and microscopic observation . The slides are graded as positive (asexual malaria parasites seen) or negative (no asexual • parasites seen), based on the inspection of 200 fields of the microscope. The staining can be • done in batches and takes approximately half an hour . • Taken from a paper by A. Jonkman, R.A. Chibwe, C.O. Khoromana, U.L. Liabunya, M.E. Cbaponda, G.E. Kandiero, • M.E. Molyneux, & T .E. Taylor in Bulletin of the World Health Organization, 1995, 73 (2): 223-227 . • • 203 • • • • GUIDEI.JNES FOR ROLE PLAY OBSERVERS Handout • Your job is to observe group interactions and to report at the end of the exercise on any • strengths or problems of group process which you observe. The following questions may help • in your analysis: • 1. Did people appear to feel comfortable in the group? • • 2. Did any participant(s) tend to dominate the discussion? • 3. Did all participants contribute to the discussion? • 4. Did any one person emerge as leader of the group? • • 5. Did you observe conflict in the group? If yes, how was it resolved? • 6. How were decisions made in the group? • Check any special behaviours you observed in the group: • daydreaming interpreting ideas • to group • encouraging others criticizing others' • to speak ideas • giving information making suggestions • interrupting others encouraging agreement • seeking information harmonizing conflict • agreeing with • others' ideas other (please identify) • making suggestions • Any other observations? • 204 • • • • ••• • Handout • CHECKLIST FOR EVALUATING YOUR OWN PERFORMANCE IN YOUR GROUP • Put a check ( ) to indicate how frequently you believe you perform each function when you are • in your group . • NEVER• • IFUNCTIONS I ALWAYS I OFTEN I SELDOM I I • Ask for information • Give information • Ask for opinions • Give opinions • Encourage others to speak • Support others' points of view • Summarize discussion • Make suggestions • Harmonize conflicts • Follow suggestions • Encourage agreement • Seek recognition •• Speak clearly and to the point • Date: _____ • • Definitions: Always: every time • Often: more than half the time • Seldom: less than half the time • Never: never! • • • 205 • • • • DESIGN FOR ACilON WORKSHEET Handout • Think about one thing you can do to improve intergroup relations. Then complete the following • action plan for implementing your strategy within the next thirty days. • • The problem or issue: • • My strategy: • • • Specific tasks involved. Who will do them and by when? • • • What are some of the obstacles that may get in the way and how may I overcome or reduce the • impact of each obstacle? • • What resources do I need? • • How will I demonstrate I have made progress on this issue? • • Source: Joe Giordano, lecture presented at Training of Trainers: Developing Cultural Diversity Programs for the Workforce, 1993. Adapted by permission of Joe Giordano, Ethnicity and Mental Health Associates, 125 Parkway Rd., • Suite 12, Bronxville, NY. • • 206 • • • • • • RECOMMENDED READING • Books & Manuals • 1. Brabec, J. and Barss, P. (1986) Laboratory Guide for Rural Health Centers in Papua ;.• New Guinea. Summer Institute of Linguistics Press, Ukarumpa PNG. 2. Cheesbrough, Monica (1987) Medical Laboratory Manual for Tropical Countries . • Vol. 1, 2nd Ed. Tropical Health Technologies & Butterworth, England . • 3. Cheesbrough, Monica (1984) Medical Laboratory Manual for Tropical Countries. • Vol. 2: Microbiology. Tropical Health Technology, England . • 4. International Dispensary Association (1990) Medical laboratories: Methods and • materials. IDA, Amsterdam . • 5. Smith, P.G. and Morrow, R.H. Eds. (1991) Methods for Field Trials of Interventions • against Tropical Diseases: a 'Toolbox'. Oxford University Press. Oxford . • 6. World Health Organization (1980) Manual of basic techniques for a health laboratory. • WHO, Geneva . 7. World Health Organization (1986) Methods recommended for essential clinical • chemical and haematological tests for intermediate hospital laboratories. Working • group on assessment of clinical technologies. Lab 86.3, WHO, Geneva . • 8. World Health Organization (1987) Laboratory Services at the primary health care • level. WHO/Lab/87.2, Geneva . • Papers • Armropriate Technology • 1. Barss, P., Brabec, J. and Gregotski, G. (1985) Training appropriate laboratory • workers for rural health centres in Papua New Guinea. Papua New Guinea Medical ,.• Journal 28 (4): 251-256. 2. Bloom, Gerald (1989) The right equipment ... in working order. World Health • Forum 10, pp 3-10 plus other letters & discussion on topic . ,.• 3. Cheng, Michael (1994) Priority for maintaining essential medical equipment. World • Health Forum 15, pp 196-199 . • 207 • • • • 4. Christian Medical Commission (1994) Guidelines on Equipment Donations. Contact, Oct. 1994, CMC - Churches' Action for Health, World Council of Churches, • Geneva. • 5. Jumah, K.B. in WHO Readers' Forum (1993) Basic Requirements for hospital • equipment. World Health Forum 14, pp 61-62. • 6. Kaminsky, R.G. (1983) Development of a rural laboratory in Tunisia. World Health • Forum 4, pp 162-165. • 7. Narayan, Ravi (1990) Technology and Health Care: Issues and Perspectives. Link Newsletter 9, No. 1, pp 8-12. • 8. Roelants, Gilbert (1995) Which health journals are most useful? World Health Forum • 16, pp 69-73. • 9. Sekiguchi, J. and Callens, S.R. (1995) Radiological services in rural mission hospitals • in Ghana. Bulletin of the World Health Organiz.ation 73, pp 65-69. • 10. Weitzel, Rolf (1992) Essential books for health workers in the Third World. World • Health Forum 13, 240-243. • 11. WHO Notes and News (1991) Health Care Equipment. World Health Forum 12, pp • 497-498. • Lab Techniques • 12. Arimitsu et. al. (1994) Evaluation of the one-point microcapsule agglutination test for •~ diagnosis of leptospirosis. Bulletin of the World Health Organi:zation 72(3), pp 395- 399. • 13. Babu,, P.G. et. al. (1993) Reduction of the cost of testing for antibody to human immunodeficiency virus, without losing sensitivity, by pooling sera. Indian Journal of • Medical Research 97, pp 1-3. • 14. Brogdon, W.G. (1987) Laboratory and field microscopy of cholinesterases in whole blood, plasma, and erythrocytes. Bulletin of World Health Organiz.ation 65(2), pp • 227-232. • 15. Gajanana et. al. (1995) A community-based study of subclinical flavivirus infections • in children in an area of Tamil Nadu, India, where Japanese encephalitis is endemic. Bulletin of the World Health Organiz.ation 73(2), pp 237-244. • • 208 • • • ,,4 . • • 16. Guimaraes et. al. (1985) Long-term storage of IgG and IgM on filter paper for use in • parasitic disease seroepidemiology surveys. PAHO Bulletin 19(1), pp 16-28 . • 17. Hubbard, J.D. and Calvo, C.L. (1994) The status of clinical laboratories in Kenyan • Mission Hospitals. Laboratory Medicine 25(3), pp 156-161. • 18. Hudson-Thomas, M., Bingham, K.C. and Simmons, W.K. (1994) An evaluation of • the HemoCue for measuring haemoglobin in field studies in Jamaica. Bulletin of the • World Health Organization 72(3), pp 423-426 . • 19. Kenyon, T.A., Kenyon, A.S. and Sibiya, T. (1994) Drug quality screening in • developing countries: establishment of an appropriate laboratory in Swaziland. • Bulletin of the World Health Organization 72(4), pp 615-620 . 20. Ko, Y.C. et. al. (1992) Successful use of pooled sera to estimate HIV antibody • seroprevalence and eliminate all positive cases. Asia-Pacific Journal of Public Health • 6(3), pp 146-149 . • 21. Lillo, F. et. al. (1992) Detection of HIV-1 antibodies in blood specimens spotted on • filter paper. Bulletin of the World Health Organization 70(3), pp 323-326 . • 22. Renganathan et. al. (1995) Evolution of operational research studies and development • of a national control strategy against intestinal helminths in Pemba Island 1988-92 . • Bulletin of the World Health Organization 73(2), pp 183-190 . • 23. Van Dyck, E., Bogaerts, J. and Piot, P. (1994) Rapid plasma reagin card test: • evaluation of a hand rotation procedure and stability of the RPR antigen. Bulletin of • the World Health Organization 72(5), pp 741-743 . • 24. World Health Organization News (1992) Testing for lllV antibody in serum pools . • Bulletin of the World Health Organization 70(2), pp 277-280 . ,.• 25. World Health Organization (1994) Laboratory diagnosis of measles infection and monitoring of measles immunization: Memorandum from a WHO meeting. Bulletin of • the World Health Organization 72(2) pp 207-211 . 26. Zwicker et. al. (1990) Mass screening for Trypanosoma crozi infections using the • immunofluorescence, ELISA and haemagglutination tests on serum samples and on , .• blood eluates from filter-paper. Bulletin of the World Health 68(4), pp 465-471. • • • 209 • • • • Quality Control • 27. Braun-Munzinger, R.A. (1989) Quality control in schistosomiasis programmes. Trop. • Med. Parasit. 40, pp 214-219. • 28. Goddard, M.J. (1980) A statistical procedure for quality control in diagnostic laboratories. Bulletin of the World Health Organization 58(2), pp 313-320. • 29. Lewis, S.M. (1988) Getting the right answers from blood. World Health Forum 9, pp • 575-581. • • 30. World Health Organization (1989) Guidelines on sterilization and disinfection methods • effective against human immunodeficiency virus (HIV). WHO AIDS Series No. 2, WHO, Geneva. • 31. World Health Organization (1991) Biosafety Guidelines for diagnostic and research • laboratories working with HN. WHO AIDS Series No. 9, WHO, Geneva. • • • • • • • • • • • 210 • • • ••••••••••••••••••••••••••••••••••••••••••• 12 Field laboratory methods

I I ntroduclion 2 Sample collection 2.1 Types of specimen 2.2 Handling specimens 2.3 Blood collection 2.4 Stool and urine collection 2.5 Sputum collection

From Smith, P.O. and Morrow, R.H. Eds. (1991) 3 Labelling and storage Methods for Field Trials of Interventions against Tropical Diseases: a 'Toolbox'. Oxford University Press, 3.1 Labelling 3.2 Storage Oxford. 3.3 Aliquoting 3.4 Storage system

Used with permission. 4 Documentation of laboratory procedures N - 4.1 Supplies - 4.2 Equipment maintenance 4.3 Procedures and staff duties 4.4 Unusual events 5 Quality control 5.1 Reproducibility of test results S.2 Internal quality.control · 5.3 External quality control 6 Links with other laboratories 7 Coding and linkage of results 8 Laboratory safety 232 Chapter 12 Field laboratory met/rods Sectio11 2 Sample collection 233 3. Food, water and environmental samples. 1 INTRODUCTION This chapter will discuss specimens collected from humans. Laboratory tests may provide the definitive basis for the measure­ ment of outcome variables in field trials, either directly by demon­ 2.2 Handling specimens stration of the presence of the pathogenic agent, or indirectly by demonstration of a host reaction to the pathogen or of biochemical The collection of samples for laboratory studies will usually in­ changes due to.the pathogen. They may also provide ev.idence of volve the following steps: the action of t.he intervention, either directly by measuring the drug or metabqlic by-products, or indirectly by measuring an im­ (1) collection of specimens from the study participants; mune response to a vaccine. In addition, they may be used to (2) placement in a suitable container; detect, or confirm the presence of adverse reactions. (3) labelling of the container; The organization and operation of a field laboratory for the (4) temporary store at appropriate temperature; support of a field trial are different from those of a routine medical laboratory. In fietd trials the emphasis is often on the collection (5) initial processing (for example, serum separation from whole and processing of large numbers of samples on which only a few blood), with appropriate relabelling; specific tests will be performed. Because of the repetitive nature of (6) transport to intermediate or final destination for further the collection and processing of the samples, it is essential to processing, testing, and storage . . include appropriate quality control procedures at all stages. N General aspects of the setting up and running of a field labora­ The procedures for collecting and. processing samples must be -N tory are discussed in this chapter. For information on specific unambiguously specified,_ including to where they are to be trans­ laboratory tests and specific laboratory methods other literature ported and how they will be labelled. All aspects of the collection, should be consulted. Useful general texts containing relevant in­ transport, and processing of samples must be piloted. Often much formation for the operation of a field laboratory and for collecting attention is paid to the proper design and testing of questionnaires, specimens include WHO (1980), Cheesbrough (1981) and Brouard but much less care is taken to find the most appropriate and et al. (1989). culturally acceptable methods for the collection of blood, stool, urine, or tissue samples. This may be crucial to ensure sustained community involvement and participation. 2 SAMPLE COLLECTION

