Bone densitometry Depot medroxyprogesterone and bone ... - NCBI

3 downloads 123 Views 618KB Size Report
only to Ward's triangle; for other regions of the femur ... the total hip, and 2-42% for Ward's coefficients of .... EDITOR,-John Gabbay and Andrew Stevens' have.
Bone densitometry Is a valuable investigation EDITOR,-Susan M Ott draws attention to some of the potential pitfalls associated with the use of bone densitometry in clinical practice.' Though I acknowledge that caution is needed in interpreting measurements of bone density, their value should not be understated. In particular, distinction should be made between the assessment of risk in an individual person and that in population screening. Because of the gaussian distribution of bone density and the gradient of the risk of fracture over its whole range, the proportional contribution to the total incidence of fracture of those with the lowest bone density is small; a major impact on the incidence can be achieved only by increasing bone density across the whole population. In clinical practice, however, the risk in an individual person is an important factor in decisions about treatment, based on analysis of the risk of disease on the one hand and the ratio of risk to benefit associated with treatment on the other. The present lack of strict densitometric criteria for treatment does not invalidate this approach, particularly for those at high risk. The figure of around 6% quoted for the precision of bone densitometry is misleading since it applies only to Ward's triangle; for other regions of the femur, such as the femoral neck and trochanter, precision of 3% or less can be obtained with careful positioning and quality control. In the spine, precision is even better, values of around 1% being reported in young and middle aged subjects.2 This compares favourably with reported annual rates of bone loss at this site in menopausal women of 1-5%/ and indicates that the ability of serial measurements to show effects of treatment in menopausal women is better than that implied by Ott. Finally, the finding in prospective studies of a relation between areal bone density and risk of fracture4 justifies the clinical use of techniques that measure areal density. There is no evidence that volumetric density measurements are superior in predicting the risk of fracture, and the ability of areal bone density to capture an element of skeletal size may indeed be an advantage since bone size, independent of density, probably contributes to bone strength and hence risk of fracture. Bone densitometry represents a real advance in the diagnosis and clinical management of patients with osteoporosis; the cautions expressed by Ott should not be interpreted as indicating otherwise. J E COMPSTON Consultant physician

Department of Medicine, Addenbrooke's Hospital,

Cambridge CB2 2QQ 1 Ott SM. Bone mass measurements: reasons to be cautious. BMJ

1994;308:931-2. (9 April.) 2 Laskey MA, Flaxman ME, Barber RW, Traffoed S, Hayball MP, Lyttle KD, et al. Comparative performance in vitro and in vivo

of Lunar DPX and Hologic GDR-1000 dual energy X-ray absorptiometers. BrJ Radiol 1991;64:1023-9. 3 CompstonJE. Osteoporosis. Clin Endocrinol 1990;33:653-82. 4 Black DM, Cummings SR, Genant HK, Nevitt MC, Palermo L, Browner W. Axial and appendicular bone density predict fractures in older women.jBone MinerRes 1992;7:633-8.

Precision error is low EDITOR,-We believe that Susan M Ott's statement that "a walk around the room causes the measurement [of bone mass] to change by up to 6% (at the hip)" is misleading and needs to be put into perspective.' The precision error of bone densitometers is usually expressed as the percentage coefficient of variation. We determined the coefficient of variation of duplicate measurements of bone mass at the hip in 19 subjects, removing the subjects from the bone densitometer between the scans. The subjects had bone mineral densities

BMJ VOLUME 308

11 JUNE 1994

ranging from normal to severely osteopenic. The coefficient of variation for our dual energy absorptiometer (Hologic QDR-2000) is 1-39% for the femoral neck, 0-92% for the trochanteric region, 0-98% for the intertrochanteric region, 0 94% for the total hip, and 2-42% for Ward's coefficients of variations reported by other investigators.23 The change of 6% in bone mass mentioned in the editorial is equivalent to a coefficient of variation of 4-12%, which is much higher than the precision error found at our institution or in earlier reports. From our data we predict that a 6% error in measurements of bone mass of the total hip or at any region of the hip would occur less than once in 1000 times. Thus a 6% change in bone mass or repeat measurement is unlikely to occur when a dual photon x ray absorptiometer is used. Any short term change in bone mass of this magnitude suggests either malfunction of equipment or defficient technique in obtaining reproducible measurements of bone mass. These opinions and assertions are the private views of the authors and are not to be construed as reflecting the views of the Department of the Army or the Department of Defense. WIT IAM E DUNCAN

