Abstract
Contraceptive prevalence estimates are indicators of the performance of family planning programs. Yet, available evidence suggests that national surveys may be underestimating the prevalence of traditional methods. The apparent underestimation of traditional methods stems from current approaches for collecting, analyzing, and reporting contraceptive data. We examined the effect of survey methodological innovations on the estimation of traditional and modern contraceptives. We used data from a cross-country comparative study conducted in the Democratic Republic of Congo (DRC), Ghana, Kenya, and Nigeria. The sample comprised 9,075 in union and sexually active women not in union aged 15–49 years. The results showed that follow-up method-by-method questioning increased the reporting of both traditional and modern methods, with the increase being much higher for modern methods, while reducing the percentage of nonusers. Revising the standard approach for computing contraceptive prevalence to account separately for concurrent traditional and modern method use revealed substantial underestimation of traditional method use, particularly in DRC and Ghana. These findings underscore the need to revise the current framing of questions and estimation approaches to improve the accuracy of contraceptive use estimates. The findings also highlight the importance of taking into account concurrent use of traditional and modern methods, which is often ignored in family planning research.