necessary. A national waiting list policy that whether these policies are subject to review and current screening programme and that a lack of guarantees access according to agreed criteria, whether such criteria are explicit. staff would limit an immediate expansion.' rather than simply guaranteeing maximum wait- We have recently completed a cross sectional Age Concern England's view is that the funda- ing times, could facilitate contracts between pur- postal survey ofthe 106 new health authorities in mental issue of the equity of the NHS breast chasers and providers on the basis of agreed England and Wales, requesting details ofpolicies screening service should not depend on further criteria for the prioritisation of waiting lists. on in vitro fertilisation. The response rate was information. The question that needs to be Points could be assigned to patients by 94% (100/106). Of the 100 responders, 67 fund answered is why the programme of breast consultants to reflect, for example, current in vitro fertilisation and 16 do not. The remain- screening invitations excludes the age group of health status, assessed using the EuroQoL or ing 17 were in the throes of the policy making women in whom the largest number of cancers SF36 health status measures; expected gain in process and were unable to offer a clear might be detected. This reduces the potential quality of adjusted life years (QALY)3 from response. effectiveness of the breast screening programme treatment; or some combination of clinical All of the 67 who funded In vitro fertilisation and, more importantly, denies older women an factors such as rate of deterioration and social did so on a restricted basis. These restrictions important step towards the early detection of factors reflecting dysfunction in usual activities. varied in severity, with some constructed so as to breast cancer. This is being pursued by New Zealand's Health effectively deny access except under exceptional The inequity of this situation will become Commission4 and in the United Kingdom is circumstances. The restrictions were invariably increasingly difficult to reconcile as women aged being piloted at Salisbury District .5 explicit and were a mixture of medical reasons- 50-64 move through the breast screening Points schemes could help establish thresholds for example, woman's age, ranging from 34 to 42 programme and then ask, at age 65, why they for entry on to elective waiting lists and help years (where age is used as a marker for clinical cannot continue to receive invitations. The standardise general practitioners' referral prac- effectiveness)-and social factors such as marital current system, in which women aged 65 or over tice for routine procedures. status, heterosexuality, and marital stability. In can request a screening every three years, has If rationing in the NHS is inevitable, waiting all cases access was also controlled by consider- proved a resounding failure, with around 97% of list points schemes could help to ensure that ing only couples (not individual) presenting eligible women failing to be screened. patients with similar need, somehow reflected by themselves for referral. Women should be informed that breast their points allocation, could be assured of simi- It is clear that in vitro fertilisation continues to cancers do occur in older women and that it is a lar access to treatment, regardless of clinical spe- be overtly rationed within the NHS. What is of good precaution to report any breast symptoms cialty, geographical location, or general practi- concern is the use of criteria which are not promptly. A recent Gallup survey showed that tioner's status. Points schemes could provide a always related to clinical effectiveness but have a 65% of older women thought that their age redefinition of the principle on which the NHS sociological construct that may discriminate group was at little or no risk from breast cancer; was founded within the constraints of resource against some members ofsociety. It is not clear at 28% thought that there was no risk at all. limitation. present whether such criteria are acceptable to Many older women do not seek advice about RHIANNON TUDOR EDWARDS local populations. They seem to reflect the values breast symptoms, believing they cannot develop Lecturer in health economics of purchaser or provider clinicians, or both. the disease. In many cases, older women mention Department of Public Health, that the screening cut off age at 65 encourages University of Liverpool, LAIN SMrrH this belief. PO Box 147, Senior lecturer in health services research Liverpool L69 3BX Nuffield Institute for Health, In view of all these factors, Age Concern Eng- University of Leeds, land is calling on the government to urgently 1 New B on behalf of the Rationing Agenda Group. The ration- Leeds LS2 9PL address the shortage of radiologists, as any ing agenda in the NHS. BMJ 1996;312:1593-601. AURORA PLOMER further delay in extending the programme would (22 June.) Lecturer 2 Samuel 0. Fundholding practices get preferences. BMJ Law Department, be unacceptable. 