Neurological deficits after a stroke are often classified according to motor function for the making of clinical decisions related to discharge and rehabilitation. Participants in clinical stroke studies are also stratified by motor function to avoid a sampling bias. This post hoc analysis examined a suite of upper limb functional assessment tools to test the hypothesis that motor function of survivors of stroke can be stratified using 2 simple tests of manual dexterity despite the heterogeneity of the population. The functional ability of the more affected hand and arm was assessed for 67 hemiparetic patients, aged 18 to 83 years (mean ± standard deviation, 59.8 ± 14.0 years), at 1 to 264 months after a stroke (23.6 ± 39.6 months) using the Wolf Motor Function Test (WMFT), upper limb motor Fugl-Meyer Assessment (F-M), Box and Block Test (BBT), grooved pegboard test, and wrist range of motion. The strength of the proposed stratification scheme was tested with a hypothesis-driven hierarchical cluster analysis using standardized raw scores and dichotomous BBT and grooved pegboard test values. The most salient discriminator between low and higher motor function was the ability to move >1 block on the BBT. Within the higher function group, the ability to place all 25 pegs on the grooved pegboard test discriminated between moderate and high motor function. The derived scheme was congruent with clinical observations. The WMFT timed tasks, F-M scores, and range of motion did not discriminate functional groups.
Two simple unambiguous and objective tests of gross (BBT) and fine (grooved pegboard test) manual dexterity discriminated 3 groups of motor function ability for a heterogeneous group of patients following stroke.