Statistical and Epistemological Issues in the Evaluation of Treatment Efficacy of Pharmaceutical, Psychological, and Combination Treatments for Women's Sexual Desire Difficulties

Abstract

We were grateful to receive responses from Leonore Tiefer, Anita Clayton and Robert Pyke, and Richard Balon and Robert Segraves, to our commentary (Brotto et al., 2016) on Pyke and Clayton (2015). These commentaries raise a number of substantive statistical and epistemological issues relating to the evaluation of treatment efficacy in pharmaceutical, psychological, and combination treatments for sexual desire difficulties, and caution researchers to remain mindful of sources of bias as we do the science. In what follows, we discuss each of these issues in turn in hopes of encouraging our field to adopt the highest possible standards when carrying out and interpreting treatment outcome research.

Modeling Female Sexual Desire: An Overview and Commentary

“…Women on the higher dose of 100 mg flibanserin reported an increase of 1.6 events. Despite the statistically significant increase in the 100 mg dose compared to controls, calculations of effect size (d = 0.18 to 0.22) and odds ratios (1.5 to 1.67) were small, indicating minimal effects on sexually satisfying events [40]. The other primary outcome measure, monthly change in "desire score" on a scale of 0-3, did not reveal significant differences between women taking placebo and those taking flibanserin.…”