A REDUCED WOMAC FUNCTION SCALE - DERIVATION AND VALIDITY

Abstract

Presently, many instruments exist for assessing both patient - and surgeon-based satisfaction after joint replacement, including both generic (measures of general health status) and disease specific measures. As such, the US PORT study (1995) recommends use of both the WOMAC and SF-36. However, this means that studies need to incorporate at least these two lengthy questionnaires into protocols, which increases the pressure on patients for both time and difficulty, but also introduces some duplication of data.

The SF-36 has been successfully reduced and validated to a 12 item questionnaire (SF-12) which can be used as a summarised generic health score. It would be of great benefit if a reduced version of the WOMAC could be derived to give a similar summarised disease-specific measurement tool.

To derive and assess the validity of a reduced function scale of the WOMAC for patients with osteoarthritis of the hip and knee.

All unilateral data from 12 centres world-wide (UK US Canada and Australia) involved in an international, multi-centre outcome study for patients undergoing TKR were included for analysis. The reduced scale was derived from pre-op and 3 month post op data using a combination of data-driven analysis and purely clinical methods. The reduced WOMAC was then extensively validated in three key areas; validity, reliability and responsiveness using 12 month post-op data from the study and data from the Medicare Hip Replacement Study.

Data from 898 patients pre-operatively and 806 patients at 3-months were used for the data driven section of analysis. For the clinical section, 30 members of the orthopaedic community were surveyed as to their opinions of which items should be retained in the reduced version of the scale. These results were then combined to produce a reduced function scale of 7 items to be used in conjunction with the 5-item pain scale. The questions remaining in the scale (and their original number in the scale) were: 2) ascending stairs, 3) rising from sitting, 6) walking on flat, 7) getting in/out of car, 9) putting on socks/stockings, 10) rising from bed and 14) sitting. This reduced scale was then scrutinised to ensure it’s validity (both construct and content), reliability (both internal consistency and reproducibility) and responsiveness (using Standardised Response Means). When examining 12 month data the reduced scale compared favourably with the full scale both overall, and when sub-divided by age, sex and country. It’s construct validity was confirmed by significant positive correlation with the SF-36 physical component score, the knee society function score, the Oxford knee score, and for the hip data, the Harris hip score and SF-12 physical component score. Cronbach’s alpha was consistently high (α> 0.85) with the reduced scale, showing it to be reliable, and the SRM’s indicated that the reduced scale may even be better at detecting change than the full scale.

This reduced WOMAC has been successfully derived and validated for use as a summarised and more practical version of the full WOMAC scale.