Norrie disease

Merit of using risk reduction rather than absolute risk for lipid lowering drugs

EDITOR--Our study of whether treatment recommendations for lipid lowering drugs should be based on absolute coronary risk or risk reduction[1] was accompanied by an editorial by Jackson in the same issue that warrants further discussion.

The chance of preventing a coronary event is the absolute risk multiplied by the relative risk reduction, but the question is whether the relative risk reduction is equal in patients of all ages. The meta-analysis of the statin trials by LaRosa et al,[2] cited as evidence for this by Jackson (together with three hypertension trials), does not render Law et al's meta-analysis of lipid lowering trials invalid[3] as LaRosa et al included both primary and secondary prevention trials, assuming that the difference between them relates only to the absolute risk of a further event.

Our study concerned primary prevention, and the two relevant statin trials in this meta-analysis suggest that age may influence risk reduction, although formal statistical analyses were not reported. One of these trials showed a relative risk reduction of 40% (95% confidence interval 16% to 56%) below the age of 55, compared with 27% (8% to 43%) above[4]; the other study showed a 46.5% risk reduction below the median age (age 58) compared with 30.4% above.[5] Both trials were consistent with the age effect predicted by Law et al's meta-analysis.

Our objective was to highlight the potential for leaving young patients with multiple risk factors untreated by assuming that relative risk reduction is not influenced by age. If treatment is based solely on absolute risk a male, non-smoking, diabetic patient with systolic blood pressure of 180 mm Hg and total and high density lipoprotein cholesterol concentrations of 6.0 and 0.9 mmol/l would not reach the risk threshold for treatment until the age of 53. By contrast, after adjustment for age a risk reduction threshold of 4.5% is reached at age 42 when absolute risk is 8.9%.

During this 11 years the patient's average annual risk of coronary heart disease is approximately 2.4%, giving a cumulative event risk of 27.5%, or more than a 1 in 4 chance of an event through the delay in treatment. If this was adjusted for life years gained--as suggested in both our study and Baker et al's study published in the same issue[6]--the benefit of starting treatment at an early age would become even more apparent.

The aim of treatment recommendations should be to maximise use of trial data to increase their relevance. Use of computer based technology enables complex guidelines to be handled easily in a clinical setting and readily updated as new evidence becomes available.

Richard Neary consultant in chemical neurology nearrh@netscape.net

Sud Ramachandran senior registrar in chemical neurology

North Staffordshire Hospital Trust, North Staffordshire Hospital, Stoke-on-Trent ST4 6QG

[1] Ramachandran S, French JM, Vanderpump MPJ, Croft P, Neary RH. Should treatment recommendations for lipid lowering drugs be based on absolute coronary risk or risk reduction? BMJ 2000;320:677-8. (11 March.)

[2] LaRosa JC, He J, Vupputuri S. Effect of statins on risk of coronary disease: meta-analysis of randomized controlled trials. JAMA 1999;282:2340-6.

[3] Law MR, Wald NJ, Thompson SG. By how much and how quickly does reduction in serum cholesterol lower risk of ischaemic heart disease? BMJ 1994;308:367-72.

[4] West of Scotland Coronary Prevention Group. West of Scotland coronary prevention study: identification of high-risk groups and comparison with other cardiovascular intervention trials. Lancet 1996;348:1339-42.

[5] Downs JR, Clearfield M, Weis S, Whitney E, Shapiro DR, Beere PA, et al. Primary prevention of acute coronary events with lovastatin in men and women with average cholesterol levels. Results of AFCAPS/TexCAPS. JAMA 1998;279:1615-22.

[6] Baker S, Priest P, Jackson R. Using thresholds based on risk of cardiovascular disease to target treatment for hypertension: modelling events averted and number treated. BMJ 2000;320:680-5. (11 March.)