Assistant chief

precision of 1-5% or better for the femoral neck.' Assessing changes in areal bone density in children is difficult, but Ott's conclusions are ambiguous. Bone area increases with growth, as does bone mineral content, but if bone area outstrips growth, bone mineral density will apparently decrease; this argues for care in interpretation. Ott's editorial could be used by purchasers as an excuse for not supporting the development of bone density services whereas it should be used to ensure that services are purchased only from providers who run services with adequate quality control, have clinical experience in interpreting results, and offer additional clinical follow up of difficult cases. The indications for assessment of bone mineral density have been fully examined and referenced in a recent publication for purchasers produced by the scientific coordinating committee of the National Osteoporosis Society.3 DAVID M REID

Consultant rheumatologist

City Hospital, Aberdeen AB9 8AU CYRUS COOPER

Clinical scientist

MRC Environmental Epidemiology Unit, University of Southampton, Southampton S09 5NH IGNAC FOGELMAN

DAVID M STOT

Medical investment technician KENNETH D BURMAN

Chief

Endocrine-Metabolic Service and Kyle Metabolic Unit, Department of Medicine, Walter Reed Army Medical Center, Washington, DC 20307-5001, USA

Consultant physician

Department of Nuclear Medicine, Guy's Hospital, London SE1 9RT RICHARD EASTELL

AUDREY S CHANGE Statistician

Department of Clinical Investigation, Walter Reed Army Medical Center 1 Ott SM. Bone mass measurements: reasons to be cautious. BM7

Senior lecturer Department of Human Metabolism and Clinical Biochemistry, Clinical Sciences Centre, Northem General Hospital, Sheffield S5 7AU ANTHONY WOOLF Consultant rheumatologist Royal Comwall Hospital, Truro TRl 3LJ

1994;308:931-2. (9 April.) 2 Chapuy MC, Arlot ME, Duboeuf F, Brun J, Crouzet B, Arnaud S, et al. Vitamin D3 and calcium to prevent hip fractures in elderly women. NEngl7Med 1992;327:1637-42. 3 Murphy S, Khaw K-T, May H, Compston JE. Milk consumption and bone mineral density in middle aged and elderly women. BMJ 1994;308:939-41. (9 April.)

Can predict those at high risk of fractures EDrrOR,-Many of the conclusions in Susan M Ott's editorial on measurements of bone mass are misleading if taken at face value.' The subheading -"Bone mass and strength not necessarily correlated"-suggests that some studies have shown a lack of correlation between bone strength and bone density. Bone mass and bone strength in vitro are strongly correlated,2 but this does not undermine the caution needed in interpreting changes in bone mineral density as indicative of changes in bone strength. Interpretaion of a single measurement of bone mineral density relies on the concept of relative risk, where by some people who subsequently sustain a fracture have a normal bone density at baseline. The same is true for the measurement of blood pressure: most sustaining a stroke have a normal blood pressure. Bone mineral density, however, is as good a risk factor for predicting fracture as blood pressure is for predicting stroke.' A recent study of over 9000 women followed up for a mean of 1 -6 years shows that the risk of sustaining a hip fracture increases by up to 2-9-fold for each standard deviation decrease in bone mineral density4-rather more than the 1-5-fold to twofold stated by Ott. Changes in bone mineral density in a particular patient with time need to be interpreted with caution, especially when measurements are made at the hip. Apparently dramatic changes can be taken as indicating improvement or dramatic bone loss but may simply be due to the precision of the measurement and poor repositioning technique. Dual energy x ray absorptiometry scanners have

1 Ott SM. Bone mass measurements: reasons to be cautious. BMJ 1994;308:931-2. (9 April.) 2 Mazess RB, Pedersen P, Vetter J, Barden HS. Bone densitometry of excised vertebrae: anatomical relationships. Cakif Tissue Int 199 1;48:380-6. 3 National Osteoporosis Society. Priorities for prevention. Bath: NOS, 1994. 4 Cummins SR, Black DM, Browner W, Cawley J, Ensrud K, et aL Bone density at various sites for prediction of hip fracture: the study of osteoporotic fractures. Lancet 1993;341:72-5. 5 Jergas M, Genant HJK. Current methods and recent advances in the diagnosis of osteoporosis. Arthritis Rheum 1993;36: 1649-62.

