Validity, reliability and minimum detectable change of the maximum step length test in people with stroke.

Stroke is a significant cause of deficits in balance, mobility and disability. Although tests of stepping speed are associated with balance performance after stroke, relationships between clinical tests of stepping distance and balance performance have not been investigated in people with stroke. A validated test of stepping distance and balance in older adults (the maximum step length [MSL] test), and two clinical measures associated with balance in people with stroke (the five-times-sit-to-stand test [FTSST] and gait speed) were evaluated in nineteen independent community-dwelling people with chronic stroke. There were strong relationships between MSL and performance on the clinical balance measures (Pearson's r 0.69 to -0.88), suggesting that MSL is a valid measure of balance after stroke. Test-retest reliability coefficients were excellent for the MSL tests (ICC both limbs; 0.98). Standard error of measurement expressed as a percentage of mean MSL was computed as 5.6% (hemiparetic leg) and 5.4% (unaffected leg), indicative of low levels of measurement error and excellent absolute reliability of the MSL test in people with stroke. Minimum detectable change expressed as a percentage of mean MSL was low (14.9% to 15.4%), suggesting that MSL may be sensitive to detecting change in physical performance in people with stroke. Advantages of the MSL test are that it is easy to administer, requiring minimal training, equipment, time or space. These advantages, together with the validity of the MSL test, its excellent test-retest and absolute reliability, and its low minimum change percent value suggest that MSL may be a useful measure of balance capabilities in people with stroke.