Joint British societies recommend their computer program for risk calculations

EDITOR--By including high density lipoprotein cholesterol concentrations the new Sheffield tables are considerably more accurate than the earlier version.[1 2] We disagree, however, with the assessment of the accuracy of risk prediction methods by applying them to the whole population when such methods are intended to identify high risk populations. Inevitably, when a whole-population approach is used most people will have a risk that is substantially below the high threshold risk for which the tables are designed. Accuracy should be tested in people who are closer to this threshold because they are the type of patient for whom the clinical decision about drug treatment is to be made in practice.

We have therefore compared the accuracy of the new Sheffield tables, the joint British societies' charts for coronary risk,[3] and the New Zealand charts for cardiovascular risk[4] with the Framingham risk equation on which they are based. We calculated them for a series of 386 patients referred to our lipid clinic for advice about whether drug treatment was justified. The joint British societies' charts identified correctly 88% of the patients to whom they could be applied, the new Sheffield tables 81%, but the New Zealand charts only 63%.

Although they are therefore similar in terms of accuracy, the new Sheffield tables were not adopted by the joint British societies[3]; they do not allow practitioners to judge the level of risk between 15% and 30%, and these guidelines recommend that, as statins become cheaper and more resources are available, people at lower risk will progressively be targeted for cholesterol lowering treatment. This decision seems to be further justified by Isles et al's finding that nurses and doctors found the new Sheffield tables difficult to use.[5]

Rather than tables or charts, the joint British societies recommended a computer program available from us, from the British Heart Foundation, or oil the British Hypertension Society's website (www.hyp.ac.uk/ bhs).[2] This program provides the risk of both coronary heart disease and stroke for both systolic and diastolic blood pressure, thus allowing a more comprehensive understanding of cardiovascular risk and rational planning of antihypertensive and lipid lowering treatment.

P N Durrington professor of medicine Department of Medicine, Manchester Royal Infirmary, University of Manchester, Manchester M13 9WL pdurrington@hq.cmht.nwest.nhs.uk

[1] Wallis EJ, Ramsay LE, Haq IU, Ghahramani P, Jackson PR, Rowland-Yeo K, et al. Coronary and cardiovascular risk estimation for primary prevention: validation of a new Sheffield table in the 1995 Scottish health survey population. BMJ 2000;320:671-6. (11 March.)

[2] Durrington PN, Prais H, Bhatnagar D, France M, Crowley V, Khan J, et al. Indications for cholesterol-lowering medication: comparison of risk-assessment methods. Lancet 1999;353:278-81.

[3] Wood D, Durrington PN, Poulter N, McInnes G, Rees A, Wray R. Joint British recommendations on prevention of coronary heart disease in clinical practice. Heart 1998;80(suppl 2):S1-29.

[4] Dyslipidaemia Advisory Group on behalf of the scientific committee of the National Heart Foundation of New Zealand. National Heart Foundation guidelines for the assessment and management of dyslipidaemia. N Z Med J 1996;109:224-32.

[5] Isles CG, Ritchie LD, Murchie P, Norrie J. Risk assessment in primary prevention of coronary heart disease: randomised comparison of three scoring methods. BMJ 2000;320:690-1. (11 March.)

Absolute cardiovascular risk is not most appropriate measure to use

EDITOR--The BMJ issue of 11 March has reduction of risk factors for ischaemic heart disease as its theme.[1] In each of the papers reduction in the absolute risk is described as an appropriate goal. It is not. Absolute risk--the chance of dying or developing serious symptomatic ischaemic heart disease for the first time--has a different value at different ages. This is not because people of different ages should have a different value attached to their life (in my view all life is equally valuable); rather it is because it is a measure of death and not of life.

It is precisely because each year of life is equally valuable that death becomes less of a tragedy as one becomes older. A 40 year old who dies of a heart attack may lose 40 years of potential life; this is unlikely to be the case for an 80 year old. This also fits in with the public perception. Even the most hardened undertaker will cry when faced with a child's body to bury or cremate: it is the thought of all the future that might have been.

To use risk prediction charts or computer programs to bring everyone's risk down to the same level will maximise the number of lives saved but not the amount of life saved. It will provide quick results to please NHS planners, and it will divert much NHS funding into selected pharmaceutical companies, but the impact will be on those near the end of their lives anyway. It will have relatively little effect on the epidemic of people dying in middle age from heart disease.

