Analysis of the underestimation of induced abortions in a survey of the general population in France

A pilot study has been conducted in France to estimate the extent of underreporting of induced abortion and the factors linked to underreporting.A representative random sample of 300 women aged between 18 and 44 years of age who had been pregnant during the previous 2 years and 100 women who had not was selected from the telephone directory. Interviews were conducted by telephone.The annual incidence of induced abortion for the 18-44-year-old age group was 6.9 per 1000 (95% confidence interval 2.2-11.6), while the rate derived from national statistics was 15 per 1000. Five women who reported a therapeutic abortion in their lifetime had confused induced abortion and therapeutic abortion. There was no significant difference in the proportion of women reporting induced abortion in their lifetime between those answering with another adult present and those who were alone when answering (13.1% versus 11.1%, p = 0.8). There was also no significant difference according to the sex of the interviewer (10.0% for men versus 12.2% for women, p = 0.6).The results suggest that the context in which the questions were asked does not affect the validity of the answers and that there is a problem in the comprehension of the terminology used. Extensive rewriting and reformulation of the questions is required to minimize the underreporting of induced abortion.