2.1 Types of specimen 2.3 Blood collection The kinds of specimen which are commonly collected in field trials The usual methods by which blood is collected in field surveys are on which laboratory tests may be performed include: by venepuncture or by finger or heel pricks. In most, but not all, communities it is easier to persuade individuals to donate a finger­ I. Human specimens, including blood, stool, urine, sputum, skin prick sample, and this will be the principal means of blood collec­ snips and other tissue biopsies, and swabs or smears collected tion in many trials. This method of collection provides an adequate from skin or mucosa! surfaces. volume of blood for many laboratory tests. 'Micro-techniques' 2. Entomological specimens for studies of vectors, and animal or have been, or are being, developed for many assays, and investiga­ malacological specimens for studies of intermediate hosts, re­ tions should be conducted before a study starts to find out the lated to effects on transmission. latest availability of such techniques (for example, by contact or •••••••••••••••••••••••••••••••••••••••••••• ••••••••••••••••••••••••••••••••••••••••••• 234 Chapter 12 Field laboratory methods Section 2 Sample co//ectio11 235 correspondence with those in a central or reference laboratory). It If repeated blood sampling is to be undertaken during the is important to verify that the methods have been adequately course of a study it is likely to be easier to maintain the co­ validated. Some tests require larger quantities of blood, however, operation of most study populations if finger-prick, rather than and it will often be desirable to collect blood by vcncpuncturc venous blood, sampling is used. For many purposes, filter paper from at least a sample of the population. samples arc just as satisfactory as blood samples collected in capil­ After collection, blood can be separated into several components lary tubes, and the former are easier to handle, label, and store including scrum, plasma, red cells, and white cells. The separation (Brouard et al. 1989). For some assays, however, larger samples must be done shortly after the blood has been collcclt?d and it is are currently needed (for example, tests for cell-mediated immun­ common foe this procedure to be carried out close to where the ity or l-ILA typing) and venous blood sampling will be required. samples havl! been collected or in a nearby field laboratory. All blood samples should be considered to be potentially infec­ 1 A samplc of blood taken from a finger-prick may be collected in tious and appropriate handling procedures must be employed to one of several ways, including: safeguard all those who will come into contact with the specimens during their collection, processing, analysis, or storage (Brouard et (1) collection into narrow glass tubes, hy capillary action; al. 1989). Special precautions should be taken when collecting (2) dropping onto a glass slide for direct examination of the blood blood by linger-prick. If possible disposable gloves should be worn smear; or and other safety measures should be employed. At the very least, (3) dropping onto strips or disks of absorbent paper (filter paper). a bucket containing water and detergent should be available for use by those taking blood. A detailed discussion is given in Brouard et al. (1989) of the method to use for puncturing the skin and of the relative advanta­ 2.4 Stool and urine collection ges and disadvantages of collection by capillary tube as compared to absorbent paper. A summary of different methods that may be used for collecting A specimen of blood obtained from a finger-prick may be suf­ urine and stool samples, with details of different container types, is ficient in volume to use for several purposes. For example, when a given in WHO (1980). The methods considered for use in a parti­ very small amount of Hemolubef) (silicon grease) is applied to cular survey should be discussed with those knowledgeable of local the finger tip before pricking, sufficient blood can be obtained to customs and taboos. In some cultures sensitivity regarding the allow the first drop to be wiped off, to obtain drops for 2 thick and collection, or public display, of stool specimens may be greater ithin malaria smears, to do one or two hacmatocrits (50J11), and than that for blood. A container which is technically appropriate to collect 50-100 µI blood in a microtubc or Microtainer® for may not be acceptable in a particular study community (for exam­ scrum and to place a drop on filter paper (WHO 1980). After the ple, due to colour, transparency, resemblance to cultural motif). haematocrits have been read, the tubes can be broken at the In advance of a survey the proposed stool and urine containers interface to the packed cells and both ends sealed with plasticine should be shown to the village leaders and the proposed methods (plastic modelling clay). (It should be noted, however, that there of sample collection discussed. As with all field procedures, it is are safety concerns regarding such procedures ...: see Section 8.) important to undertake pilot testing to ensure that the procedures The tube with the plasma can be stored in a freezer or in liquid planned will be acceptable (both to the investigator and to the nitrogen. The amounts of plasma and serum recovered from study population). ~ finger-prick samples will be sufficient to perform various ELISA­ As stool samples can rarely be collected 'on the spot' it is usually bascd serological tests and is sufficient for the determination of necessary to leave the container with an individual overnight and some micronutrients such as vitamin A or zinc (minimum scrum to arrange to pick up the specimen on the following day. A poten­ requirements 25-40 µI). tial hazard in doing this is that containers may be exchanged 236 Chapter 12 Field laboratory methods Section 3 Labelling and storage 237 between individuals or, for example, one person may provide a collection, the type of specimen, if not evident, and possibly the sample for the whole family. It is difficult to rule out this possibility name of the village. Individual names may also be recorded but completely, but it is important for fieldworkers to stress .the im­ this can create problems with blinding and confidentiality, and portance of participants ndhcring to the correct procedures and to oflcn names arc not a unique idcnlificr as several individuals may be on the look out for possible cheating. have the same name. Containers should usually be labelled using water-proof marker 2.5 Sputum collection pens (but see note 1, below), writing directly onto the container The WHO ( f980) manual gives a concise description .of recom­ labelling area or onto adhesive labels attached to the container. If mended methods of collecting sputum samples using different the container has a cap, the marking should be on the body of the kinds of jarsJ boxes, and containers including transport media. containcr(and possibly on the cap as well, but neveron the cap only). Two points merit special attention: A 'Hag' can be attached to capillary tubes, made from an adhesive strip, for the identification information, written with a ( 1) all sputum samples should be considered potentially infectious; waterproof markci: pen, or tubes may be stored in labelled (2) careful attention should be given to the cold-chain require- envelopes as they are collected in the field. Alternatively, capillary ments if sputµm samples have to be sent for culture to another tubes which are 75 mm long will fit into an ordinary microscope laboratory. slide box. The tubes can be put into a numbered sequence according to subject, and multiple tubes may be taken from a subject and stored together. The box facilitates transport and 3 LABELLING AND STORAGE obviates the need for ftags, which are awkward and may have to be 3.1 Labelling removed before spinning. Staff need to be careful, however, to record and maintain the correct numbering and not invert or tip Proper labelling of samples is essential. The labelling scheme the box. Packing with cotton wool will keep the tubes in place. should be made as simple as possible, consistent with the study Filter paper can be written on directly. objectives, and must take due account of the size of containers and It is not possible to recommend a single standardized form of how the specimens will be handled, transferred, and stored. Sclf­ labelling for different sample containers that will be appropriate in adhesive, preprinted labels (possibly prepared using a microcom­ all circumstances. It will be necessary in a particular study to puter), with the individual identification numbers duplicated on establish a method that guarantees the reliability of the labelling data sheets, can speed processing. Also, lab<:ls in a variety of from the time the sample is first collected, through transportation, materials suitable for differing storage conditions, and with each processing, analysis, and storage. Useful advice on these aspects is number duplicated several times, are available commercially (for given in Brouard et al. (1989). Using sets of labels with series of example, from W. H. Brady & Co Ltd., Daventry Road Industrial identical numbers on them, for coding samples and associated Estate, Banbury, Oxfordshire OX16 7JU, UK, or Shamrock record forms, reduces the chances of labelling errors. · Scientific, 34 Davis Drive, Bellwood, Illinois 60104, USA, or A few warnings are appropriate. Thomas Scientific, 99 High Hill Road, Swedesboro, New Jersey 08085, USA). 1. If the transport cold chain include~ a stage where samples are The information recorded on a label will vary according to frozen in salt-alcohol mixtures, never use felt pens (even water­ particular requirements. It may include just a single number or proof ones). Always use plain pencils. code which is utilized during laboratory processing and which may 2. Numbers and letters must be written in a clear and standard­ be linked back to an individual by reference to records kept at the ized form (for example, 191 looks the same as 161 upside-down!). time the sample was taken. In some circumstances it will be Advice on some of these aspects is given in Chapter 13, Section appropriate to include on the label a record of the date of 4.2 .