To achieve a proper balance we need to treat the young more, and the old less, than the charts suggest. Remember that a 40 year old with a 30% risk of death over 10 years will have a 70% chance of reaching 50, a 49% chance of reaching 60, and only a 34.3% chance of reaching 70 if no action is taken.

Daniel Albert joint chair, Leeds South Primary Care Group

Fountain Medical Centre, Morley, Leeds LS27 9EN daniel@albert4.demon.co.uk

[1] Risk in cardiovascular disease [theme issue]. BMJ 2000;320 (7236). (11 March.)

Having so many different guidelines about reducing risk is confusing

EDITOR--The 11 March issue of the BMJ provided a wealth of information and advice on how we might best bring some logical order into our efforts to reduce cardiovascular risk in the population.[1] As a reasonably conscientious general practitioner, I read all the relevant papers (some of them twice) and the accompanying editorial, yet I came away feeling that I was floundering around in a muddy present rather than striding out into a brave new evidence based future.

I want a simple chart or computer program that will allow me to assess and reduce the risk of cardiovascular disease in patients who may or may not already be taking hypotensive treatment. I also wish to give patients some idea of the likely benefit they can expect from treatment. If I have understood things correctly the following five statements are true.

* The Sheffield tables allow for hypertensive patients already taking treatment; the joint British societies' and the New Zealand tables do not.

* The joint British societies' and the New Zealand tables are easier to use in practice than the Sheffield tables, but by excluding hypertensive patients already taking treatment they are useless for a large number of patients whose risks I wish to address.

* Only the New Zealand tables include estimates of expected benefit from treatment, in the form of numbers needed to treat and events prevented; but they use five year total cardiovascular rather than 10 year coronary heart disease risk estimates.

* The choice between 30%, 15%, or any other cut-off point for 10 year risk is arbitrary and depends on an as yet unrealised consensus or government diktat.

* None of the tables addresses impact on total mortality or morbidity, and their recommendations are therefore of less interest to the individual patient than to the cardiovascular lobby.

So, until someone can clear the waters for me, I think I'll just continue to muddle along. Despite my apparent lack of enthusiasm for the cardiovascular cutting edge I can be contacted on email.

Dougal Jeffries general practitioner Bemerton Heath Surgery, Salisbury SP2 9DJ dougal.j@virgin.net

[1] Risk in cardiovascular disease [theme issue]. BMJ 2000;320 (7236). (11 March.)

Subclinical hypothyroidism is risk factor for coronary heart disease

EDITOR--The BMJ issue of 11 March is concerned with the risk factors for cardiovascular disease.[1] At a rough count there are 10 items on these, with 33 named contributors and the combined strength of the British Cardiac Society, British Hyperlipidaemia Association, British Hypertension Society, and British Diabetic Association.

It has been obvious to many clinicians that subclinical hypothyroidism is a risk factor for coronary heart disease in women,[2] and now the large Rotterdam population study has confirmed that it is an independent risk factor as important as the other risk factors for coronary heart disease.[3] Nowhere is this fact mentioned despite its obvious importance for prevention.

P B S Fowler consultant endocrinologist 152 Harley Street, London W1NN 1HH

[1] Risk in cardiovascular disease [theme issue]. BMJ 2000;320 (7236). (11 March.)

[2] Fowler PBS, Swale J, Andrews H. Hypercholesterolaemia in borderline hypothyroidism. Stage of pre-myxoedema. Lancet 1970;ii:488-91.

[3] Hak AE, Pols HAP, Visser TJ, Drexhage HA, Hofman A, Witteman JCM. Subclinical hypothyroidism is an independent risk factor for atherosclerosis and myocardial infarction in elderly women: the Rotterdam study. Ann Intern Med 2000; 132:270-8.

COPYRIGHT 2000 British Medical Association
COPYRIGHT 2000 Gale Group