Oral contraception was not associated with venous thromboembolic disease in recent studyBMJ 1998; 316 doi: https://doi.org/10.1136/bmj.316.7137.1090 (Published 04 April 1998) Cite this as: BMJ 1998;316:1090
- R D T Farmer, Professor,
- J-C Todd, Research assistant,
- K D MacRae, Reader,
- T J Williams, Research assistant,
- M A Lewis, Director
- Department of Public Health, Imperial College School of Medicine (University of London), London SW10 9NH
- Epidemiology Pharmacology and Systems Research, Berlin, Germany
EDITOR—Three studies published in 1995-6 reported odds ratios for venous thromboembolic disease among women using third generation oral contraceptives compared with women using second generation products that were in the range 1.5 to 1.8.1-3 The design limitations and interpretation of the results of these studies have been widely questioned. A further study, based on computer records from 147 general practitioners in Britain, failed to show any significant difference in the risk of venous thromboembolic disease between generations of oral contraceptive.4
To investigate the issue further we analysed data from the German MediPlus database (October 1992 to September 1995), which is similar in principle to the British MediPlus database, and the General Practice Research Database. Cases were 42 women aged 18-49 with venous thromboembolic disease who were exposed to an oral contraceptive at the time and were treated with an anticoagulant. We randomly selected four controls per case (168), matched by year of birth and exposure to an oral contraceptive on the event day.
For each woman we extracted the history of use of oral contraceptives, recent medical history, and consultations for physical or psychotherapeutic complaints between January 1992 and the event day. Records of body mass index (or weight), blood pressure, or smoking habit were insufficient to warrant the inclusion of these variables as possible confounders.
More women used second than third generation pills, and none used progestogen only pills (table). We found no significant difference between cases and controls in the type of oral contraceptive used on the event day (unadjusted odds ratio for users of third compared with second generation pills was 0.77 (95% confidence interval 0.38 to 1.57)); this is consistent with the results of the British MediPlus study.4
In both the cases and controls there was no significant difference in age between users of second and third generation pills. There was also no significant difference in the number of prescriptions for oral contraceptives (means 0.29 and 0.27), consultations for psychotherapeutic complaints alone (means 0.29 and 0.27), or mixed psychotherapeutic-physical consultations (means 0.26 and 0.11). The cases did, however, have significantly more consultations than the controls for purely physical reasons before the event (means 13.57 and 5.44; P<0.0001). This difference might indicate poorer health status or prodromal symptoms in the cases or unequal diagnostic surveillance of cases compared with controls, with diagnoses being more likely in frequent attenders. This area, although requiring further investigation, was not related to the type of progestogen in the oral contraceptive used.