••••••••••••••••••••••••••••••••••••••••••• _ ··········································-······- .. 238 Cliapter 12 Field laboratory metlwd.f . Section 4 Documentation of laboratory procedures 239 3. The methods to be used for collection, storage, and transport thawing. This can be avoided if samples are divided into small of specimens should be pilot tested. This testing should include portions before freezing. Ideally, the size of aliquots should be evaluation of how labels stick under the varying conditions (for chosen so that there is just sufficient material in each aliquot to example, freezing), how pen and pencil writing is conserved from perform the tests required at a particular time. This is not always the field collection to the final place of analysis and storage, and possible and, in practice, compromise procedures may have to be where there is the possibility of transcription errors. Labelling of adopted (for example, on grounds of coi;t). It is important that the specimens with commercially available self-adhesive labels should laboratory recording procedures arc such that the histories of each be undertaken with care. They often curl off plastic tubes and aliquot arc adequately documented (especially recording how sometimes inscriptions on them rub off with frequent himdling or many times each one has been thawed and re-frozen) so that any wetness, even if waterproof markers have been used. Some recipient of the samples can be given detailed information about workers have found that fixing transparent adhesive tape around their preparation (for example, whether volumes are precisely the labels helps to overcome these problems, in the absence of measured or are approximate): better labels. Special containers and labels arc required if samples arc to be stored in liquid nitrogen. 3.4 Storage system Quality control ~ocedures in the laboratory arc likely to require When large numbers of samples are collected and stored, a storage that some samples are stored and coded in replicate, both for check­ and record system must be devised that allows the rapid retrieval ing analysis and handling procedures and for establishing the of particular samples. If this is not done, sorting through large variability of tests (see Section 6). Procedures for dividing and numbers of samples can be a very time-consuming activity. The labelling such samples should be devised and laid down in the field particular storage system used should be tailored to the design of procedures. the specific study. In general, it is appropriate to store samples in hutches according lo the date they were collected or frozen, with u J.2 Storage careful record being kept of the contents of each batch. This The storage area of a field laboratory should be designed to be information can be recorded on a microcomputer so that, if adequate for a particular study (or studies). This will require necessary, a computer search can be made to find the location estimation of the rate at which samples will be collected and (batch) of a particular specimen. processed and for how long they must be stored before being 1 transported on to another location (for example, for processing or storage in the base laboratory). Serum and plasma samples should 4 DOCUMENTATION OF LABORATORY PROCEDURES be frozen as soon as possible after separation, and storage in a field laboratory at -20°C is adequate for most purposes, at least for There should be clear and explicit documentation of all laboratory several weeks. For only a few tests is immediate storage at less procedures to ensure reproducibility and to enable appropriate than -70°C required. comparisons to be made with results from other laboratories. Stool, urine, and tissue samples may be stored under various Records should be made of equipment maintenance, the batches conditions using appropriate fixatives and stabilizers; different of supplies and reagents used al different times, the detailed test possibilities are summarized in WHO (1980). procedures and the duties and responsibilities of staff members. Records should also be made of unusual events that may affect the results of a test (for example, power failures-though in some J.J Allquotlng circumstances these may not be 'unusual'). . Biological samples arc easily damaged by repeated freezing and Depending on the size of the laboratory and the variety of tests

--·--·-----~--~----.<--··-····------~------·------·------·------·------~------240 Clrapter 12 Field laboratory met/rods Section 5 Quality control 241 and procedures undertaken, the documentation should be Maintenance procedures are usually described in the instruction arranged in a single, or several, log-books that are arranged booklets for the relevant equipment, but these are often inade­ chronologically. quate. Complete maintenance instructions for each piece of equip­ ment should be incorporated into a manual and a log-book with 4.1 Supplies check-lists kept for each piece of equipment. The supervisor should review these log-books regularly with the appropriate The laboratory log-book should provide information on: the members of the laboratory team. reagents, test kits, laboratory equipment (including b~and names), the shelf-life of reagents, storage conditions, batch or lot numbers, and the rel~vant re-ordering arrangements (for example, when, 4.3 Procedures and staff duties how much, 1by and through whom). Every laboratory procedure should have detailed, step-by-step, instructions described in a manual (i.e. what to do with specimens 4.2 Equipment maintenance from their receipt in the laboratory to the storage of specimens and the communication of the results). Each step should be clearly Regular checks.. should be made on each piece of equipment to detailed and the responsibility of each staff member indicated. In ensure that it is in good working order. Such checks should be addition, staff members need to know to whom they should report recorded and, for some items, publicly displayed. Among items and how they should record results, additional observations, that should be checked regularly are: · mistakes, and other unusual events. These include any change of: (1) daily (morning and evening) recording of temperatures of kit or batch number of sera, media, preservatives, and so on; N refrigerators, freezers, and cool-rooms, using maxima and working arrangements, such as a new brand of glassware, changes °'- minima thermometers; in incubation time or temperature; and coding procedures for samples and results. Staff members involved in distinct sequences (2) checking on the position of the cap and the level of nitrogen in of the procedures should be indicated on relevant flow charts and ·liquid nitrogen containers; these should be written into the log-book(s). (3) regular and systematic inspection of all items of equipment which require clean lenses (for example, microscopes) and checks on focus and adjustment of light sources; 4.4 Unusual events I (4) periodic checks on the position of centrifuge rotors (tight The log-book(s) should be used to keep a record of errors in test centre bolts) and regular cleaning; procedures, and in the preparation of reagents, power failures, (5) regular zero-line/level calibration of balances and cleaning of temperature and humidity changes that might influence the results balance pans; of the tests or the quality of stored samples. (6) checking the filters and calibration of spectrometers and varipipettes. 5 QUALITY CONTROL In moist areas of the tropics fungus may grow on glass lenses and damage microscopes and similar equipment. In these circum­ Quality control is an inherent component of a good study. A stances it may be worth storing microscopes, cameras, and similar comprehensive work plan will facilitate the systematic implemen­ equipment in a 'light-box' (a box containing, and heated by, a low­ tation of quality control mechanisms. Appropriate references on wattuge electric light bulb) to prevent condensation, or with silica­ c1uality control include Braun-Munzinger ( 1989), Goddard ( 1980), gel (as a dessicant) in air-tight boxes for lenses . Russell (1974), and Tonks (1972). ••••••••••••••••••••••••••••••••••••••••••• ...... ,...... •...... 242 Chapter I 2 Field laborc1tory metlwcls Section 6 Links with other laboratories 243 involves extcrnul monitoring, such as the duplicate testing of 5.1 Reproducibility or test results samples in another laboratory to serve as a 'gold standard'. The reliability of data should be tested by regular checks on its The essence of internal quality lies in a tight circle of checks, reproducibility. There is no general rule for an acceptable level. reporting, evaluation, and action. It is essential to have detailed This will depend on the test involved and on the test conditions, manuals of every procedure, with a check-list to be consulted each and information on these aspects should be sought in manuals or time the procedure is run. Well-kept records with regular review papers in which test procedures arc described. by the supervisors arc key clements in quality control. Quality In addition to ensuring that all procedures arc carried out in control procedures must be an integral part of the work-plan for accordance .with the instructions as detailed in the manuals, that the study. all cquipme11t is in good working order, and that materials arc properly and kept up-to-date, the basis of the approach to sh~red 5.3 External quality control internal quality control is through checking on the reproducibility of test results. Many test systems have inbuilt controls using A major reason for having specimens sent to external laboratories · standardized reagents for this purpose of known concentration is to check the accuracy of test results. Reproducibility can be or quantity. 'J.:~e use of such standard controls is necessary but assessed adequately by internal quality control procedures, but not sufficient to monitor the quality of test procedures. Whenever checks on accuracy arc best done, for many tests, in collaboration possible, coded duplicate samples should be tested. The frequency with other laboratories. The results from a laboratory may be with which such duplicates are included depends upon how highly reproducible within that laboratory, but might be consis­ smoothly the laboratory is running and how long it has been tently incorrect. Arrangements with other laboratories should be doing the test. Typically, when a test is first introduced, a high organized well in advance and should be included in the work frequency of such checking may be appropriate, with a decreased plan. The actual checking procedures involved are similar to those frequency as the procedures become more familiar. In many cir­ for internal reproducibility checks as outlined above. cumstances it will be appropriate to ensure that duplicate analyses Samples that are sent should be accompanied by an investiga­ are done on between 5 and 10 per cent of sample& on a routine tion request form and every effort should be made to ensure that basis. transport conditions are appropriate and the same for all samples Reproducibility should be checked within batches, between (for example, route, packing conditions, type of container). The batches and from day-to-day or week-to-week by the use of field laboratory may make arrangements with the main laboratory appropriate controls. Intra-observer variatio11 can be determined to have samples sent to an external laboratory, rather than sending by having duplicate samples processed by the same observer at them direct, but these samples should have been tested by the field different times, and inter-observer variations measured by having laboratory within its routine programme. At the beginning of a the same samples processed independently by two different staff field study and during training phases this part of quality control members. Inter-product variation is tested by comparing new activities should be strengthened. batches of staining solutions, media, reagents, and so on with the old on a group of the same samples. 6 LINKS WITH OTHER LABORATORIES 5.2 Internal quality control A field laboratory may be set up specifically ~or the conduct of a Two types of quality control can be distinguished, 'internal' and particular study and may have no regular links with other labor­ 'external'. Internal quality control comprises procedures that arc atories. Often, however, there will be a link to another laboratory introduced within the field laboratory. External quality control in the home-based institution at which some of the laboratory tests 244 Chapter 12 Fie/cl laboratory methocl.~ Section 8 Laboratory safety 245 may be checked or performed. There should be a clear specifica­ know which intervention group any sample is from, and it should tion in the study protocol of which procedures and checks will be not be possible for this to be deduced from the labelling system performed at which laboratory, of how arrangements will be made employed. Specimens must be labelled in such a way, however, for transport of specimens and supplies between them, and how that each is identified uniquely and any test results can be linked and which records will be interchanged. Links with an external back later to records of the individual from whom the specimen reference laboratory may be desirable for independent checks and was taken. While this seems to be stating the obvious, the quality control procedures or for the conduct of new or specialized problems that arise with these aspects in large studies are often tests. . substantial. Special care is necessary in longitudinal studies where If samples arc to be sent to other laboratories for further storage, individuals may he followed for many years, in studies involving processing, or 4nalysis (for example, blood, scrn, slides), it may he many different research groups or laboratories, and in studies important to gi'vc attention to the suggestions listed below: where results need to be linked with census information that may I. It is unwise to send all samples to another laboratory at the be updated over time (for example, individuals may move house same time. Duplicates should be kept, even when storage facilities and this may cause problems if the coding system for individuals is arc limited, to guard against loss during shipment. too closely linked to a house code). Laboratory results will usually be recorded in laboratory books 2. Samples should not be sent to another laboratory without a or on specially prepared forms to facilitate computer entry. For clear agreement as to what analyses will be done and how these such recording it will be important to develop result codes that will be reported back. It is essential to know wlro does what, witlr idcntify particular problems or features. For example, codes might wlrat, and wlren ('the rule of the 5W's'). This point is very be developed to indicate lost and broken samples, technical important. A laboratory may be tempted to do more tests on N problems with batches of samples (for example, staining, storage, ..... samples than are requested (for example, to try out a new test on a 00 transport), and the identification of the technicians involved with variety of samples), but it must be made clear that this should only each test (to check variations between observers). be done if there is prior agreement for this. No tests should be If the study uses laboratory numbers in addition to individual performed other than those previously agreed. It is always a very identification numbers (as is often the case), both numbers should sound policy to send samples to another laboratory in such a way be entered on a computer form for data entry so that cross-checks that they arc analysed 'blind' (for example, no details arc sent of can be made in the computer. which intervention group the samples are from or of the age and If multiple laboratory tests are being performed on samples sex of the individual subjects). Agreement with respect to publica­ from the trial population, it may be best to wait until all the results tion procedures should also be made before specimens are sent. have been assembled and collated before entering them into the The arrangements agreed between the field and other laborator­ computer so that the checking and linkage back to other data on ies should be part of the study protocol, in which the division of each individual can be done in relatively few steps. This will responsibilities should be specified. All parties must also adhere to depend on how the computing system is organized, but repeated the provisions of the agreement to undertake the study (for processing of many small sets of data is liable to lead to confusion example, local research clearance, ethical clearance). and may be unnecessarily time-consuming.

7 CODING AND LINKAGE OF RESULTS 8 LAHORATORYSAFETY

The coding of specimens is discussed in Chapter 7. In order to Detailed attention to safety aspects is as important in a field remove the possibility of bias, those in the laboratory should not laboratory as in any kind of laboratory. Indeed, this is of special ••••••••••••••••••••• ••••••••••••••••••••• ••••••••••••••••••••• , ••••••••••••••••••••• 246 Clrapter 12 Field luboratory methods I References ~-., importance in field laboratories as there may be much greater World Health Organization (1988). Guidelines on sterilization and lligh­ accessibility by the public than is usual in a laboratory. It is level disinfection methods effective against human immunodeficiency important to ensure that proper procedures are documented and vim.v (HIV). WI 10 AIDS Series 2, Geneva. implemented for activities such as the disposal of needles, blood, stool, urine, and sputum samples, and of used reagents, chemicals, and detergents. Special attention should be paid to precautions concerning the transmission of blood-borne infections such as hepatitis B and HIV (WHO 1988). Shattering 9f packed cell volume (PCV) tubes in a microhacma­ tocrit ccntrifu~e is likely to carry special hazards and, if possible, the centrifuge should be put in a hood. Also, cutting PCV tubes with a diamond to obtain scrum is very hazardous and should be discouraged. Laboratory safety guidelines are given in WHO ( 1980). Safety procedures should be reviewed regularly by laboratory supervisors and all staff conce'rncd.

REFERENCES

N :O Braun-Munzingcr, R. A. (1989). Quality control in schistosomiasis pro- grammes. Tropical Medicine and Parasitology, 40, 214-9. Brouard, Y. J., Blackwell, J.M., and Fine, P. E. M. (1989). Blood collection, fractionation and storage methods: a practical manual for immuno-epidemiological studies with emphasis on work in developing COl4ntries. Ross Institute Bulletin No. 15, London School or Hygiene and Tropical Medicine, London. Cheesbrough, M. ( 1981). Medical laboratory manual for tropical .cow1- tries. Austin & Sons Ltd, Hertford, England. · Goddard, M. J. (1980). A statistical procedure for quality control in diagnostic laboratories. Bulletin of the World Healtlr Organization, 58, 313-20. Russell, R. L. (1974). Quality control in the microbiological laboratory. In Manual of clinical microbiology (2nd ed), (ed. Lenette, E. H., Spauling, E. H., and Truant, J.P.) pp. 862-70. American Society for Microbiology, Washington. Tonks, D. B. (1972). Quality control in clinical laboratories. Warner­ Chilcott, Scarborough, Ontario. World Health Organization ( 1980). Manual of basic techniql4es for a health laboratory. WHO, Geneva.

