Incidence of induced abortion determined by the randomised response technique.

OBJECTIVE: To estimate the incidence of induced abortion using the randomised response technique, and to compare the results with estimates obtained by direct questioning. DESIGN: Cross-sectional study. SETTING: Medical Officer of Health area, Homagama. SUBJECTS: Married women in the reproductive age group. RESULTS: The annual incidence of induced abortion for the three year period 1988 to 1990 was 15.7 per 1000 married women of the reproductive age group as estimated by the randomised response technique. Direct questioning led to 30.5% under-reporting of induced abortion. The application of the method failed only in 8.1% of the population studied. CONCLUSIONS: The randomised response technique is a feasible method to use in our population and provides the best estimates of induced abortion.