Depot medroxyprogesterone and bone density EDrrOR,-It is gratifying when publications arouse a vigorous debate, but the comments of correspondents on our paper' are disappointing. Either our writing is not clear or the correspondents have not read the paper properly. Anne Szarewiski and colleagues say that we described the theoretical risk of oestrogen deficiency and osteoporosis in long term users of depot medroxyprogesterone.' Our study in fact showed that spinal bone density increased in women who stopped using depot medroxyprogesterone and became oestrogen replete, whereas it did not either in women who continued to use this contraceptive or in women who had never used it. We deduced this from sequential measurements of bone density. To us this ranks as a prospective study, not a cross sectional one. We are also taken to task for not having given details of plasma oestradiol concentrations. We reported the values in women using depot medroxyprogesterone in our previous study.3 In the recent study 12 of the 14 women who stopped taking depot medroxyprogesterone resumed menstruation, and we took this as evidence of their being oestrogen replete. As described, we measured oestradiol concentrations in the women who remained amenorrhoeic.

1567

J B Sharma and colleagues raise the question of whether smoking could account for the lower bone density in women using depot medroxyprogesterone.4 We agree that smoking is a potential confounder and discussed this in our earlier publication.3 We are puzzled by Sharma and colleagues' reference to the two groups in our recent study. There were three groups. The control subjects certainly smoked less than the women who used depot medroxyprogesterone, but the proportion of women smoking was the same in those who continued to use depot medroxyprogesterone and those who stopped using it. Bone density increased only in those who stopped using the contraceptive. Finally, we note Sharma and colleagues' (unreferenced) assertion that measurements of the bone density of the femoral neck and wrist by dual energy x ray absorptiometry might be more accurate and reliable than measurements of the bone density of the spine. This is contrary to our experience and to most published data. We believe that the reason we saw greater changes in the spine than the hip is that the spine contains a larger proportion of trabecular bone and is more sensitive to oestrogen. TIM CUNDY Senior lecturer in medicine IAN REID Associate professor

Department of Medicine, University of Auckland Medical School, Auckland 1, New Zealand 1 Cundy T, Cornish J, Evans MC, Roberts H, Reid IR. Recovery of bone density in women who stop using medroxyprogesterone acetate. BMJ 1994;308:247-8. (22 January.) 2 Szarewski A, Hollingworth B, Guillebaud J. Depot medroxyprogesterone and osteoporosis. BMJ 1994;308:717. (12 March.) 3 Cundy T, Evans M, Roberts H, Wattie D, Ames R, Reid IR. Bone density in women receiving depot medroxyprogesterone acetate for contraception. BMJ 1991;303:13-6 (correction 303:202). 4 Sharma JB, Newman MRB, Smith RJ. Depot medroxyprogesterone and osteoporosis. BMJ 1994;308:717. (12 March.)

Towards investing in health gain Evaluation of interventions is needed EDITOR,-John Gabbay and Andrew Stevens' have misrepresented the process through which the Welsh Office's protocols for health gain were written. As a member of two of the panels, I disagree that informed consensus was actively "substituted for scientific proof." The cancers protocol2 was the first one attempted; a technical document was commissioned from Professor Jocelyn Chamberlain to give a starting point; this was sought from outside the principality to ensure an impartial view. For the pain, discomfort, and palliative care protocol,3 the group members wrote a series of technical documents that were critically discussed by the group and externally reviewed by Dr H McQuay (Oxford) and Dr A Diamond (Bristol). The imposition of deadlines is always required to make a group function, but they did not inhibit consultation or discussion. Unfortunately, very few medical outcomes have been subjected to rigorous scientific scrutiny. If the groups had waited to include only research data, they might have waited for ever. However, the protocols have acted as a stimulus to research into needs assessment in an attempt to gain firm data. Such a project on palliative care is under way, but will take at least a further year to complete. The Welsh targets aim at improving mortality and morbidity. The consultation was wide and in reaching the targets the views of the consumers were actively sought. Quality of life is a concern of patient groups and the "patient centred" approach has been welcomed by the Expert Advisory Group on Cancer at the Department of Health. It is easy for those who failed to have the imagination to initiate such an exercise elsewhere

1568

to criticise a "first," which these documents are. They provide a stimulus to improve the quality in the care for patients, not "demands from the centre." Changes are already evident in the approach to cancer patients and to palliative care to increase scientific evaluation of interventions; the importance of this for patients should not be underestimated. ILORA G FINLAY Consultant in palliative medicine Holme Tower Marie Curie Centre, Penarth, South Glamorgan CF64 3YR 1 Gabbay J, Stevens A. Towards investing in health gain. BMJ 1994;308: 117-8. (30 April.) 2 Welsh Health Planning Forum. Protocol for investment in health gain: cancers. Cardiff: WHPF, 1990. 3 Welsh Health Planning Forum. Protocol for investment in health gain: pain, discotnfort andpalliative care. Cardiff: WHPF, 1992.

patients) working towards a consensus document, which was then presented as a menu of targets for local purchasers to choose their own priorities for local strategies for health. In chaotically diverse European provision for health services for people with learning disabilities, the document has been warmly received for what it is: a protocol based on sound evidence from professionals and consumers to help decision making at national and local levels. W I FRASER Professor of leaming disabilities University of Wales College of Medicine, Ely Hospital, Cardiff CF5 5XE 1 Gabbay J, Stevens A. Towards investing in health gain. 1994; 308:1117-8. (30 April.) 2 Welsh Health Planning Forum. Protocol for investment in health gain, mental handicap (learning disabilities). Cardiff, WHPF, 1992.