1992;305:1497. University of Leeds 3 Gudex C, Williams A, Jourdans M, Mason B, Maynard J, SALLY GREENGROSS Director general O'Flynn P, et al. Prioritising health care. Health Trends 1 of Health. the national survey the 1990;22:103-8. College Report of of funding Age Concern England, and provision of infertility services. London: College of London SW16 4ER 4 National Advisory Committee on Core Health and Disability Health, 1993. Support Services. Core servicefor 1994/95. Wellington, New 2 College of Health. Report of the second national survey of the Zealand: NACCHDSS, 1994. funding andprovision ofinfertdity services. London: College of 1 Rubin G. Wrong comparison quoted for breast screening. 5 Edwards KT. An economic perspective of the Salisbury Health, 1995. BMJ 1996;312:1674. (29 June.) waiting list points scheme. In: Malek N, ed. Settngpriorities in health care. Chichester: Wiley, 1994. Don't confuse personal choice with Selection ofgroup members needs to be collective choice Ratio ofwaist circumference to clarified EDITOR,-MoSt of the problems we encounter height is strong predictor of EDrroR,-I think clinicians would take a little with rationing decisions' arise through collective intra-abdominal fat more notice of the Rationing Agenda Group if it choice in publicly funded systems. In other included more people who actually had to face words, decisions are taken by a collectivity that EDrrOR,-Waist circumference' and the ratio of binds individuals. In a privately financed system waist circumference to height' have been sick people and their families day by day over the proposed as better indicators of the need for years. It is difficult to identify any working clini- individuals are more able to make personal cians in the group.' choices (admittedly constrained by their management of weight than the classic body It would be interesting to know the method of income). Decisions are less likely to be taken by mass index. The relevance of the different proxy others on their behalf. Even in most privately measures is presumed to relate to their selection for members to sit in this group. King's prediction ofintra-abdominal fat, but a compari- Fund Policy Institute is presumably a registered funded systems third party decisions are charity-that is, receives a lot of money from the commonplace, but the distinction between son has not yet been reported. taxpayer-and its function should be more personal choice (making your own rationing Thirty one women and 16 men presenting for democratic. decisions) and collective choice (making deci- routine computed tomography had their weight, sions on behalf of others) is an important one. height, and waist circumference measured; all C J BURNS-COX but three men and three women also had their Consultant physician RAY ROBINSON hip circumference measured. The subjects were Frenchay Healthcare Trust, Director , Institute for Health Policy Studies, aged 18-73, and the range of body mass index BS16 ILE Southampton SO17 1BJ (kg/m2) was 20.148.5. The cross sectional areas of both intra-abdominal and subcutaneous fat 1 New B on behalf of the Rationing Agenda Group. The ration- 1 New B on behalf of the Rationing Agenda Group. The ration- were calculated from images taken immediately ing agenda in the NHS. BMJ 1996;312:1593-601. ing agenda in the NHS. BMJ 1996;312:1593-601. (22 June.) (22 June.) cranial to the iliac crests, as previously described.3 Pearson product-moment correlations for the Purchasing policies for in vitro fertilisation logarithms of selected anthropometric variables vary considerably Breast screening has failed and ratios with estimates of total fat and the two older women fat compartments showed that the ratio of waist EDITOR,-The reports from the College of circumference to height had the highest correla- Health' 2 continue to confirm a considerable ED1TOR,-Gary Rubin, service director of the tion with intra-abdominal fat (r = 0.83, variation between health authorities in their pur- East Sussex Breast Screening Service, said that P<0.001). This was higher than the correlations chasing intentions for in vitro fertilisation. Some invitations to breast screening should not be of waist circumference (r = 0.75), body mass have also included sociomedical criteria for allo- extended to women aged 65 or over until more index (r = 0.69), and the ratio ofwaist to hip cir- cating these scarce resources. It is unclear as to information is available on the success rate ofthe cumference (r = 0.54) with intra-abdominal fat.

