Co-variation of tests commonly used in stroke rehabilitation

The aim of the present study was to analyse the co-variation of different tests commonly used in stroke rehabilitation, and specifically used in a recent randomized, controlled study of two different physiotherapy models in stroke rehabilitation.

Correlations of the performed tests and recordings from previous work were studied. The test results from three-month, one-year and four-year follow-up were analysed in an SPSS Version 11 statistical package with Pearson and Spearman correlations.

There was an expected high correlation between the motor function tests, both based on partial and total scores. The correlations between Nottingham Health Profile Part 1 and Motor Assessment Scale (MAS), Sødring Motor Evaluation Scale (SMES), the Berg Balance Scale (BBS) and Barthel Activities of Daily Living (ADL) index were low for all items except physical condition. The correlations between registered living conditions, assistive devices, recurrent stroke, motor function (MAS, SMES), ADL (Barthel ADL index) and balance (BBS) were high. The same variables showed weak or poor correlation to the Nottingham Health Profile (NHP).

The co-variations of motor function tests and functional tests were high, but the co-variations of motor, functional and self-reported life-quality tests were poor. The patients rated themselves on a higher functional level in the self-reported tests than was observed objectively in the performance-based tests. A possible reason for this is that the patients may have been unaware they modified their performance to adjust for physical decline, and consequently overestimate their physical condition. This result underlines the importance of both performance-based and self-reported tests as complementary tools in a rehabilitation process.