---·------·------·------~·-·------··-·--·-·------·· ... _..... --. _ ___,_,, .. ~--- • • • CHAPTER 8 • TRANSLATING RESEARCH INTO ACTION • C. Larson • OVERVIEW An important premise to these introductory remarks is the belief that health research conducted • in the least empowered, poorest populations of our world can result in improved health and • quality of life. Furthermore, it is not only the quality of research methods or conduct, but, as • well, the relationships established among researchers, communities, and those involved in solving priority health problems that will determine the success of a research project and the • likelihood of translating the research into action . • When we participate in health research, especially in less-developed countries, it is often implied that our research will help solve important health problems and lead to improved health • and quality of life. Unfortunately, research findings are often not translated into action and have • little or no impact on health. The translation of research findings into action is the last, but most • important, step in the research process. We will examine strategies to improve the link between • research and action. The objectives of this session are to sensitize trainees to barriers in translating research into action and to explore mechanisms to improve the likelihood that • research findings will result in action . , .• • SUGGESTED SCHEDULE • 20-30 min Give an overview of health research for development and • discuss limitations in how successful researchers are at translating research into action. Stress that it is important that • we learn more and establish the gaps that exist in our • knowledge about this topic. You could modify this section to • adapt to local realities. It might be helpful to start with a • personal experience so that the trainees will feel less inhibited • to discuss their own past successes and mistakes . • • • 220 • • • • NOTABLE QUOTE • "Medical care or health services research, in contrast to biomedical research, has • had little discernable impact on public health policy, the behavior of health • professionals, or the organization of health facilities. . . . There is little evidence that the initiation of policy flows directly from research findings . . . "1 • • • It is evident to many researchers that their work has not had the impact they had hoped it would. • Why? Let us look at some important issues that arise in community-based health research in • developing countries which may explain the shortfall. • A. ORIGIN OF THE REsEARCB QUESTION • Researchers develop a protocol from a research question (Figure 1). Who is asking the research • question? While some communities or service organiz.ations might approach researchers, most • research questions arise in universities or research institutes. Does it really matter who asks the • question? It does, because it establishes a sense of ownership, i.e., who will buy into the project • and its aims. • • Current Identify Identify Identify Research • Practice Priority subjects/ - Problem - Question - Community • • Alternative Identify Identify Identify Practice Community Priority Research • - (subjects) - Problem - Question • Figure 1. Origin of the research question • In most studies we have participated in, we began by identifying priority health problems • and then developed a specific research question. Only after deciding on a research design and • sample size did we consider which communities to enroll. Communities were selected on the basis of suitability to the research question and study design. These communities did not have • a direct involvement in asking the research question nor did they have a sense of ownership. • 221 • • • • • As an alternative, it is suggested that we begin by identifying communities to create • research partnerships with. Through working with these communities, priority health problems • can be defined and appropriate research questions formulated. Using this approach, research • questions are selected which apply to the community directly affected. Keeping some of the • participatory action research principles covered previously in this manual in mind will help • ensure that the research question is relevant to the community . • B. IIEALm ENHANCED ENVIRONMENTS Let's consider two randomized field trials of oral rehydration therapies in • the home management of acute childhood diarrhea. In both instances, large • portions of the populations enrolled did not have access to safe water, • which is a well established determinant of diarrheal disease. These trials • were not appropriate given our current knowledge, as a safe water supply must be available prior to assessing the efficacy of the oral rehydration • therapies . • We all accept that clinical studies should comply with ethical standards of conduct that guarantee all participants a minimum standard of . . . . • . . . . . care consistent with current knowledge, resource availability, and • .r . • affordability. Should field trials be any less accountable? No, they should : . .•.... • not. But they are, in fact, less accountable. They often remain detached . .... from participating communities, which greatly reduces the likelihood of the • translation of research findings into meaningful change...... , . . • . .•· .. . • Community health research should be conducted in a health .. • . . . ' .. • • enhancing environment, which means that basic health services known to .. enhance health are in place. For example, these could include safe water, ..•.. . . . • immunization, essential drugs, or access to health services. As health . . . • researchers and practitioners we have an obligation to provide these basic :- . . • health services to communities participating in research. Our obligation is . ... . , . • to insure that research is conducted in a technologically appropriate and . . . affordable health enhancing environment. For example, as a precondition • to carrying out a study related to early childhood growth, the research protocol might ensure that • each participating community will have in place a safe source of water, outreach EPI-MCH • programs, trained functional community health agents, and a referral plus follow-up protocol for • children identified as malnourished . • c. PARTICIPATION: WHO? WHAT? How? • A broad base of participation is a prerequisite to for effective community-based health research . • The researcher must consider who should participate, what they should participate in, and how . As indicated in Figure 2, identifying who and what is fairly straight forward. Far more complex • and less understood are the dynamics of how the who and what interact. We talk a great deal • 222 • • • • about participation, but have not looked at how best to put it into practice. For instance, how would one bring consumers into the data analyses phase of a study and the presentation of • results? • The study of participation is sadly neglected in health research. We need the objective • guidance such studies can provide, or our good intentions will not translate research into action. Despite this lack of knowledge, we recommend the completion of a table, such as in Figure 2, • as part of the planning of a research project. It will serve to maximize a broad based of • participation in a study. • WHAT? • WHO? PLAN CONDUCT ASSESS INFORM APPLY • Researchers • Practitioners • Communities HOW? • Ombudsmen • Funders • Policy Makers • Figure 2. Participation in research: Who? What? and How? • D. BARRIERS TO PARTICIPATION • Additional reasons why a broader base of participation does not occur are: • • It threatens autonomy and scientific integrity. By broadening the base we introduce • competing aims beyond that of knowledge and problem solving, e.g., support for special • interests. • • There may exist a lack of confidence or trust in the proposed partners. Researchers may • not be convinced that lay community representatives have the required skills or insight. • Often communities do not trust researchers, who they believe have their own agenda. • • There is a lack of incentive to change the way we do things. Those funding agencies that • require that participation at various levels occurs are an exception to this. • • • 223 • • • • • • Many people involved in problem solving such as policy makers, planners, or community • leaders, are not necessarily interested or comfortable with research findings. This could • be because they are not familiar with research terminology, the wish to use only • supportive data, or the lack of interaction with the researchers . • Broadening the base of participation can create insecurity and a sense of being out of • one's element for the researcher. There can be a loss of control over the process and the •••• researcher must learn to interact with people or groups s/he is not familiar with . • • The process takes time to bring in additional partners and mistakes can be made in trying • to work with new people, which can result in painful experiences . • Stress to the class that researchers involved in community-based health research need to • critically reassess how research projects are carried out. We need to identify important limitations to its application and impact. We need to develop a research agenda to objectively ,.• assess alternative approaches to participation in health research and the application of research findings. Finally, and equally important, researchers must develop interpersonal and • communication skills to match their scientific abilities, to deal with the many people at different levels of the problem-solving and decision-making process. i • • • Notable Quote "There is a probabilistic logic to scientific work and the fact that it is probabilistic • leaves conclusions open. Then there's the issue of generalizability--you can never • do the perfectly generalizable study . . . so whenever we move from the data to • extrapolating to the more general, . . . there will always be opportunities for people • on the other side of the fence to find holes." • • • • • • • 224 • • • • CASE HlsTORY • The instructor may choose this or another more familiar case history. This case and the • questions are provided as an example. • Setting: Central Highlands District in East Africa (elevation 3,000 meters) • Population: Mix of two dominant ethnic groups, one Muslim, one Christian • About 100,000 residents • Mean age 17 years • Services: One health center • Two health stations • A few health posts (village based) • Political Structure: Elected leaders at district, subdistrict, and village levels. Health • committees at district and subdistrict levels. An appointed District • Health Manager. • Problem: The District Health Manager (DHM) recently completed a health profile of the district. A leading cause of morbidity was found to be diarrhea. It primarily • affects children under five and is referred to as acute childhood diarrhea (ACD). • No mortality data is available for the district, but it is estimated that in the • country as a whole, over 1 % of children under five years of age die of ACD. • Environment: • Physical: - Highly contaminated surface water - Few safe water sites and no immediate plans to change situation • Health Services: - Recently upgraded and functional village based community health • workers in about half of the villages • - 50 % access to health services - very limited MOH financing • -Ministry of Health Control of Diarrheal Diseases (MOH-CDD) program: • specifically recommends oral rehydration solution (ORS) sachets, which • are manufactured locally • - Recent MOH directive to charge a small fee for ORS sachets • Other Vested Interests: - Recently opened ORS sachet manufacturing plant • "Knowledge" (research): - Recent in-country findings; - Diarrhea #1 killer in children under five • - Most mothers initially "starve" diarrhea and withhold fluids • - Most mothers know about ORS, but few use it ( < 10% episodes) • - A local cereal mixture is traditionally used and accepted • 225 • • • • • Research Objective: The DHM has met with the district and subdistrict health committees, presented the results of • the health profile, and indicated his concern about ACD and the need to prevent unnecessary • early childhood deaths. These committees support his recommendation to study the effectiveness • of setting up "oral rehydration therapy (ORT) comers" in each village for the early detection of ACD and to train mothers in the preparation of a home made-cereal based oral rehydration • therapy . • Human Development Objective: Decrease under-five mortality and increase local ,.• competence in the management of ACD . • • Action (Application) Objective: If demonstrated to be acceptable to communities, • effective, and safe, then to implement similar programs in other villages with the support of ~e • MOH . • • , .• • • • • : .• , .• • • • 226 • • • • • • • • • • • • • • • • • • • • 227 • • • • • • POSSIBLE SMALL GROUP REsPoNSES ,.• Leader's Guide • The following responses to small group assignments are meant to serve as examples. They are not exhaustive and are open to constructive criticism. They may be raised as appropriate during • the group presentations to the plenary . • Group 1: Should this research proceed • a. For (yes) • • addresses a priority health problem • • direct link to district health office and eventual program planning at a local level • • meetings held with district and subdistrict health committees • the proposed intervention is consistent with current local practice of preparing a cereal • mixture • b. Against (no) ;.• • not clear about community ownership and participation (informing does not equal participating) • • who is asking the question • • what basic health services are (and are not) in place • not evident that any meetings or role for MOH and other stakeholders has been defined ;.• • incomplete understanding of why ORS not used • • Group 2: • a. Ministry of Health Several options available to the research team. Should include j.• early discussions with the National Diarrheal Disease Control program and/or Division of Maternal & Child Health. One ' • could inquire about current policies and anticipated changes. ,. Questions to be asked could include; ,.