The minimal clinically important difference of the mini-balance evaluation systems test in patients with early subacute stroke

Abstract

Background: There is insufficient evidence regarding the minimal clinically important difference (MCID) of the Mini-Balance Evaluation Systems Test (Mini-BESTest).

Objective: To determine the MCID of the Mini-BESTest in patients with early subacute stroke.

Patients and methods: In this prospective cohort study, the Mini-BESTest score of 50 patients with stroke was obtained within 1 week of their admission, their Mini-BESTest and Global Rating of Change Scale (GRCS) scores were obtained at discharge. The GRCS scores were reported by both the patients and their physical therapists. We evaluated the correlation between the Mini-BESTest change scores and the GRCS by determining Spearman's rank correlation coefficient. The MCID was calculated using 0.5× standard deviation (SD) for the distribution method and the change difference and receiver operating curve (ROC) for the anchor method.

Results: The mean (SD) number of days between evaluations was 15.4 (4.8), and the Mini-BESTest score at admission was 17.7 (5.2) and 23.1 (3.5) at discharge. The correlation between the GRCS and the change in the Mini-BESTest score was 0.28 (p = .04) for the patients and 0.54 (p < .001) for the therapists. The MCID based on the distribution method was 3 points for 0.5× SD. The MCID values based on the anchor method were 2.3 for the change difference and 0.5 for the ROC in the patient-rated GRCS, and 4.2 for the change difference and 4.5 for the ROC in the physical therapist-rated GRCS.

Conclusions: The MCID based on the anchor method was 4.2-4.5 points, and the MCID based on the distribution method was 2.3 points.

Keywords: Stroke; minimal clinically important difference; outcome assessment; postural balance; rehabilitation.