Minimum Detectable Change and Minimal Clinically Important Difference for the Box and Block Test and Fugl-Meyer Assessment in Moderate to Severe Chronic Stroke

Abstract

Objectives

To quantify the Minimum Detectable Change (MDC) and the Minimal Clinically Important Difference (MCID) of the Box and Blocks Test (BBT) and Upper Extremity Fugl-Meyer Assessment (FMA-UE), and to determine whether there is a BBT floor effect for individuals with chronic stroke and moderate to severe motor impairment.

Design

Analysis of pre- and postinterventional data using anchor and distribution-based methods.

Setting

Research laboratory.

Participants

Fifty-three adults (N=53) with chronic stroke. Age (mean ± SD)=58.5±10.7; 71% men; mean time poststroke=7.4±6.8 years; baseline FMA=20.6±7.6.

Interventions

Randomized controlled trial, including 24 sessions over 8 weeks. Each session included 40 trials of reach-grasp-retrieve-release practice using the ReIn-Hand, with or without robotic support of the arm weight.

Main Outcome Measures

A BBT floor effect emerged among participants with baseline FMA-UE (FMA_bl)≤18. Above this threshold, the linear regression model reported that for every 1-unit increase in FMA_bl, MDC for the BBT (MDC_BBT) increased by 0.46 blocks/min. MDC for the FMA (MDC_FMA) followed a concave quadratic pattern with FMA_bl, increasing until reaching a peak when FMA_bl was 21-26, then declining. Global Rating of Change was not correlated with changes in BBT (ρ=0.22; P=.29) or FMA (ρ=0.19; P=.23). Using distribution-based methods, we found that MCID for the BBT (MCID_BBT) increased linearly with baseline FMA-UE: for every 10-unit increase in FMA_bl, MCID_BBT increased by 1 block/min. There is no linear relationship between baseline FMA-UE and MCID_FMA. Overall MCID_FMA=2.1 points.

Conclusions

Because of its floor effect, the BBT may not be appropriate for those with severe motor impairment poststroke. MDC values for both BBT and FMA-UE depend on motor impairment severity and should be interpreted using regression model-based estimates. In valid ranges of BBT and FMA-UE, subjective perceived change (Global Rating of Change) showed limited association with them. Distribution-based MCID values were small but meaningful, though they may not reflect participants’ perceived importance.