• • What is their current position vis-a-vis home made M.O.H. preparations? ep i d~miol 09i sf • • What type of information would be needed to change ere- • current policy? • • 228 • • • • Does the CDD program or MOH plan to conduct any field work on oral rehydration • therapies or do they know of others studying this problem? • Is the MOH interested in participating in this proposed study? If yes, how? • • Does the MOH have specific questions which can be addressed by the proposed study? • b. Concrete steps: Following an initial meeting, the following options can be considered; • • Establish a project advisory team (committee) with CDD program representation • • Schedule follow-up meetings with CDD/MOH staff • • Provide written up-dates of project development and conduct • • Encourage and incorporate commentary/feedback • c. Other interest groups • • ORS manufacturers • Academic institutions (universities, institutes, technical schools) • • National or regional scientific and technology committees or agencies • • Health practitioners {public and private sector) • Group 3: Dissemination of results can be accomplished through several means. These include; • a. Publication • • scientific journals • • news letters • • local publications • b. Mass media • • radio • newspapers • electronic (internet) • • • c. Oral presentations to • • community groups • • women's and other special interest groups • • district and regional health offices • • MOH/CDD program • scientific meetings • • NGO's and multinational organizations • 229 • • • • • • d. Consider creative presentation techniques such as • • theatre • • puppets • Group 4: MOH: What should be happening • Options include: • setting up additional demonstration projects • • holding regional or national workshops on the management of ACD • • direct discussions regarding the policy implications of the research findings • • summarizing the current state of knowledge 1• • community and district lobbying for the further development of community-based early • detection and treatment programs • Group S: Community participation: Using Figure 2 as a starting point, the following • considerations are made; • a. Planning: In theory, communities can be actively involved throughout the planning phase . • Most important, they must view the research question as relevant to their concerns and local ,. needs. Beyond the question being asked, those directly affected by a health problem, such as ACD, can have valuable input into the specifics of the information sought. For example, this • might include qualitative data based upon informal interviews or community group meetings . • From this would stem a more comprehensive range of data collection at more formal and , .• quantitative stages of observation and measurement. b. Conduct: Community representatives can play key roles in the recruitment of subjects, • informing individuals and groups about the project, and carrying out specific data gathering • tasks . i. c. Assessment: An important role here is data interpretation or the exploration of alternatives . • d. Inform (dissemination): Ideas here can be matched with the responses from Group 3 . e. Application: Ideally, communities should be direct beneficiaries of new knowledge in terms • of improved services • • • • 230 • • • • REFERENCES • 1. Mechanic D., Politi.cs, Medicine, and Social Science. John Wiley & Sons, Inc. 1974. • • RECOMMENDED READING • 1. Murray C.J.L. Rational approaches to priority setting in international health. J Trop Med • Hygiene 1990;93:303-311. • 2. Taylor C.E. Health systems research: how can it be used? World Health Forum • 1983;4:328-335. • 3. Trostle J. Research capacity building in international health: definitions, evaluations and strategies for success. Social Science in Medicine 1992;35:1321-1324. • 4. Vlassoff C., Tanner M. The relevance of rapid assessment to health research and • interventions. Health Policy Planning 1992;7:1-9. • 5. Wilmoth T.A., Elder J.P. An assessment of research on breastfeeding promotion • strategies in developing countries. Social Science in Medicine 1995;41:579-594. • • • • • • • • • • 231 • • • • • • CHAPTER9 • WORKING WITH GROUPS • J. Allen-Peters • Much of our personal and professional lives is spent in groups. Most of us live in a group, • learn in groups, play as part of teams or groups, and carry out tasks and programs in groups . • Sometimes we act as the leader of a group and at other times we operate as a member of a team . • The degree of satisfaction we get from a group experience often depends on how it is structured and the dynamics among group members during the group's activities. To understand a group's • effectiveness in achieving its goals, we must understand what is happening to, and among, • individual members of a group . When we become involved in projects or programs which require community • participation, our group skills are just as important as our professional expertise. Yet, we are • often much less prepared for group work than for the medical, health, or research components • of a task. In this session we will look at a brief overview of group process and some guidelines • for working effectively with groups . • We have gathered our information from several years of lectures on group • communications given to university and college students and in workshops held for professionals . • We have adapted and summarized material from texts, manuals, and other sources. Some of • these are included in the reference list at the end of the chapter . • How DO GROUPS OPERATE? Every group operates on three levels: group task level, group maintenance level, and individual • needs level. None of the three levels should be neglected or ignored as they are equally • important and occur simultaneously . • Group Task Level • Most groups exist only, or primarily, to carry out a task. For example, the task of the staff at • a clinic may be to treat sick patients who come to the clinic. The need at this level is to • accomplish the task, but leaders and members of a group must remain aware of the other two • levels of need, which are going on at the same time . • Group Maintenance Level • The maintenance level refers to what is happening to the people in a group as a task is being accomplished. When people are working together on a task, they are doing something to and • with each other so there is a constantly changing network of interactions and relationships . • 232 • • • • Groups need to be aware of themselves as a group that has a lifecycle, a unique personality of • its own, and an energy and dynamism beyond that of any individual group member. The relationships within the group must be maintained if the task is to be accomplished successfully. • Individual Needs Level • :Each member of a group has a particular set of needs. These individual needs are often well hidden behind the task drive of the group. Yet if these needs are ignored or not recognized, the • efficiency of the group in achieving its task can be greatly reduced. • When a group finds the balance of these three levels of need -- task, maintenance and • individual - it becomes more effective and more mature. • • • • • • • • LEADERSHIP FuNCilONS IN A GROUP • It is easy to forget that leadership is not just the responsibility of the elected or chosen leader • of a group. For a group to be effective at all three levels of need, the leader and sometimes other members of the group must perform leadership functions. Depending on the situation or • task at hand, leadership passes around among the various members of the group. When • leadership functions satisfy the needs of group members, the group moves toward achieving its • objectives. • Leadership functions are required to meet the group's needs at the task level and the • maintenance level. • Task Level Functions • Leadership functions at the task level facilitate and coordinate group effort in choosing and • defining a common problem, and in working toward solving that problem. The leader should: • 233 • • • • • initiate: define the group problem, suggest a procedure or ideas for the solution, propose • tasks or goals • • seek information or opinions: ask for facts, seek relevant information about the group's • concerns, ask for ideas and suggestions • • give information or opinions: offer information, provide relevant facts about the group's • concerns, state a belief, make suggestions • • clarify or elaborate: interpret or reflect ideas and suggestions, clear up confusions, indicate • alternatives and issues to the group, give examples • • summarize: pull together related ideas, restate suggestions after the group has discussed • them, offer a decision or conclusion for the group to accept or reject • • consensus test: check with the group to see how many members agree • Maintenance Level Functions Leadership is needed to change or maintain the way group members work together, and develop • loyalty to one another and to the group as a whole. The leader should: • • encourage: be friendly, warm and responsive to others and their contributions, show regard • for others by giving them an opportunity for recognition • • express group feelings: sense feelings, moods, and relationships within the group; share • feelings with other group members • harmonize: try to reconcile disagreements, reduce tensions by pouring oil on the troubled • waters, have people explore and talk about their differences • •compromise: offer to compromise when one's own idea or status is involved in the conflict, • admit mistakes; discipline oneself to maintain the group's cohesion • • gatekeep: try to keep communication channels open, facilitate the participation of others, • suggest procedures for discussing group problems • • set standards: express standards for the group to achieve, apply the standards in evaluating • group functioning and production • Groups are likely to be most efficient when members perform both task and maintenance functions, and when these functions are seen as the responsibility of all members . • The two types of functions are not mutually exclusive. As a member performs a task function, • s/he may also be answering a maintenance need as well. For example, in summarizing related • items one may also express the feelings of the group members. In most groups it does not • matter who does what functions, as long as all the appropriate functions are performed by • someone. The functions any group member chooses to perform may reflect that person's • 234 • • • personality as well as the official position s/he holds in the group. For a group to work • effectively, the leader and the group members must recognize the individual needs of its members. • Four characteristics of individual needs: • Identity: Each person in the group may seek to answer personal questions such as: Who • am I in this group? What resources do I have that will be useful to the group? • What roles will I be asked to play in this group? Will being part of this group • affect the way I see myself? • Power and influence: Group members often are concerned with who will have the power and • influence in the group. How much will others exert their influence? How much • can I influence others? • Goals and needs: What are the needs of other group members? Will my needs be met? • These are questions asked silently and often by both leaders and members of groups. • Sometimes there is little information available in a group to answer these questions, • because the members are not aware of their needs, or are unwilling to share their • concerns and feelings with others. • Acceptance and intimacy: As part of a group, members often have to confront issues of their • needs, their hopes, and their feelings of adequacy and inadequacy in forming trusting, • close and personal relationships with others. For some people being alone is threatening; for others being close is even more difficult. Achieving an appropriate balance between • independence and closeness is often being worked out behind the scenes as the group • works at its task. • When the individual needs of group members are ignored, or unmet, people often become self­ • centred and do things such as: • • resist authority, or lean too much on the leader for authority; • • fight with others, or try to get one's own way in discussions; • withdraw from the group, physically or mentally; • • block the ideas of the group; • • form a small sub-group to resist the leadership of the larger group. • DILEMMAS OF LEADERSHIP • Each of us may find that our basic leadership dilemma is balancing what we think is right and • desirable, and what we actually do in practice as a leader. We may ask ourselves: How • democratic can I be and get things done? How authoritarian must I be to get things done on time? • • 235 • • • • • • Actually we face a whole series of dilemmas . . . • • In our tradition we tend to work competitively, but now we must be cooperative. • • We are under pressure to get things done, but we believe everyone's point of view must • be heard . • We are pushed for time, but we want participation in decision making, and this takes • time. • • We see the opportunity for quick results in one person decisions, but we believe that • shared responsibility means better, longer lasting solutions . • A Group Leader can Help a Group • An effective leader will give attention to both the task in hand and the people in the group, that • is to interactions at task, maintenance, and individual levels . • Before a session an effective leader will: • The Group Task The Group Maintenance • - prepare relevant material; - provide coffee or some form of welcome; • - organize equipment, rooms, and furniture; - consider the individuals and decide work, language • and materials, particular individual needs, size and • composition of group; - communicate task to group and give time for - work out rationale for the task to present to the • preparation; group members; • - define clear objectives for self and recognize - provide necessary prior information; • participants will have objectives also; look at skills, • knowledge, attitudes; - plan procedures to achieve task; - plan procedures that involve participants and • balances task/people; • - work out timing to fit task; - spend time with members if appropriate. - consider implications of time of day, make timetable • of what comes before, what will follow; • - choose the methods and teaching media appropriate • to the task and the group . • • • 236 • • • • During a ~ion the effective leader will: • The Group Task The Group Maintenance • - settle the group in; - be punctual • - clarify the task and procedures; - use warm, welcoming behaviour; • - present rationale, show why work is important; - build contract with group re what the group will do, how it will work, set ground rules; • - establish limits and boundaries; - use humour when appropriate; • - arrange physical environment to best effect; - avoid personal attacks or put-down; • - focus group on task; - use behaviours that convey respect, understanding, and genuineness to members; • - be sensitive to participants' wants; - listen, respond, bring in, achieve balanced • participation, ask questions; • - support relevant contributions; - encourage, support, give feedback; • - provide appropriate information; - manage conflict, model openness and acceptance • of individual; • - give time checks at intervals; - use names: check out feelings from time to time; • monitor member involvement; • - check understanding in group from time to time; - provide breaks or