Documents should be updated regularly EDITOR,-We support the call from John Gabbay and Andrew Stevens for a consolidation of the scientific base as an essential prerequisite for an evolving purchasing plan.' We have recently written a technical document for East Anglia: Coronary Heart Disease-Opportunities for Prevention and Treatment. This document is aimed at purchasers, most particularly directors of public health, and reviews the recent evidence on effective interventions while illustrating the subject in the context of local epidemiology. In preparing this document we came across several similar documents, none of which we felt able to cite as they neither clearly specified their aim as informing purchasing based on evidence nor included recent research findings. We firmly believe that these documents have great value. Their value will, however be sustained only if they are kept readable and up to date. Such documents should be produced at a national level and updated annually. The new arrangements for public health within the regional office structure as part of the NHS Executive could ensure the coordination of such documents. These publications, however well received today, will be left to gather dust unless the need for regular view and updating is recognised as a central responsibility. A R NESS

Registrar in public health medicine (Anglia) CELIA DUFF Consultant in public health medicine (Anglia) Anglia and Oxford Regional Health Authority, Chesterton, Cambridge CB4 I RF 1 Gabbay J, Stevens A. Towards investing in health gain. 1994;308: 117-8. (30 April.)

BM_

Supporting protocols exist EDITOR,-John Gabbay and Andrew Stevens's praise for the Welsh Health Planning Forum's series of protocols for investment in health gain is qualified by their criticism of the comparative lack of evidence to support many of the recommendations.' Accepting that, for many medical interventions, the evidence simply does not exist, as chairman of one protocol panel2 I can confirm that a 500 page technical document, copiously referenced, supported the protocol. The technical document had two external referees; the authors separately ranked the strength of the evidence to provide for learning disabilities interventions-a "shopper's guide" from 1, "substantiated as increasing health gain," to 4, "should be abandoned in light of available evidence." Gabbay and Stevens also consider the protocol exercise as "top down" target setting. This was not my experience. The process consisted of consumers (voluntary organisations, parents,

New public health Research is part ofthe political process ED1TOR,-Jan P Vandenbroucke, in commenting on the new public health, acknowledges the perennial relevance of equity, ecology, and the environment but advises that the energy behind the rhetoric be directed in "good ways" to avoid "academic downfall."' His point is well made, but he says nothing of the opportunities and progress that are being made. We recently addressed these issues at a conference in Liverpool (Health in cities: research and change in urban community health) drawing on experience of the WHO Healthy Cities Project2 and Health for All initiatives to address two key questions: (1) what research methods are appropriate for assessing needs and evaluating the Health for All process and outcomes, and (2) how can the research findings be translated into policy, and what are the barriers? The conference heard that, if research is to be made more relevant to policy makers, there needs to be greater recognition that research is part of the political process, more integration of qualitative and quantitative research, and selection of research topics more in line with current priorities. The research community also needs to be aware of how policy makers view research; often it is seen as fine in principle, but time consuming and threatening in practice, and always subject to the receptiveness of the political environment. The growing experience of community participation in needs assessment was reported through varied case studies, using focus groups, surveys, geographical information systems, and artistic expression. A theatre performance showed how powerfully drama can be used to research children's experience of relationships, drugs, and family and to present this in a way that teachers and children can work with. Community participation and variety in methods give breadth, depth, and policy relevance. The focus on implementation included primary care, health promotion, housing, and transport. It drew out the opportunities presented by community participation and intersectoral collaboration, but also the practical difficulties. These processes-central policy3 4-require commitment, resources, and experience. The conference heard much about that experience, and about the first steps towards evaluation-for instance the Health Education Authority/Wessex multisectoral collaboration for health project. This is an important stage for the "new public health," as all sectors of society get to grips with implementing Health for All ideas. There is ample substance for academic endeavour, although conceptual and practical challenges in developing methods for needs assessment and evaluation,

BMJ VOLUME 308

11 JUNE 1994