Condom effectiveness: where are we now?

Abstract

After performing a review of prospective cohort evaluations, a focussed assessment of the current knowledge base and methodology pertaining to condom effectiveness against sexually transmissible infections, including HIV, was also conducted. Key observations included the point that studies of condom effectiveness are inherently complex and the potential forms of study bias all generally favour the null hypothesis. Perhaps the most challenging obstacle to rigor in these studies lies in determining which events of condom-protected sex occurred before infection as opposed to after infection when, in fact, infection occurs. This problem leads to misclassification bias; however, other sources of misclassification bias are common. Greater attention to the selection of a recall period, improved precision of self-reported measures, and accounting for condom use errors and problems are critical steps that must occur to promote rigor in these studies. Despite multiple shortcomings, prospective studies of condom effectiveness provide a reasonably favourable evaluation. Subsequent studies, however, should be designed to greatly reduce the error variance that predisposes condom effectiveness studies to type 2 errors that mask the potential value of condoms.