Practice effects and test–re-test reliability of the Five Digit Test in patients with stroke over four serial assessments

Abstract

Objectives: To investigate practice effect and test–re-test reliability of the Five Digit Test (FDT) over four serial assessments in patients with stroke.

Design: Single-group repeated measures design.

Methods: Twenty-five patients with stroke were administered the FDT in four consecutive assessments every 2 weeks. The FDT contains four parts with five indices: ‘basic measures of attention and processing speed’, ‘selective attention’, ‘alternating attention’, ‘ability of inhibition’ and ‘ability of switching’.

Results: The five indices of the FDT showed trivial-to-small practice effects (Cohen’s d = 0.03–0.47) and moderate-to-excellent test–re-test reliability (intra-class correlation coefficient = 0.59–0.97). Practice effects of the five indices all appeared cumulative, but one index, ‘basic measures of attention and processing speed’, reached a plateau after the second assessment. The minimum and maximum values of the 90% confidence interval (CI) of reliable change index modified for practice (RCIp) for this index were [−17.6, 11.2].

Conclusions: One of five indices of the FDT reached a plateau, whose minimum and maximum values of the 90% CI RCIp are useful to determine whether the change in an individual’s score is real. However, clinicians and researchers should be cautious when interpreting the test results of these four indices over repeated assessments.

• Attention
• executive functions
• practice effect
• reliable change index modified for practice
• stroke

Acknowledgements

This study was supported by a research grant from the E-Da Hospital (EDAHT-101013).