Measuring Condom Use: Asking “Do You or Don't You” Isn't Enough

Abstract

This article compares cross-sectional measures of condom use among 2,269 female sex workers in Cameroon randomly assigned to receive one of five different questionnaires measuring condom use. We found that the level of reported condom use varied depending on the type of survey questions used. Measures based on 2-week coital logs or the past 10 acts categorized more women as "100%" or "0%" users than always-to-never scales categorized women as "always" or "never" users. Consistency of use also varied by type of partner. Internal consistency of responses was high. Future studies should assess differences in prospective measures of condom use and the level of association between various measures and infection with sexually transmitted disease.