BMJ vOLUmE 313 31 AUGUST 1996 559 2 Ashwell MA, Lejeune SRE, McPherson K. Ratio of waist cir- trust will be covered by a registration held by the cumference to height may be a better indicator of need for 250- Meno trust as data user (section 1(5) of the act). When weight management. BMJ 1996;312:377. (10 February.) they hold personal data in connection with their 4 3 Ashwell MA, Cole TJ, Dixon AK. Obesity: new insight into the 200-l oWomen 0 anthropometric classification of fat distribution shown by own research they need to register individually as computed tomography. BMJ 1985;290: 1692-4. data users. When health professionals hold 4 Samaras K, Spector T, Nguyen TV, Baan K, Campbell LV, -~ISO- Kelly PJ. Smoking and oestrogen replacement are personal data for accreditation purposes they E ~0 oO0 associated with lower total and central fat in menopausal must consider who is the data user in respect of 0 100- o o women. Int 7 Obes Relat Metab Disord 1996;20(suppl such data. If neither the employer nor the E.100- 0 4):144. that it controls the con- cc 5 Cox BD, Whichelow MJ, Ashwell MA, Prevost AT. Compari- accrediting body accepts 50- 8°°°° ° son of anthropometric indices as predictors of mortality in tents and use of this personal data individuals 50 British adults. Int Y Obesity Relat Metab Disord have to accept direct personal responsibility. 1996;20(suppl 4):137. Although there is an exemption that applies to II 0.4 0.5 0.6 0.7 0.8 personal data held by an individual for domestic or recreational purposes, this cannot sensibly Waist circumference: height apply to sensitive confidential data held in Data protection, health care, connection with pursuing individual research, or 250- * o for professional accreditation, as the data are and the new European directive held in connection with an individual's profes- E 200- sional development or interests (section 33(1)). 0 0 Data Protection Act applies to discussions Data users must observe the Data Protection t, 150- by email and on the Internet Principles, which are set out in the schedule to C ~~ * the act and set enforceable standards for collect- E ~~0 0 o 100 C EDITOR,-Many doctors who hold log books or ing, holding, and using personal data (schedule do clinical research on computers are probably 1, part 1 to the act). The first principle requires acting illegally if they are not registered under that individuals should be aware, at least in broad the Data Protection Act (Data Protection Regis- terms, ofthe purpose or purposes for which their trar, personal communication).' personal data may be used or disclosed. Further- Hospital registration may cover research more, when health professionals use confidential 60 75 90 105 120 135 performed in a hospital on that hospital's Waist circumference (cm) medical data for a purpose other than the imme- patients or their records but will not cover diate health care of the patient concerned-such research performed at home or a collection of as private research or maintaining a record for Fig 1-Relation between ratio of waist circumference data from more than one hospital, such as a sur- accreditation purposes-this raises the question to height and intra-abdominal fat and between waist. gical log book. The position may deteriorate fur- of whether this entails a breach of confidentiality circumference alone and intra-abdominal fat in 31 ther when manual records are included in unless the patient's consent is obtained. When women and 16 women legislation.2 Anonymising records is only valid if personal data are processed to do something that the patient cannot be identified by any means, entails a breach of a duty of confidentiality there Multiple regression was used to find the best which is seldom the case. This places junior doc- will be a consequent breach of the lawful predictor of intra-abdominal fat for the two sexes tors in a difficult position. They may be required processing requirement of the first principle. after adjustment for covariates. The effects of sex to keep log books and are encouraged to perform Furthermore, when health professionals main- and subcutaneous fat were not significant, so the clinical research. As the law stands, they should tain sensitive records on palmtop computers, sexes were combined (fig 1). Even after therefore register under the Data Protection Act, whose portability makes them easy to lose or adjustment for age (t = 2.9, P = 0.006) and body the first principle of which places them under an steal, this calls into question compliance with the mass