It is not easy to perform informative prospective studies of contraceptive failure. This paper refers to the most recent analysis of data from the International Active Surveillance Study of women taking oral contraceptives, which is a comprehensive 6-monthly follow-up of up to 5 years of pill users in multiple European countries and the USA.

The current analysis refers to a USA population of 52 218 women through 73 269 woman-years of use. In this population there were 1634 unintended pregnancies of which 229 (14%) had no apparent failures in compliance, 46.3% simply forgot their pills, 21.1% used antibiotics and 9.5% had diarrhoea or vomiting.

As expected, the greatest predictor of failure was age, with women aged 20–24 years having three times the failure rate (3 per 100 woman-years) than 35–39-year-olds. Women less than 20 years also had a slightly reduced failure rate. Some 23% of these North American women were obese, and although this did affect failure, the effect was small with an increase in body mass index (BMI) from 20–25 to 30–35 associated with a rise in failures from 2 to only 2.5 per 100 woman-years. Smoking, first year of use, higher parity and lower educational achievement also increased risk of failure.

Different contraceptive regimes also led to different failure rates. As expected, those regimes with a shortened pill-free interval (4 rather than 7 days) led to fewer failures. There was also benefit from preparations containing drospirenone, and the authors attributed this to the longer half-life (30 hours) of this progestogen effectively reducing the hormone-free interval still further. It is also interesting to note that, in line with other studies, failures in the USA were four times as common as failures in the European arm of the study.