changes to help concentration; • - apportion task if that fits; - appreciate all contributions, builds them in; • - summarize progress from time to time; - manage time sensitively, ensure each member is • okay by the end; • - adjust task if anything more relevant emerges; - make physical environment as comfortable as • possible; • - manage time effectively; - use the •here and now•; • - keep to the point, but be flexible if more - respond appropriately to difficult members; significant topic arises; • - vary the work or method, maintain a flow; - avoid interruptions and outside distractions; • - contribute own ideas without taking over; - finish session positively. • - identify agreed action plan for follow-up; • - check what has been achieved; • - final summary • • 237 • • • • • • After a session the effective leader will: • The Group Task The Group Maintenance • - decide evaluation criteria; - ask for and give feedback; - evaluate achievement and learning. Have - check out individual experience and perceptions; • objectives been achieved? • - prepare any follow-up materials or work produced - follow up individuals who might need support, • by group; check out, give further help or special attention; • - follow up on group discussions and action plans, - allow a winding-down time; • or ensure that others do so; • - produce report on session if appropriate; - recognize achievement of each person; • - return borrowed equipment or other resources; - return members' work; • - return room to former condition or prepare for - spend time with members; • next; • - work out learning from session to apply next time; - monitor effect of group's work on system; • - file away materials that will be useful in future . - ask how the session might have been better; • - thank everybody who has helped . • • • • • • • • • • 238 • • • • Problem Behaviour in Groups • Every person comes to a group situation with needs s/he expects the group to meet. Some of • the needs are positive and may help the group to function. Other needs may disturb group • members and interfere with achieving the group's goals. • Here are a few ideas about handling problem behaviour: • • Culprit Possible Reasons for Behaviour P«mibilities for Dealing With It • Terry the Overly Talkative One May be an "eager beaver" Don't be embarrassing or • sarcastic; you may need those May also be exceptionally well­ traits later on • informed and anxious to show it or just naturally talkative Slow Terry down with some • difficult questions • May need to be centre of attention Interrupt with: "That's an • interesting point; now let's see • what the group thinks of it." • In general, let the group take care • of Terry as much as possible • • Andrew asks for your Opinion Trying to put you on the spot Generally, you should avoid solving problems for people. • Trying to have you support one Point out your view is relatively view • unimportant, compared to the view of other members at the • May simply be looking for your meeting. Don't let this become a advice phobia. There are times when • you must and should give a direct • answer. Before you do so, try to determine the reason for asking • your view. Say, "First, let's get • some other opinions. Joe, what do you thing about this point?" • (Select a member to reply.) • • • • 239 • • • • • Culprit Possible Reasons for Behaviour Possibilities for Dealing With It • Lee the Quiet One Bored Your action will depend on what • motivates the individual. Arouse Indifferent Lee's interest by asking for an • opinion. Draw out a member Timid nearby; then ask Lee to respond • that seated to person. H Lee is • Insecure near you, ask for the opinion • directly so Lee will respond to • you, not to the group . H Lee acts "superior" show that • you respect experience, then ask • for his opinion, but don't overdo it. You might cause the group to • resent Lee . Give praise the first time a • contribution is made. Be sincere. • Say, •1 think you have some good ideas on this. Can you collect • your thoughts and I'll ask you to • give them in a few minutes." • Pat the Quick, Helpful One Really trying to help; actually Cut across Pat tactfully by • makes things difficult as it keeps questioning others . others from participating • Thank Pat and suggest "we put • others to work," then use Pat for • summarizing . • • • Highly Argumentative Two or more members clash Emphasize points of agreement, • minimizing points of disagreement • Can divide your group into (if possible). factions Draw attention to objective. • Cut across with direct question on • topic . Bring a neutral member into the • discussion. • Frankly ask that personalities be • omitted . • • 240 • • • • Culprit Po~ible Reasons for Behaviour Pcmibilities for Dealing With It • Ozzie the Obstinate Won''. '.ludge. Throw Ozzie's view to the group • and have group members Prejudiced straighten him out. • Hasn't seen your point of view Say time is short and you will be glad to discuss it later. • Ask Ozzie to accept group • viewpoint for the moment. • • • Has pet peeve Get the feeling of the group. • Professional griper Suggest private discussion later. • Has legitimate complaint Have a member of the group • answer. • • Sue and Sally Side Conversations May be related to subject Don't embarrass them, but call each one by name and ask an easy • May be personal question. • Distract members and yourself Call one by name, then restate last • opinion expressed or last remark • Don't feel comfortable voicing made by group member, and ask views to whole group but want to for an opinion. • have a say • If, during meeting, you are in habit of moving around the room, • saunter over and stand casually behind those talking. This should • not be obvious to the grouo. • • • • 241 • • • • • Culprit Possible Reasom for Behaviour Pouibilities for Dealing Wdh It • Inez the Inarticulate Lacks ability to put thoughts in Don't say, "What do you mean by • proper words this?" • Gets an idea but can't convey it. Say, "Let me repeat that" (and • Needs help. then put it in better words.) • Twist Inez's ideas as little as • possible, but have them make • sense . • Definitely Wrong Member comes up with a Say, "I can see how you feel" or • comment that is obviously "That's one way of looking at it. • incorrect . Say, "I can see your point, but can you reconcile that with the • true situation?" • Must be handled delicately . Ask, "How do the rest of you feel • about this?" • • Handling Confrontation in Meetings • The following ideas should help you handle confrontations in a positive way . 1. Clarify Objectives: An important first step is to clarify and reach agreement on objectives . • Conflict can develop if participants have different understandings of the objectives of the • meeting . • 2. Strive for Understanding: As the meeting leader, you may have to stop a confrontation and make sure each party can state the opposing party's position and supporting reasons. When • involved in an argument, people often do not listen carefully to the opposition's side because • they are thinking of a rebuttal . • 3. Focus on the Rational: For the benefit of the outcome of a confrontation, you should keep attention focused on rational consideration -- facts, supporting reasons, potential problems if a • certain course of action is followed, etc. Emotional involvement is a natural part of • confrontation but sound decisions cannot be reached when participants are too emotional. 4. Generate Alternatives: A challenging part of the process is finding alternative solutions to • satisfy the needs of the diverse points of view involved in the confrontation. Participants often • cannot accept an alternative solution to their own. Moderate group members could generate • some reasonable alternatives . • 242 • • • S. Table an Issue: Tabling can be an effective way to deal with conflict when you feel a party • needs time to consider the arguments that have been presented. It works particularly well as a facesaving device. People sometimes find themselves in a position of having argued so strongly • for a position they cannot gracefully change even after being convinced of the logic of a different • position. Tabling gives a person time to work this out. • 6. Use Humour: If you are good with humour you can use it to reduce the emotional tension • in a confrontation. It can serve as a release and clear the way for more rational problem­ • solving. • Other items that will help handle conflict in meetings include: • • acknowledge deadlines • • involve everyone in the process • • allow time to think • take a break (i.e. call a time out) • • refer items to a subcommittee • • allow expression of strong feelings • • protect the group from early closure • When there is a disagreement between two groups on how to resolve a problem, you can stop • the discussion and ask each group member to indicate where he or she stands on the following • scale: . • FOR S 4 3 2 1 0 1 2 3 4 S AGAINST • When group members have indicated their positions, ask them to post their score on a • flip chart. Then, ask group members who are for a given position to explain why other • members are against the position. Ask those who are against the position to explain why the • others are for the position. In this way, you can: • • provide a quick way to get the issues surfaced • • get all pro and con ideas out in the open • • ensure that one group is listening and understands what the other group is saying and • why • finally, have those who are neutral on the issue offer alternatives to integrate the needs • of those for and those against • • • • 243 • • • • • • CONFLICT • Conflict should not be avoided in meetings. It is a natural outcome of strongly held points of • view. However, it must be contained and focused toward resolution . • All conflict can be resolved. Not that it always will be - but it can. Most often it is resolved • through some communication. One expert estimates that 70 3 of conflict can be handled by simply using clear communication, 203 will require negotiation, and the remaining 103 can • be resolved through arbitration or the use of a third party . • • • DECISIONS BY CONSENSUS • Consensus is a decision-making process that makes full use of available resources and • resolves conflicts creatively. Consensus is sometimes difficult to reach, so not every group decision can be made in this way. Complete unanimity is not the goal - it is rarely • achieved. But each individual should be able to accept the group's decision on the basis of • logic and feasibility. When all group members feel this way a consensus has been reached and the judgment may be entered as the group's decision. Following are some guidelines to • achieve consensus: 1. Don't argue for your position. Present your position as clearly and logically as • possible, but listen to the other members' reactions and consider them carefully . • 2. Don't assume that someone must win and someone must lose when discussions reach • a stalemate. Instead, look for the next-most acceptable alternative for all parties . • 3. Don't change your mind simply to avoid conflict. Be suspicious when agreement • seems to come too quickly and easily. Explore the reasons and be sure everyone accepts the solution for basically similar or complementary reasons. Yield only to • positions that have objective and logically sound foundations . 4. Avoid conflict-reducing techniques such as voting, averaging, and bargaining. When • a dissenting member finally agrees, don't feel that he or she must be rewarded by • being allowed to "win" on some later point . 5. Differences of opinion are natural and expected. Seek them out and try to involve • everyone in the decision process. Disagreements can help the group's decision • because with a wide range of information and opinions, there is a greater chance that • the group will hit upon a more adequate solution . • • 244 • • • 1 • HOW TO BE AN EFFECTIVE LEADER • LEADERS: BORN OR MADE? • • You've heard the expression "He's a born leader." Are all leaders born leaders, or are leadership skills learned? What makes a good leader? Governments, businesses, the military • and other organizations have been asking those questions for centuries. • At one time it was thought that leaders were born. This is particularly true of autocratic leaders who must act decisively in times of extreme pressure. A great emphasis • was placed on action as leadership and on the personality or charisma of the leader. • Today, it is felt that leadership is more of a helping process. It is the process of • meeting the needs of people through an understanding of human behaviour. • More recent studies have found that: • • leadership skills can be learned • different styles of leadership are appropriate for different situations • • LEADERSHIP SKILLS • Effective leaders have many common qualities. Successful leaders make an effort to learn • and practice all of the skills listed here. Good group leaders: • • have a knowledge or understanding of the goals and purpose of the organization • are able to communicate that understanding or vision of the future to both the • members and to those outside the organization • • know how (or have learned how) to work with others • • have made time in their lives for organization -- they are well organized • • have good communication skills -- they can express themselves clearly and with confidence • • can live with a certain amount of conflict and realize it is usually impossible to • please everybody all the time • • can inspire others to do more than ever thought possible • • 1(This information used with permission of the Ontario Ministry of Agriculture and Food, from their Factsheet • Series, 1988) • 245 • • • , . • • LEADERSHIP STYLES • Theories of leadership styles have been developed through years of research with groups and • organizations. The most popular current theory is situational leadership. The appropriate • choice of style or method for a leader to use depends mostly upon the group's level of readiness; that is, how willing and able the members are to take responsibility as a group . • Willingness relates to attitudes of the group and means that the group has the confidence, • commitment, and motivation to accomplish a specific job or activity. Being able or having • ability implies that the group has the knowledge, skill, and experience to accomplish a • particular task. , .