VALIDATION OF SCALES USED TO ASSESS OUTCOME OF CMC ARTHROPLASTY

Abstract

This study determined the validity of three patient self-report scales (PRWE, DASH and AUSCAN) to assess outcomes of CMC arthroplasty. Factor analyses did not support the described structure of the three scales. There was a strong relationship between pain or function subscales across different instruments (r> 0.80). Known construct testing regarding WSIB status and arthritis severity supported the discriminative validity of all scales (p< 0.05) except for the function (PRWE) and stiffness (AUSCAN) subscales (p=0.08). Separation of pain/function concepts may be difficult when evaluating outcomes in hand arthritis. The DASH is not unidimensional in this population

This study determined the concurrent validity of patient self-report scales to assess outcomes of CMC (carpometacarpal) arthroplasty.

The subscale structure of the PRWE, DASH and AUSCAN is not valid for a patient population with hand arthritis – Pain, function and stiffness do not differentiate as separate concepts. It appears as though function can be separated into separate components addressing strength and fine motor hand function on all three scales. The DASH was not unidimensional.

Reporting of outcomes following CMC arthroplasty should utilize either the questionnaires total scores or validated factors. Use of unvalidated subscales should be avoided.

Factor analyses did not support the described structure of any of the three scales. PRWE three subscales- two factors; AUSCAN – pain and stiffness items loaded on one factor, function items separated into two factors; DASH – four factors. The largest factor on the DASH contained items relating to symptoms and participation restrictions. Items relating to hand function also separated into a separate factor. Correlational analyses indicated a strong relationship between pain or function subscales across instruments ( r> 0.80) and low correlation with hand appearance (r< 0.20). Tests of known constructs on WSIB status or arthritis severity supported the discriminative validity of all scales (p< 0.05), except for the function subscale off the PRWE or the stiffness subscale of the AUSCAN (p=0.08).

Factor analyses, inter-scale correlations and tests of known constructs were conducted on the Patient Rated Wrist Evaluation (PRWE), Disability of the Arm, Shoulder, Hand (DASH) and the AUSCAN (osteoarthritis of the hand).