Comparison of responsiveness of the Barthel Index and modified Barthel Index in patients with stroke

Purpose: To compare the group- and individual-level responsiveness of the Barthel Index (BI) and modified BI (MBI) in patients with early subacute stroke. Materials and methods: The BI and MBI scores of 63, 63, and 55 patients were retrieved at 3 time points with a 3-weeks interval. The group-level responsiveness was examined using paired t-test and standardized response mean (SRM). The individual-level responsiveness was examined by the percentage of patients who achieved significant improvement exceeding the corresponding minimal detectable change. Results: At the group level, the MBI showed significantly larger SRMs than did the BI in the 1st–2nd assessment (1.10 vs. 0.81 [95% CI of mean difference = 0.05–0.38]) and the 2nd–3rd assessment (0.94 vs. 0.72 [95% CI of mean difference = 0.04–0.41]). At the individual level, the MBI detected significantly more patients with significant improvement than the BI for the 1st–2nd assessments only (34.9 vs. 25.4% [95% CI of mean differences = 3.2–17.5]). Conclusions: The MBI has better responsiveness than the BI at both the group and individual levels in the patients with early subacute stroke. The MBI is recommended for clinical and research use as an outcome measure for patients with stroke.IMPLICATIONS FOR REHABILITATION The MBI is recommended for clinical and research applications because of its superior ability to detect subtle changes in ADL performance in patients with subacute stroke. The MBI and the BI have equal responsiveness for patients whose magnitude of improvement of ADL is substantial.