Rapid Responses to:

PAPERS: Nicholas Dunn, Margaret Thorogood, Brian Faragher, Linda de Caestecker, Thomas M MacDonald, Charles McCollum, Simon Thomas, Ronald Mann, and Øjvind Lidegaard Oral contraceptives and myocardial infarction: results of the MICA case-control study • Commentary: Oral contraceptives and myocardial infarction: reassuring new findings BMJ 1999; 318: 1579-1584 [Abstract] [Full text]

Rapid Responses: Submit a response to this article

Rapid Responses published:

Large selection bias likely

Rosemary Kirkman   (23 June 1999)

MICA paper on oral contraceptives

Paul O'Brien   (12 July 1999)

Re: Large selection bias likely

Nicholas Dunn   (17 July 1999)

Large selection bias likely 23 June 1999
Rosemary Kirkman Send response to journal: Re: Large selection bias likely	Sir The finding by Dunn and co-workers1 of no significantly increased risk of myocardial infarction in users of oral contraceptives should be put into the context that a large selection bias will have occurred at the time when the prescribing doctor assessed whether it was appropriate to prescribe the combined oral contraceptive to each woman requesting it. I think the results give a real life example of research findings having been widely adopted by British GPs and thereby producing a detectable reduction in clinical risk of a medication. There has been a strong emphasis against use of combined oral contraceptives by older women who smoke ever since the initial Royal College of General Practitioners findings of an increased relative risk of myocardial infarction by age and smoking in combined oral contraceptive users2. The study by Dunn took cases of myocardial infarction between October 1993 and October 1995 when clinical practice had already changed in response to wide publicity of these adverse findings, both from the 1981 paper and from a re-analysis in 19893 which showed that excess risk with oral contraceptive use was concentrated in the smokers. Their results should not be taken as contradiction of the RCGP study findings, but to reinforce the message that the pill does not give heart disease as long as women are selected according to the presently established criteria of excluding the older women who smoke. Dr Rosemary Kirkman Palatine Centre Manchester References 1 Dunn N, Thorogood M, Faragher B, de Caestecker L, MacDonald TM, McCollum C et al. Oral contraceptives and myocardial infarction: results of the MICA case-control study. BMJ 1999;318:1579-84 2 Royal College of General Practitioners' Oral Contraception Study. 1981 3 Croft P, Hannaford PC. Risk factors for acute myocardial infarction in women: evidence from the Royal College of General Practitioners' Oral Contraceptive Study. Br Med J 1989;298:165-8
MICA paper on oral contraceptives 12 July 1999
Paul O'Brien Send response to journal: Re: MICA paper on oral contraceptives	Dear Editor Dr Lidegaard is mistaken in his commentary on the MICA study.(1) Third generation oral contraceptives were, in fact, preferentially prescribed to women at lower risk of myocardial infarction. The odds ratios of third versus second generation pills increased rather than decreased when more potential confounders were entered into the logistic model.(2) Similar preferential prescribing in favour of third generation contraceptives was found in the only empirical study of actual prescribing in the UK.(3) The MICA study should now lay to rest the lipid hypothesis of oral contraceptives and myocardial infarction which was influential in the massive shift in prescribing to third generation pills (which have less effect on blood lipids) in the early 1990's. We have known for some time from angiography studies following myocardial infarction that oral contraceptive users had very little atherosclerosis, which is consistent with repeated epidemiological findings that there was no effect of past use of oral contraceptives on infarction risk. These should preclude a lipid effect. The other main study comparing second and third generation pills, the Transnational study, also found a tendency towards lower risk with second generation oral contraceptives in the UK.(4) A higher risk with second generation pills was only found in Continental Europe. However, 55% of cases there did not have their blood pressure measured prior to starting the pill (17% in UK), and blood pressure was measured less frequently in users of second generation contraceptives (Lewis, personal communication). The apparent increased risk with second generation pills was probably due to inadequate screening for hypertension in older women on these pills. The WHO study found no difference in risk between the generations after controlling for blood pressure measurement. Dr Lidegaard and others, who have argued that for most women the balance in terms of cardiovascular safety lies with third generation contraceptives,(5) a view widely promoted in the last few years, now need to reassess this advice in the light of the MICA study. There appears to be no reduction in risk of myocardial infraction with third generation contraceptives to compensate for the increased risk of venous thromboembolism, a risk recently restated by the UK Medicines Control Agency. The remaining question about third generation oral contraceptives is whether there is any benefit from these pills with lower androgenicity on the 'minor' but important symptoms of acne or break-through bleeding. We await a Cochrane Collaboration review that should provide the answers. Potential conflict of interest: I have provided expert advice to the legal team representing women who developed a venous thrombosis on third generation oral contraceptives. I have been paid for this work. Dr Paul O'Brien Services for Women Parkside Health NHS Trust St Charles Hospital Exmoor St London W10 6DZ 1 Lidegaard O. Oral contraceptives and myocardial infarction: reassuring new findings. BMJ 1999;318 (7198):1583-1584. 2 Dunn N, Thorogood M, Faragher B, de Caestecker L, MacDonald TM, McCollum C, et al. Oral contraceptives and myocardial infarction: results of the MICA case- control study. BMJ 1999;318 (7198):1579-1584. 3 Dunn N, White I, Freemantle S, Mann R. The role of prescribing and referral bias in studies of the association between third generation oral contraceptives and increased risk of thromboembolism. Pharmacoepidemiology & Drug Safety 1998;7(1):3-14. 4 Lewis MA, Heinemann LA, Spitzer WO, MacRae KD, Bruppacher R. The use of oral contraceptives and the occurrence of acute myocardial infarction in young women. Results from the Transnational Study on Oral Contraceptives and the Health of Young Women. Contraception 1997;56(3):129-40. 5 Lidegaard O, Bygdeman M, Milsom I, Nesheim BI, Skjeldestad FE, Toivonen J. Oral contraceptives and thrombosis. From risk estimates to health impact. Acta Obstet Gynecol Scand 1999;78(2):142-9.
Re: Large selection bias likely 17 July 1999
Nicholas Dunn, Senior Research Fellow Drug Safety Research Unit Send response to journal: Re: Re: Large selection bias likely	Sir, Dr Kirkman has misunderstood the results from the MICA case-control study. Although it may well be true that doctors are cautious about prescribing combined oral contraceptives (COC) to women at risk of myocardial infarction (especially smokers), there were still some cases in our study who were users of COC and smoked (n=26, 15 of whom smoked 20+ cigarettes per day). We specifically analysed our data to see whether there was an interaction between COC use and smoking, and could find none. The details of these data are currently in press. The implication of this finding is that it does not seem to matter about COC consumption, if a woman is a smoker. It is the smoking that does the harm regarding risk of myocardial infarction, and this is the message that doctors should pass on to their patients. Dr Nicholas Dunn.