index (t = 0.2, P = 0.8) the ratio of waist obligation to state the purpose for which data requirement of the eighth principle to take circumference to height remained by far the best will be used on collection. This is difficult for a "appropriate security." predictor of intra-abdominal fat (t = 4.2, junior doctor acting alone to achieve, and may be Finally, Gerrand is right to counsel caution P = 0.0002) and the intercepts for men and impossible when performing, for example, a ret- when participating in email discussion groups to women were the same (t = 0.1, P>0.9). Waist rospective review of case notes. Indeed, even ensure that any data communicated are either circumference alone was less predictive (t = 3.3, seldom meet the requirements of the anonymised or shared with consent. P = 0.002), and the intercepts were significantly act in this respect, and when they do, might I ask different for men and women (t = 3.0, that they consider the needs of junior doctors? PHILIP J JONES P=0.01). also covers data Assistant registrar The act transfer, including of Data Protection Registrar, We conclude that the ratio of waist circumfer- images such as radiographs, over the Internet. Wycliffe House, ence to height is the best simple anthropometric However, registration may not help. Although Wilmslow, predictor of intra-abdominal fat in men and the intention to transfer data abroad can be reg- Cheshire SK9 5AF women. It is a better predictor than waist istered, the destination must be controlled, and circumference alone because of the correlations the nature of the Internet makes this difficult. between waist circumference, height, and intra- Doctors who participate in email discussion abdominal fat. Others have reached the same groups should exercise caution in discussing Doctors in distress conclusion for women after using dual x ray patients' details or even politically sensitive absorptiometry.4 On the basis of the results issues with anyone, let alone the disparate group Same points were made 120 years ago reported here, our previous cross sectional that makes up a forum on the Internet. There is evidence,2 and new evidence from prospective a parallel with publishing, and the same care EDrrOR,-I congratulate Bob Bury on at last epidemiological studies,' we propose that the should be taken with anonymity and consent. highlighting the positive aspects of a career in ratio of waist circumference to height should be medicine.' These points were summarised 120 CRAIG GERRAND used for the management of weight in a public Registrar in orthopaedics years ago by Lord Lister in his address to gradu- health context so that increased emphasis can be Victoria Infirmary NHS Trust, ates: "If we had nothing but pecuniary rewards put on treating the people with the greatest Glasgow G42 OTT and worldly honours to look to, our profession metabolic risks of obesity. would not be one to be desired. But in its 1 Millman A. Keeping your computer healthy and legal. BMJ MARGARET ASHWELL 1995;311: 1289-93. practice you will find it to be attended with Director 2 Smith MF. Data protection, health care, and the new peculiar privileges; second to none in intense Ashwell Associates, European directive. BMJ 1996;312:197-8. (27 January.) interest and pure pleasures."2 Ashwell, We should remember this rather than continue North Hertfordshire SG7 5PZ to be doctor. Ifwe TIMOTHY J COLE3 to bleat about how awful it is a Senior scientist Data users must observe all principles of do not then there is a real danger that, sooner or MRC Dunn Nutrition Unit, Data Protection Act later, people will start to believe us. Cambridge CB4 1XJ ADRIAN K DIXON EDrrOR,-Craig Gerrand's letter raises important DAVID HEWN Professor of radiology questions. Health professionals who hold log Senior surgical registrar University Department of Radiology, Department of Surgery, Addenbrooke's Hospital, books or do clinical research that entails holding , P0 Box 219, personal data on computer must ensure this is Bristol Royal Infirmary, Cambridge CB2 2QQ covered by an appropriate registration under the Bristol BS2 8HW 1 Han TS, Lean MEJ, Seidell JC Waist circumference remains Data Protection Act 1984. Health professionals 1 Bury B. Doctors in distress. BMJ 1996;312:1235. (11 May.) useful predictor of coronary heart diseasee.BM maintaining computerised records relating to 2 Lister J. Graduation address. Edinburgh Medical journal 1996;312:1227-8. (11 May.) treatment they have provided as employees of a 1876;XXII:280-4.

560 BMJ VOLUME 313 31 AUGUST 1996