Objectives: Weaknesses of the nine-hole peg test include high floor effects and a result that might be difficult to interpret. In the twenty-five-hole peg test (TFHPT), the larger number of available pegs allows for the straightforward counting of the number of pegs inserted as the result. The TFHPT provides a comprehensible result and low floor effects. The objective was to assess the test-retest reliability of the TFHPT when testing persons with stroke. A particular focus was placed on the absolute reliability, as quantified by the smallest real difference (SRD). Complementary aims were to investigate possible implications for how the TFHPT should be used and for how the SRD of the TFHPT performance should be expressed.
Design: This study employed a test-retest design including three trials. The pause between trials was approximately 10-120 s.
Participants, setting and outcome measure: Thirty-one participants who had suffered a stroke were recruited from a group designated for constraint-induced movement therapy at outpatient clinics. The TFHPT result was expressed as the number of pegs inserted.
Methods: Absolute reliability was quantified by the SRD, including random and systematic error for a single trial, SRD2.1, and for an average of three trials, SRD2.3. For the SRD measures, the corresponding SRD percentage (SRD%) measure was also reported.
Results: The differences in the number of pegs necessary to detect a change in the TFHPT for SRD2.1 and SRD2.3 were 4.0 and 2.3, respectively. The corresponding SRD% values for SRD2.1 and SRD2.3 were 36.5% and 21.3%, respectively.
Conclusions: The smallest change that can be detected in the TFHPT should be just above two pegs for a test procedure including an average of three trials. The use of an average of three trials compared with a single trial substantially reduces the measurement error.
BMJ Publishing Group Ltd, 2019. Vol. 9, no 12, article id e032560