IMPACT‐III Is a Valid, Reliable and Responsive Measure of Health‐related Quality of Life in Pediatric Crohn's Disease

Abstract

Funding for this study was provided by Centocor Inc. Background: IMPACT is a health-related quality of life (HRQL) questionnaire for pediatric patients with IBD. During the cross-cultural adaptation/translation process in Europe, modifications made to IMPACT resulted in IMPACT-III. Aim: To evaluate the validity, reliability and responsiveness of IMPACT-III in the context of a randomized, controlled trial. Methods: As part of a multi-centre randomized trial of infliximab for moderate to severe pediatric Crohn's disease (CD), HRQL was assessed using IMPACT-III at the North American sites. IMPACT-III is a 35 item, self-administered questionnaire, with total score ranging from 35 (poor) to 175 (best). Using week 10 data from the REACH study, validity was assessed based on established criteria, including change in IMPACT-III score corresponding to change in disease activity (ie, hypothesized that IMPACT score increases as disease activity decreases). Reliability statistics for the IMPACT-III total and domain scores were calculated using the Cronbach's α statistic. The responsiveness ratio of IMPACT-III was analyzed. Using linear regression, the cutoff values for IMPACT-III corresponding to remission and clinical improvement were calculated. Results: IMPACT-III results were available for 76 patients (38% female, mean age=13±2 yrs, mean baseline PCDAI=40±7). Mean IMPACT-III total scores across four disease activity categories (none [n=2], mild [n=7], moderate [n=56], severe [n=10], based on baseline physician global assessment) were significantly different (ANOVA, p=0.046); higher disease activity related to lower IMPACT-III score. The IMPACT-III total score showed excellent reliability (Cronbach's α=0.92). The reliability scores for the 6 domains were good except for the tests/treatment domain at 0.56. The responsiveness ratio for ΔIMPACT was 6.9. Regression line for IMPACT-III (r2=0.158; b=-0.771, p<0.001) resulted in a cutoff value of 143 points for remission (PCDAI ≤10) and for ΔIMPACT-III (r2=0.286; b=-0.877, p<0.001) resulted in an increase of 10.8 points for clinical improvement based on ΔPCDAI of -15 points. Conclusions: We have demonstrated that IMPACT-III, like IMPACT, is a valid and reliable measure of HRQL in pediatric CD. Also, IMPACT-III is responsive to change in HRQL over a 10 week period. These findings provide important measurement characteristics that can be used in the design of subsequent pediatric CD clinical trials.