Psychometric validation of symptom severity measures in irritable bowel syndrome with constipation

Background Historically, measures of symptom severity of irritable bowel syndrome with constipation (IBS-C) in clinical trials have not met the evidence requirements described in the FDA guidance on patient-reported outcomes (PROs), which describes the evidentiary requirements and review criteria for patient-reported outcome measures intended to support product approval or labelling claims. Aim Data from two phase 3 trials (N = 1608) of linaclotide for the treatment of IBS-C were analysed to evaluate the psychometric properties of patient-reported outcome measures assessing changes in the severity of abdominal and bowel symptoms. Methods A set of patient-reported outcome assessments addressing abdominal and bowel symptoms, the IBS-C Symptom Severity Measures, were administered daily using interactive voice response system technology. Intraclass correlation coefficients (ICCs), Pearson correlations, factor analyses, F-tests and effect sizes were computed to evaluate the reliability, construct validity, discriminating ability and responsiveness of the IBS-C Symptom Severity Measures in a clinical trial context. Results The IBS-C Symptom Severity Measures showed highly satisfactory test-retest reliability (ICCs ranging from 0.79 to 0.95) and construct validity. Factor analyses indicated one factor for abdominal symptoms and another for bowel symptoms. Known-groups F-tests comparing subgroups based on various responder definitions were statistically significant and in the expected direction, substantiating the discriminating ability of the IBS-C Symptom Severity Measures. Responsiveness statistics (ranging from 0.6 to 2.1) demonstrated these measures are also capable of detecting change. Conclusions The psychometric analysis results strongly support the reliability, construct validity, discriminating ability and responsiveness of the IBS-C Symptom Severity Measures and substantiate the conclusion of linaclotide treatment benefit.