Cumulative Distribution Plots to Enhance Interpretation of Treatment Differences on the Self-Esteem And Relationship Questionnaire for Men with Erectile Dysfunction
Andrew G. Bushmakin, Pfizer Inc 
*Joseph C. Cappelleri, Pfizer Inc 
Martin O. Carlsson, Pfizer Inc 
Tara Symonds, Pfizer Ltd 
Kelly H. Zou, Pfizer Inc 

Keywords: cumulative distribution plots, patient-reported outcomes, interpretation, treatment effect, erectile dysfunction

Objective: To use cumulative distribution plots (CDP) to enrich clinical meaning and enhance the interpretation of scores on the Self-Esteem And Relationship (SEAR) questionnaire, a patient-reported outcome, with respect to treatment differences for men with erectile dysfunction (ED).

Methods: Data were from a randomized, parallel group, multicenter, double-blind, placebo-controlled trial of flexible-dose sildenafil for ED treatment, which randomized and enrolled 256 patients. The SEAR questionnaire was assessed at baseline and at Weeks 8 and 12 (end of study). The SEAR questionnaire consists of 14 items measuring sexual relationship satisfaction, confidence, self-esteem, and overall relationship satisfaction; the questionnaire also includes a total (overall) score. Scores on each of these aspects were transformed onto a 0-to-100 scale, with higher scores being more favorable. The CDP showed a continuous plot of the numeric change from baseline to Week 12 on the x-axis and the percent of patients experiencing that change on the y-axis. Kolmogorov-Smirov (KS) and chi-square tests were conducted.

Results: The CDP showed a clear and distinct profile of psychosocial factors in favor of sildenafil over placebo (all KS p-values = 0.01), with essentially no overlap between sildenafil and placebo treatments. At a given change score, the percentage of subjects at that score or better was higher in the sildenafil group than that in the placebo group. For example, in sildenafil vs. placebo groups, respectively, 71.7% vs. 37.4% of patients improved by at least 10 points on sexual relationship satisfaction; 70.8% vs. 41.4% on confidence; 76.1% vs. 48.7% on self-esteem; 60.2% vs. 44% on overall relationship satisfaction; and 75.2% vs. 38.3% on the overall score. All treatment differences were statistically significant (all chi-square p-values = 0.01).

Conclusions: CDP can enrich clinical interpretation of scores on patient-reported outcomes in general. In particular, on the SEAR questionnaire, the CDP clearly, distinctly, and meaningfully highlighted the favorable profiles of responses with sildenafil, compared against placebo, with respect to psychosocial factors.