The 1995 pill scare revisited: anatomy of a non-epidemic

Abstract

Two years after the October 1995 pill scare that received worldwide attention, this synthesis of evidence goes back to the earliest research on risks of first generation oral contraceptives (OCs). It also covers epidemiological data published since, emphasising the 1995-1996 findings. Late breaking data are also examined. The key issue: are there differences in the risk profiles of second and third generation OCs. The ultimate question is: did any epidemics of venous thromboembolism (VTE) occur? This synthesis of evidence leads to the following conclusions and observations: (i) all OCs on the market are becoming progressively safer; (ii) relative risks of about 2 for VTE, even if real, are clinically unimportant and of no public health significance; (iii) the weak odds ratios contrasting third and second generation OCs, ranging from 1.5 to 2.3 in the 1995-1996 studies are more likely explained by bias than by a causal relationship; (iv) incidences of VTE among users of any OC have been declining over the past three decades; (v) absolute rates of VTE for third generation OC users reported in 1995-1996 are lower than those for users of second generation OCs in 1988 and 1991; (vi) there is no difference in risk of VTE between first starters on second generation OCs versus first starters on third generation OCs; (vii) users of third generation OCs are at much lower risk of acute myocardial infarction than users of second generation OCs; (viii) among users of any OC, the occurrence rates of stroke are low, they are declining, and no differences between second and third generation OCs are apparent; (ix) 2 years after the pill scare there are no epidemics of VTE; (x) there have been excessive rates of therapeutic abortions in some countries; and (xi) the benefit-risk ratio is favourable for users of any OC.

PIP: Presented is a synthesis of the research data on the association between oral contraceptive (OC) use and the risk of venous thromboembolism (VTE), including epidemiologic studies published after the 1995 "pill scare" in the UK. The review suggests 11 general conclusions and observations: 1) all OCs on the market are becoming progressively safer with time as a result of dosage changes; 2) even if OCs are associated with a VTE risk as high as 2.0, this risk is of no clinical or public health significance due to the rarity of VTE; 3) the low odds ratios (1.5-2.3) for third compared with second-generation OCs found in the most recent studies are likely a result of selective referral bias; 4) the incidence of VTE among OC users has been steadily declining over the past three decades; 5) absolute rates of VTE for third-generation OC users in 1995-96 are lower than those for second-generation OCs; 6) there is no difference in VTE risk between first starters on second-generation OCs compared with first starters on third-generation formulations; 7) users of third-generation OCs are at a substantially reduced risk of acute myocardial infarction compared with those of second-generation OCs; 8) occurrence rates of stroke are low among users of both second and third-generation OCs and continue to decline; 9) 2 years after the "pill scare," no epidemics of VTE have occurred; 10) excessive rates of therapeutic abortion have been recorded in some countries as a result of anxieties induced by mass media coverage; and 11) the risk-benefit ratio is favorable for users of any OC.