Psychometric analysis of the University of Wisconsin Interstitial Cystitis Scale: implications for use in randomized clinical trials.

摘要

PURPOSE: A psychometric analysis of the University of Wisconsin Interstitial Cystitis Scale was conducted on 30 females previously enrolled in a phase II double-blind randomized controlled trial evaluating the efficacy of six weekly intravesical instillations of TICE BCG. The analyses were to: (1) evaluate the adequacy of the seven individual IC component items for measuring the range of patient responses; (2) verify the 2-factor (IC versus reference) construct of the scale; (3) evaluate the internal consistency and reliability of the IC items; (4) better define the scale's applicability and limitations; and (5) if possible, make recommendations for improvements in the scale. MATERIALS AND METHODS: Standard psychometric analyses were used to perform the evaluation, and included descriptive analysis of individual items, computing of item-total correlations and Cronbach's internal consistency measures, and the application of factor and Rasch analyses. RESULTS: The original 7-item IC scale was found to have ceiling effects that could limit its use in detecting small therapeutic differences. It was also found that the Pelvic item originally assigned to the reference set of items of the scale should be included as an IC item when used in a comparable IC population. After including this item into the IC scale Cronbach's alpha was 0.84, compared with 0.82. CONCLUSIONS: The UW-IC Scale has psychometric properties similar to other measurement instruments used in clinical research, and appears worthy of further study in well-characterized IC populations. The reference items suggest that IC patients do not indiscriminately report high values for generalized body complaints, but do so on bladder related symptoms as recorded by the IC items of the scale. Although the scale has limitations it appears applicable for use in future IC intervention clinical trials.