• The Four Leadership Styles and How to Use Them • 1. The Telling Leader • This leader tells the members what to do, and does not worry too much about the feelings or • relationships within the group. This is appropriate where the members are new or • inexperienced and need a lot of help and direction in order to get the job done . • 2. The Selling Leader • Here the group is a little more responsible or experienced but the leader sells and persuades • the group to buy into the job psychologically. Direction and guidance by the leader is still • necessary at this level. • 3. The Participating Leader Leaders using this style know their followers have the ability to do the job. Therefore the • group members and leader make decisions and carry them out together. Here the leader puts • more emphasis on the relationships within the group than on the tasks to be done . • 4. The Delegating Leader • A leader chooses to be a delegator and trusts the group to do their own thing when the group • members are both willing and able to run their own affairs . • How to Choose the Most Useful Leadership Style • You must keep in mind that no one style of leadership is appropriate for every occasion or • situation. You may feel most comfortable with one of the four styles and so use it the most. To be effective as a leader, however, it is critical to know your group -- where they are at in • ability and knowledge, and desire and willingness -- and be ready to adapt your leadership • style to suit the occasion . • • 246 • • • • When you have figured out approximately how the group is functioning, then you can • decide which leadership style to use for that occasion (See Figure 1). • For instance, if you are normally comfortable as a participating leader, but your • group has a lot of new members and an important job to get done in a hurry, it would be • appropriate for you to become more of a selling or even a telling leader, and provide more direction to the members. • In summary, keep in mind that: • • while some leaders are born, most learn the skills of leadership through trial and error on their own and from the experiences of others • • group leaders have many qualities in common that we can develop in ourselves • • there are various styles of leadership that we can use for different group situations • • • • • • • • • • • • • 247 • • • • • • If the members in your group: Use This Leadership Style By using these specific actions • -are new (inexperienced) -give clear directions • -have little knowledge or ability -follow-up • about the job TELL -give feedback • -lack confidence in doing the work • -are somewhat experienced -explain decisions • -need help " buying into" the -give lots of • decisions that have to be made SELL opportunity for • clarification -need coaching as to how to do the • job • • -know how to do the job, but -share ideas • lack self confidence PARTICIPATE-+ -help members in decision making • -need to share decision making -encourage and • with each other and the leader compliment • • -are quite experienced -tum over responsibility • -are capable of doing the job for decisions to the • -want to do the job DELEGATE -+ members and allow the -are responsible for directing members to carry out • their own affairs those decisions • • Figure 1. How to Choose the Most Useful Leadership Style • • • 248 • • • • LIFE CYCLE OF A GROUP SYSTEM • Groups are complex systems. They have a life cycle and personality of their own. Many • things can cause success or failure and there are many different ways to reach the group's • objectives. • Phases in Group Growth and Development • Forming: the group comes together and relationships emerge. • Storming: conflicts break out and are resolved through discussion, listening, etc. • Norming: rules and standards for group and individual behaviour emerge; cohesion begins to • develop. • Performing: roles are assigned within the group. Members begin to give attention to the • task at hand. • During the forming stage, often a kind of primary tension is felt because of the • newness of the group. Symptoms are uneasiness, boredom, hesitation, politeness, soft • voices, pauses in the conversation. To help overcome this tension, use exercises to "break • the ice," get-acquainted activities, jokes, social chitchat and a welcoming cup of coffee. • During the storming stage, secondary tensions related to the task at hand are usually • felt. People disagree, interrupt each other, fidget, speak loudly, even shout. To overcome • secondary tension and progress to the norming stage, individual members can show their • agreement with ideas as they are expressed, indicate that even if they don't agree, there is still solidarity in the group and an allegiance to a common goal. Humour is always a great • help! • Tertiary tension will erupt as the group performs and works toward its goals. Power • struggles, disagreements over who is making the rules, the rights of individual members, and • questions of who is in charge emerge. Members should demonstrate tolerance and understanding for the disagreements, and know that their ideas are appreciated and valued • even when they represent the minority opinion. The leader should express confidence in the • group's dependability and use humour and a light touch, with care, to break and relieve • tension. • Remember: Conflict is a normal part of group growth and can be constructive. • Group Norms • Group norms are standards of behaviour that group members impose on each other and • represent the collective evaluation and expectations of the group. A group culture evolves • 249 • • • • • • from the group's values and ideals, which in tum generates norms that contribute to its cohesiveness, the personal satisfaction of individual members, and its effectiveness. Any • individual may challenge group norms, and norms do change and evolve over time. • Sometimes norms are rigid and lack creativity if they cannot adapt to changing circumstances • and expectations . • in • The Importance of Cohesiveness Groups • A group is distinguished from just a bunch of people sitting around and talking by its sense • of togetherness, belonging, or cohesiveness, as it is called by the specialists . • Cohesive groups: • • interact more with each other • are more satisfied with their group's performance • • develop useful norms for their group behaviour • • work well as a team in an emergency • • How does a group become cohesive? • If you were to give tips for encouraging a group's cohesiveness, they might include the •e following: • Encourage lots of effective communication • • Stimulate each member to act in ways that help the group • • Be sure the group develops its own norms for acceptable behaviour • • Face conflicts openly and resolve them in a way that the group can agree on (e.g., • consensus) • Cohesive groups are productive and satisfied with their performance . • Group Think • Group think represents the downside of cohesiveness. It refers to the tendency for cohesive • groups to fail to critically analyze group information, ideas, proposals, decisions, etc . Groups that have been together a long time, or have a lot of power in their particular area, • may avoid analysis of information and arguments, avoid seeking outside opinions and input, • avoid controversial issues, take decisions quickly without enough consideration, or exert • pressure on any member whose opinion might vary . Brainstorming alternative solutions to problems might help to prevent group think • from occurring. By definition, brainstorming is done to bring out as many creative ideas as • possible, without any evaluation of ideas allowed. Then, give the alternative solutions to • 250 , .• • • • small subgroups to discuss and develop, and encourage each group to prepare a case in • support of each alternative. Each group presents its alternative to the whole group and defends it. The final solution may be one of these alternatives but often is a modified • version, or a synthesis of the best features of each alternative. • Working Together as a Team • Even with the best will in the world, people don't just walk into a room, introduce • themselves as part of the same team, and start working together productively. Teams • develop slowly, by establishing good relationships among members, norms and procedures • for the group, and by being very clear on what the goals are, and how committed each team member is to achieving those goals. • You might fmd this model helpful in starting a new team activity. • A. Forming the team: Who are we? • Have members of the team share their expectations about the team task, their personal needs • and hopes that the group can meet, and how they are feeling right now. Use incomplete sentences to start the discussion, such as: • "When I first become part of a new group I feel . . . " • "I usually try to make people think I am . . . " • "What I'm afraid might happen is ... "" • B. Deciding on procedure, or how the group is to work together • Have the group first decide how it will operate, e.g., how to make decisions, how to • approach the task, and the best procedures in view of the task. The group should anticipate some of the problems that could block the team from working together effectively, and • decide on methods to work through the blocks. It's easier to avoid problems if procedures • are established and agreed upon before the work actually begins. • C. Clarifying the goals and tasks, or what the group will do together • Have the group establish goals at the beginning to avoid team problems. A team may find it • helpful to dream about the potential ideal results, then explore the strengths and resources of • each group member, and the different ways of stating the group's goal. Review the purpose to be sure that everyone understands it. Make sure the group thinks about how the goal and • the tasks ahead relate to the individual needs of group members. • • • 251 • • • • • • D. Analyzing the factors which will help or hinder achievement of the goal • A SWOT analysis lists the strengths, weaknesses, opportunities and threats in relation to a • particular plan of action. Have members of the team do a SWOT analysis or other exercise • to identify the forces they expect may affect the achievement of the goal . • E. Planning the tasks necessary to achieve the goal or how the group will get there • Have the team brainstorm the action steps needed to accomplish the goal. They should then • rank the different ideas or plans by desirability or disability. After choosing one plan and working out the procedure to implement it, have the team work out an alternative Plan B, • just in case! • F. Planning the evaluation or how the group will know the goal is accomplished • Have the team plan an evaluation of its effort from the very beginning and identify the • criteria they will use to decide if and how well the goal has been accomplished . • G. Reflecting on process • Have the team think about what happened during this exercise, and how effectively they have • worked as a team in this activity . • H. Summarizing • Have each team member summarize their conclusions about the team building that occurred • during this activity . • FEEDBACK • We can learn how our behaviour in a group affects others through feedback. Simply, • feedback is providing information to a person or a group about how the person or group affects other people. Feedback helps group members become sensitive to their actions and • the reactions that they cause in others, and helps us to consider changing our behaviour if it • is communicating something that we aren't feeling, or trying to communicate . • Feedback can be: • CONSCIOUS - nodding agreement, or UNCONSCIOUS - falling asleep! • SPONTANEOUS - "wow!" or SOLICITED - "tell me . . . " VERBAL - "No." or NONVERBAL - leaving the room • FORMAL - a questionnaire or INFORMAL - applause • • 252 • • • • Feedback is helpful: • Feedback can reinforce or confirm desirable behaviour by encouraging you to repeat it. • "You really helped me understand better when you clarified that position ... " • Feedback can help people identify or clarify their relationships. • "Jane, I thought we were on opposite sides of this issue, but we really aren't are we?" • Feedback should be: • • descriptive rather than evaluative. If you describe how you feel or react to something • someone has said or done, the individual is free to use the feedback or not. Avoid • evaluative language so that the person is less likely to react defensively. • • specific rather than general. Use specific examples to illustrate a point. For • example, if someone thinks someone dominates decisions, they could say, "Just now • when we were deciding the issue and you didn't listen to what others were saying, I felt I had to accept your arguments or face an attack from you." • • appropriate to take into account the needs of the giver and the receiver of feedback. • If feedback only serves your own needs and fails to consider the needs of the person • on the receiving end, it can be destructive. • • usable. Reminding a person of some shortcoming over which they have no control is • not usable or helpful. It only increases frustration. • • requested rather than imposed on people • • timely, at the earliest opportunity after the behaviour or situation occurs • • clear; the receiver could paraphrase the feedback to be sure that s/he heard it correctly • • accurate. Is the feedback an impression shared by others in the group or just one • person's impression? • If you want feedback, ask for it, accept it, ask for clarification if you don't • understand and then respond to what you've heard. • • • 253 • • • • • If you want to give feedback, it should be done in a way that is helpful to others. Be • aware of your feelings and let the person know that's how you feel. For example, • "When you ... (state the behaviour), I feel ... " (state how the behaviour makes you feel) This provides information to the other person or group without arousing defenses and allows • them to decide whether they will use it. • Observing Your Group • Why should you observe the group with which you are working? It can enable you to reflect •• on the group process, or how the group is working together. Observing helps you answer • questions, such as: Are we communicating? Is everyone participating? How are decisions • being made? • Even if you are a group participant, observing can provide data to use to make • changes to increase the group's effectiveness, or to support the useful behaviours that are • making the group experience successful and worthwhile to the group members . • A number of observation tools are useful for observing your group. • Some allow you to reflect on your own actions within the group, others allow evaluation of • meetings of the group to enable each group member to think through how the group is • growing and developing. These tools are provided as guidelines for reflection and evaluation • of your group and its performance. • Group work is an essential component of community work, whether it involves needs • assessment, project implementation or participatory action research. The more you • understand about groups the more satisfying your experience will be. Good Luck! • • • • • • • 254 • • • • GROUP OBSERVATION TOOIS • A. How DO I BEHAVE IN A GROUP? • Answer the following as honestly as you can. The answers will help you to decide