Methodological challenges in research on sexual risk behavior: III. Response to commentary

Background: We recently provided two detailed reviews of sexual risk behavior research (1,2) in which we identified methodological challenges, reviewed new developments, and made recommendations to guide future research. In this issue, Catania et al. provide a commentary on our two reviews. Purpose: Clarifications. Method: We carefully reviewed the commentary provided by Catania et al. and respond to their critique. Results: Catania et al. emphasized three major points in their commentary: (a) When assessing sexual risk behavior, categorical assessments of unprotected intercourse may be preferable to count data; (b) quantification of sexual risk via counts of unprotected intercourse does not reflect the complexity of the sexual transmission of HIV; and (c) analysis of variance and analysis of covariance options in analyzing the effects of randomized HIV risk-reduction trials were overemphasized in our review. We address each of these points in turn and clarify the three points that were misunderstood. Conclusions: We remain convinced that count data of sexual risk behavior are the more useful, versatile, and valid indicators of sexual risk reduction in the evaluation of sexual risk behavior interventions. We urge interested readers to review our original articles, which provide a more detailed account of the challenges and opportunities of sexual behavior research.