Premium
Test−Retest Reliability and Convergent Validity of Three Manual Dexterity Measures in Persons With Chronic Stroke
Author(s) -
Ekstrand Elisabeth,
Lexell Jan,
Brogårdh Christina
Publication year - 2016
Publication title -
pmandr
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.617
H-Index - 66
eISSN - 1934-1563
pISSN - 1934-1482
DOI - 10.1016/j.pmrj.2016.02.014
Subject(s) - intraclass correlation , convergent validity , rank correlation , spearman's rank correlation coefficient , cohen's kappa , test (biology) , reliability (semiconductor) , physical therapy , stroke (engine) , medicine , kappa , concurrent validity , correlation , cronbach's alpha , psychology , physical medicine and rehabilitation , statistics , clinical psychology , psychometrics , mathematics , internal consistency , mechanical engineering , paleontology , power (physics) , physics , geometry , quantum mechanics , engineering , biology
Abstract Background Decreased manual dexterity is common in persons after stroke. Different measures are used to assess manual dexterity, but a lack of knowledge exists about their reliability and how they are related. Objective To evaluate the test−retest reliability and convergent validity of 3 manual dexterity measures after stroke. Design A test−retest design. Setting University Hospital. Participants Forty‐five persons (mean age 65 years) with mild‐to‐moderate impairments in the upper extremity at least 6 months after stroke. Main Outcome Measures Manual dexterity was assessed on 2 occasions, 1 week apart using the Box and Block Test (BBT), the Nine‐Hole Peg Test (NHPT), and the modified Sollerman Hand Function Test (mSHFT). The reliability of the BBT and NHPT was evaluated with the intraclass correlation coefficient together with systematic and random measurement errors. Reliability of the mSHFT was evaluated with the Kappa coefficient and the Svensson rank‐invariant method (percent agreement and systematic and random disagreements). Convergent validity of the total scores was evaluated with the Spearman rank correlation coefficients (rho). Results The intraclass correlation coefficient for the BBT and the NHPT ranged from 0.83 to 0.99. Significant systematic measurement errors were found for both tests and hands. The Kappa coefficient for the total sum score of the mSHFT was 0.95 for the more affected hand and 0.59 for the less affected hand. One of the 3 items showed systematic disagreements for both hands. The convergent validity (rho) for the more affected hand ranged from 0.41 (BBT versus mSHFT) to −0.68 (NHPT versus mSHFT). Conclusion The test−retest reliability of the BBT, NHPT and mSHFT was high but all measures showed learning effects. The relationships between the 3 measures indicate that they partly complement one another. The BBT may be preferred for persons with moderate impairments of the upper extremity and the NHPT and the mSHFT for persons with milder impairments. As the mSHFT has the advantage of reflecting activities in daily life it may be a suitable alternative to the NHPT. Level of Evidence III