The responsiveness and minimally important difference for the Accidental Bowel Leakage Evaluation questionnaire.

Abstract

We describe the responsiveness and minimally important difference (MID) of the Accidental Bowel Leakage Evaluation (ABLE) questionnaire. Women with bowel leakage completed ABLE, Patient Global Impression of Improvement, Colo-Rectal Anal Distress Inventory, and Vaizey questionnaires pretreatment and again at 24weeks post-treatment. Change scores were correlated between questionnaires. Student's t tests compared ABLE change scores for improved versus not improved based on other measures. The MID was determined by anchor- and distribution-based approaches. In 266 women, the mean age was 63.75 (SD = 11.14) and 79% were white. Mean baseline ABLE scores were 2.32 ± 0.56 (possible range 1-5) with a reduction of 0.62 (SD = 0.79) by 24weeks. ABLE change scores correlated with related measures change scores (r = 0.24 to 0.53) and differed between women who improved and did not improve (all p < 0.001). Standardized response means for participants who improved were large ranging from -0.89 to -1.12. Distribution-based methods suggest a MID of -0.19 based on the criterion of one SEM and -0.28 based on half a standard deviation. Anchor-based MIDs ranged from -0.10 to -0.45. We recommend a MID of -0.20. The ABLE questionnaire is responsive to change, with a suggested MID of -0.20.