which skills and behaviours you wish to work on. • Circle your answer. • When I am in a group, I usually ... • 1. trytogetthingsstarted YES NO • 2. sit back and wait for someone to lead YES NO • 3. daydream and think of something else YES NO • 4. have so many ideas I can't help but interrupt YES NO • 5. try to support others YES NO • 6. criticize ideas that seem stupid or irrelevant YES NO • 7. feel inferior to others in the group YES NO • 8. feel superior to others in the group YES NO • 9. contradict, disagree or argue with others YES NO • 10. agree with everyone in order to get along YES NO • 11. don't speak up, but tell my friends afterward how I YES NO • felt • 12. prefer not to lead because others will only criticire YES NO • whoever takes charge • *usually - more than 50% of the time • Write here any other behaviours that you think you exhibit when you are in a group. • • 2SS • • • • • B. CHECKLIST FOR EVALUATING YOUR OWN PERFORMANCE IN YOUR GROUP • Put a check (4) to indicate how frequently you believe you perform each function when you are in your group. • Fill this out early in your group project work. Repeat this evaluation at the end of your group experience. Has • there been any change in your performance? • Functiom Always unen Seldom Never • Ask for information • Give information • Ask for opinions I • Encourage others to • speak I • Support others' points • of view • Summarize discussion I • I • Make suggesllons l • Harmonize conflicts I • I • Follow suggestions ! • Encourage agreement I • Seek recognition ;:speaK clearly and to the • point • I • Definitions: I • Always: every time I Often: more than half the time • Seldom: less than half the time • Never: never! • Date: • • 256 • • • • C. AFrER THE MEETING IS OVER • • 1. I thought this meeting was: (check one of the following): • • very poor fair mediocre good very excellent • poor good • • • 2. The strengths of the meeting were: • • • • 3. The weaknesses of the meeting were: • • • 4. This meeting would have been better if: • • • 5. At our next meeting I would like to see: • • 257 • • • • • This next part of the evaluation deals with your reactions to what went on at the meeting . • Fill in the blanks as directed: • THINGS PEoPLE Dm: (Fill in the blanks with descriptions of behaviour or actions regardless • of who the person was.) ••• 1. The most helpful thing that anyone did today was: • • 2. I was somewhat surprised at: • • • 3. Today we really needed someone to: • • 4. Of the things people did, I tended to admire: • • • 5. I really hindered our progress when: • • 6. It was disappointing to me that: • • • 7. I personally felt somewhat uncomfortable when: • • 258 • • • • 8. It gave me a comfortable feeling that: • • 9. I felt I disapproved of : • • PEoPLE IN THE GROUP: (Fill in the space below with the names of people in the group. Put • in as many names as you feel should go in a space, including yourself). • 10. During the meeting, I felt I agreed mostly with what was said by • • 11. Of everyone, seemed to get the most • out of the meeting. • 12. ------and I seemed to understand one another quite • well. • 13. I felt the group didn't treat------very fairly. • 14. In general, I could not agree much with what was said by • 15. I thought------got less out of the meeting • than the rest of us. • 16. There seemed to be a communication block between ------and me. • 17. ------got quite fair treatment from the • group. • THINGS I Dm: (Fill in the blanks below with what you did). • 18. I felt I had to control my impulse to------• • 259 • • • • • • 19. I felt I did not do a very good job at------• 20. I had a feeling of accomplishment when I ------• • 21. I feel I would like to change the way in which I ------• • General comments on the meeting: • • • • • • , .• • • • • • ,•• 260 • • • • How FREE Do WE FEEL: • Below is a list of typical group reactions. Please indicate how free you felt to express these • reactions in your meeting. Use this key. • 1. Put I in the appropriate space at the right to show how free YOU personally felt. • 2. Put 0 in the appropriate space to show how you think most of the OTHERS in the group • felt. This may be the same or different from how you felt. • 3. Put W in the appropriate space to show how free you personally WISH you had felt or could feel. This may be the same or different from the I and the 0. • ** See example below for a typical reactions and answers. • Reaction Degree of • Freedom • Very free Fairly free Hampered Restricted • ** Example: Admit an w 1,0 • ineffective or undesirable behaviour • of one's own. • A. Point out lack of progress when the • group seemed bogged • down. B. Offer a comment in • the capacity of • spokesperson for the group. • C. Say something to put the group at ease in • a moment of tension. • D. Comment on ineffective or • undesirable behaviour • of other members. E. Indicate to the • group one's own real • feelings about what was going on. • F. Criticize the leader • (chair) openly when her/his efforts did not • satisfy the group. • G. Ask for comments on own behaviour. • • 261 • • • • • • ENERGIZERS CAN BE USED IN MANY WAYS • • • ; .• • • Energizers can: • • prepare people for new learning • • alter the pace of activities • emphasize important points • • summarize discussions • • provide invigorating breaks • • stimulate creativity and lateral thinking , .• • Use these energizers anytime during the workshop. They may be particularly useful to wake • people up if a session becomes long and attention seems to be drifting . • Remember our Strings • Give each group of 5-8 people a long rope with several knots tied in it. Each person holds onto the rope with one hand. Without moving the hand that holds onto the rope, the group • unties the knots . • The Point: We all bring our own "strings" to group work. We may be unable or • unwilling to drop these strings. Such strings include professional bias, gender bias, religious beliefs and cultural experiences. By acknowledging these strings, especially our own, • seemingly insurmountable problems can be solved. New ways to "untie the knots" are • needed. Often the most obvious approach i.e. dropping the string is not the best solution . , .• • • 262 , .• • • • Don't Forget the Villagers • Create a "story" which is relevant to the task. Give participants directions as to their • identities in the story and the actions linked to specific words. For example, when your • profession is mentioned - stand up; when an animal is named - make the sound of that animal; at the word "village" everyone stand and link arms across the table. Use your • imagination. Keep people moving. • Now, proceed to tell the story . . . • "Once upon a time in the highlands of Papua New Guinea, the villagers gathered to • celebrate the marriage of a young girl and a young man. At the wondrous feast, the • villagers enjoyed the delicacies of the region - wild parrot, roast chicken and pig as well as • many kinds of fruits and vegetables. Soon after this event some of the villagers fell ill and • eventually five died. • With such a tragedy in the villages, the elders gathered to discuss what they might • do. Were the gods angry, the villagers asked? Had women broken a sacred taboo? Had the • men? A young man stepped forward for he had remembered the people from the city who • had been asking questions about deaths in the village. • Continue the story with physicians, nurses, data collectors, veterinarians, all • becoming involved in "solving" the problem. Until villagers are involved however none of • the solutions work. • The Point: Those who participate share ownership in a project and therefore have a • stake in its success. "Don't forget the villagers" can become a rallying cry for participation. • Remember the Pigs • A variation on the above can be used to bring awareness to occasions when the community • takes an action that outsiders deem inappropriate. • Tell a story in which the pigs are found to be the • source of the infection. The villagers decide to kill all pigs. • The Point: Any form of new knowledge or technology changes a community. It is often difficult to forecast such • changes. Therefore, everyone has a responsibility to assist in • the most beneficial integration of new information while • remembering that the community must lead in this decision • making process. "Remember the Pigs" can become a cautionary note when change is contemplated. • 263 • • • • • • Comfort Zones • Ask two volunteers to sit side by side on the floor. Cover them with a blanket. They are • asked to remove one item which will make them cooler. When the correct item is produced • you can join the group . ,.• "Now imagine that you are in a desert with the sun beating down upon your heads. You are getting hotter and hotter but there is no water close by." Continue to build the • story . • The Point: When faced with an uncomfortable or potentially embarrassing situation • most people do not think at their best. Rarely do they make the wisest or most obvious • decisions - such as removing the blanket. Instead, rings, shoes, or necklaces appear from • under the blanket. Many of us respond in much the same way when facing a new experience. We are nervous, shy, anxious and not "at our best". To ensure that we all work • more effectively in groups it is important to take time to make everyone a little more at ease . • • Thanks to Kay Wotton for writing up the Energizers • • • • • • • • • • • 264 • • • • RECOMMENDED READING • 1. Brilhart, JK and Galanes, GJ. Effective Group Discussion. Seventh Edition. Wm. • C. Brown Publishers, 1992. • 2. Cragan, JF and Wright DW. Communication in Small Groups: Theory, Process and • Skills. Fourth Edition. West Publishing Company. 1995. • 3. Dimock, HG. Leadership and Group Development: Four Monographs • Factors in Working with Groups, Second Edition 1983 How to Observe Your Group, Second Edition 1985 • How to analyze and Evaluate Your Group, Second Edition 1985 • Planning Group Development, Second Edition 1986 • 4. Galanes, GJ and Brilhart, JK. Communicating in Groups: Application and Skills. • Second Edition. Wm.C. Brown and Benchmark Publishers. 1993. • • • • • • • • • • • • 265 • • • • • • WRAP-UP • J. Pickering • • The two objectives of this session are: • 1. to help the participants to consolidate their learning during the workshop, and to think • concretely about how they will apply what they have learned 2. to help the facilitators learn what could be done to improve this or other similar • workshops in the future • Objective 1. • The following two activities were used at the Montreal workshops as wrap-up activities . • a. The Talking Stone Circle: This is a tradition from Canadian aboriginal groups . Everyone sits in a circle and a stone is passed from person to person. As the stone is • passed around, each person states the most important thing he/she learned during the • workshop. Only the person holding the stone may speak . • b. Letter to Myself: Participants wrote letters to themselves, expressing what they intended to change over the next year as a result of the workshop. The participants were given • the option of sealing the letter, or leaving the letter open for the facilitators to read. • Letters were collected and mailed back to the participants 8 months after the workshop • as follow-up so the participants could see if they had done the things they had written in • their letters . • Facilitators or participants may have other ideas for activities that would be suitable as • wrap-up activities . • Objective 2 • • Much useful information about how to improve this or similar workshops will have been • gathered by observing the day-to-day activities of the workshop, and during the daily evaluation • sessions . • An evaluation questionnaire can also be completed by the participants. This is ,. particularly useful for participants who may be inhibited about speaking up about their feelings • during the workshop. Both quantitative and qualitative information should be collected on the • 266 • • • • questionnaire. Often the qualitative information, i.e., the comments written in, are the most useful. The sample questionnaire on the following three pages is very similar to the one used • during the Montreal workshop and can serve as an example. • Loose ends can also be dealt with during wrap-up. For example, at the Montreal • workshop, a list of all the resource material at the workshop and a list of funding agencies for health research in developing countries were requested. Since there was insufficient time during • the workshop to complete these lists, they were typed up and mailed to the participants after the • workshop. • • • • • • • • • • • • • • • 267 • • • • EVALUATION • Evaluation is crucial to the development and refinement of future courses and workshops. Please take a few • moments to share your feedback with us. Please rate each session on its CONTENT, FORMAT and RELEVANCE to your needs, and the READINGS provided, by circling the corresponding number. Please • use the back of the page for your additional comments . • • THE WHY AND HOW OF RESEARCH • Poor Somewhat Adequate Good Excellent ,.• Adequate CONTENT 1 2 3 4 5 • FORMAT 1 2 3 4 5 RELEVANCE 1 2 3 4 5 • READINGS 1 2 3 4 5 • Comments: • • INTRODUCING PARTICIPATORY ACTION RESEARCH • Poor Somewhat Adequate Good Excellent • Adequate • CONTENT 1 2 3 4 5 FORMAT 1 2 3 4 5 • RELEVANCE 1 2 3 4 5 • READINGS 1 2 3 4 5 • Comments: • EPIDEMIOLOGIC, DEMOGRAPIDC, AND BIOSfATISI1CAL METHODS: • ASSF.SSING NUMBERS, RATES, CAUSES AND DETERMINANTS OF MORTALITY • AND MORBIDITY IN DEVELOPING AREAS • Poor Somewhat Adequate Good Excellent • Adequate CONTENT 1 2 3 4 5 • FORMAT 1 2 3 4 5 • RELEVANCE 1 2 3 4 5 • READINGS 1 2 3 4 5 • Comments: • • 268 • • • • METIIODS FROM SOCIAL SCIENCES OVERVIEW OF QUANTITATIVE AND • QUALITATIVE METIIODS • Poor Somewhat Adequate Good Excellent • Adequate • CONTENT 1 2 3 4 5 • FORMAT 1 2 3 4 5 • RELEVANCE 1 2 3 4 5 READINGS 1 2 3 4 5 • Comments: • • EVALUATING COMMUNITY PARTICIPATION • Poor Somewhat Adequate Good Excellent • Adequate • CONTENT 1 2 3 4 s • FORMAT 1 2 3 4 s RELEVANCE 1 2 3 4 s • READINGS 1 2 3 4 s • Comments: • • PRACTICAL ISSUES IN HEALm RESEARCH IN DEVELOPING COUNTRIES • Poor Somewhat Adequate Good Excellent • Adequate • CONTENT 1 2 3 4 s • FORMAT 1 2 3 4 s RELEVANCE 1 2 3 4 s • READINGS 1 2 3 4 s • Comments: • • • 269 • • • • • LABORATORY ISSUES IN DEVELOPING COUNTRIES Poor Somewhat Adequate Good Excellent • Adequate CONTENT 1 2 3 4 5 • FORMAT 1 2 3 4 5 • RELEVANCE 1 2 3 4 5 • READINGS 1 2 3 4 5 • Comments: • TRANSLATING RESEARCH INTO ACTION , .• Poor Somewhat Adequate Good Excellent • Adequate • CONTENT 1 2 3 4 5 • FORMAT 1 2 3 4 5 RELEVANCE 1 2 3 4 5 • READINGS 1 2 3 4 5 • Comments: • • WORKING WITH GROUPS • Poor Somewhat Adequate Good Excellent • Adequate • CONTENT 1 2 3 4 5 • FORMAT 1 2 3 4 5 RELEVANCE 1 2 3 4 5 • READINGS 1 2 3 4 5 • Comments: • • • 270 • •