Estimating the annual abortion rate in Kerman, Iran: comparison of direct, network scale-up, and single sample count methods

Background: Abortion is a sensitive issue surrounded by social, cultural and religious stigmata. Therefore, estimation of its prevalence involves methodological challenges. The aim of this manuscript is to estimate the abortion prevalence, stratified by type, using a direct and two indirect methods. Materials and Methods: This cross-sectional study was done in 2016, we recruited 1020 women aging 18-49 years. Three methods were applied to estimate the abortion prevalence: direct question, network scale-up (NSU), and single sample count (SSC). In the direct method, to guarantee anonymity, data were collected by means of a self-administered questionnaire. In other methods, data were collected through gender-matched street-based inter-views. Results: The annual rate of abortion estimated by direct and NSU methods were 2 respectively 9 (10 intentional, 4 therapeutic and 15 spontaneous) and 23 (9 intentional, 3 therapeutic, and 11 spontaneous) per 1000 women aging 18-49 years. The annual rate of intentional abortion estimated based on SSC method was higher (15 per 1000) women than other methods. Conclusion: The present estimates are higher than previously reported ones. The results of three methods more or less supported each other confirming the internal validity of our estimates.