Study design and outcomes measures: The influence of composite endpoints and other design variables on outcomes in a study of a novel device for stress urinary incontinence

• Published in: Neurourology and urodynamics Vol. 37; no. 8; pp. 2919 - 2927
• Main Authors: Dmochowski, Roger R., Rovner, Eric S., Connors, Kevin G., Winters, J. Christian
• Format: Journal Article
• Language: English
• Published: United States Wiley Subscription Services, Inc 01.11.2018
• Subjects: Clinical Trials; endpoints; Motivation; outcome measures; Quality of life; stress urinary incontinence; SUI; Urinary incontinence; women's health; endpoints; stress urinary incontinence; outcome measures; SUI; women's health; quality of life; Clinical Trials
• ISSN: 0733-2467; 1520-6777; 1520-6777
• DOI: 10.1002/nau.23817

Objectives To perform a post hoc analysis of the outcome data from a phase III study of a novel incontinence therapy for female stress incontinence (SUI) and to understand the pitfalls inherent to performing trials in SUI. Methods Results of the SUCCESS Trial, a multicenter, prospective, single‐blinded, randomized, sham‐controlled study were assessed post hoc for parameters and variables consistent with changes in the symptoms or signs of SUI. Subjects were randomized to either the investigative device (Vesair Balloon) placement or placebo. The efficacy endpoint analysis herein evaluated a composite of both a >50% reduction from baseline on 1‐hour provocative pad weight test and a ≥10‐point improvement in symptoms on the Incontinence Quality of Life (I‐QOL) questionnaire assessed at the three‐month study visit. Further analysis was performed utilizing more challenging success criteria for each endpoint and a subset of subjects with more severe baseline symptoms. Results The three‐month composite efficacy endpoint was achieved in 42.1% of treatment group subjects compared with 28.1% of control subjects on an intention‐to‐treat analysis (P = .046). This difference between groups was not evident when analyzing each component of the composite endpoint separately. These endpoints diverged from endpoints involving improvement in PGI‐I score and a 50% reduction of incontinence episode frequency as reported by subjects on a diary. Conclusions Individual endpoints may not demonstrate a significant treatment effect as standalone SUI measures but may as components of a composite endpoint. Defining patient selection and outcomes assessments is a critical aspect of study design, experimental flaws, and therapeutic accuracy.