Minimal detectable change and clinically important difference of the stroke impact scale in stroke patients

Objectives. The purpose of this study was to establish the minimal detectable change (MDC) and clinically important differences (CIDs) of the physical domains of the Stroke Impact Scale (SIS) and to assess the proportions of patients' change scores exceeding the MDC and CIDs after stroke rehabilitation. Methods. Seventy-four patients received 1 of 3 treatments for 3 weeks and underwent clinical assessment before and after treatment. The MDC was calculated from the standard error of measurement to indicate a real change with 95% confidence for individual patients (MDC95). Anchor-based and distribution-based approaches were adopted to triangulate the ranges of minimal CIDs. The percentage of patients exceeding MDC95 and minimal CIDs were also calculated. Results. The MDC95 of the strength, activities of daily living/instrumental activities of daily living, mobility, and hand function subscales were 24.0, 17.3, 15.1, and 25.9, respectively. The respective minimal CIDs for these 4 subscales were 9.2, 5.9, 4.5, and 17.8 points, respectively, and the MDC95 and CID proportions were 14% to 43%, 16% to 49%, 10% to 50%, and 23% to 64%, respectively. Conclusions. The change score of an individual patient has to reach 24.0, 17.3, 15.1, and 25.9 on the 4 subscales to indicate a true change. The mean change scores of a stroke group on the 4 subscales should reach 9.2, 5.9, 4.5, and 17.8 points to be regarded as clinically important changes. Future research with larger sample sizes is warranted to validate these estimates. ? 2010 The Author(s).