Abstract
Purpose: The Quality of Recovery-40 questionnaire (QoR-40) has been developed, validated and extensively used to assess the quality of life of patients following major surgery. It is composed of 40 questions answered by the patient and organized into six dimensions: emotional state, physical comfort, psychological support, physical independence, pain, and global score. However, this questionnaire has not been validated in an orthopaedic population. The objective of the study was to assess the psychometric properties of the QoR-40 in a population of patients undergoing total knee or hip arthroplasty.
Method: Sixty seven patients undergoing total knee arthroplasty and 65 patients undergoing total hip arthroplasty were recruited. Patients were assessed with the QoR-40 and the SF-12, a validated generic quality of life questionnaire, at the following seven times: three and one week pre-operative; one, two and three days post-operative (short-term); one and three months post-operative (long-term). The following psychometric properties of the QoR-40 were assessed: reliability between three and one week pre-operative using intra-class correlation coefficients (ICC), construct validity with the SF-12 using Pearson product moment correlations, responsiveness to change using effect sizes, floor and ceiling effects, and predictive validity of short-term QoR-40 scores of long-term SF-12 scores.
Results: All dimensions and global scores of the QoR-40 appeared reliable, with ICCs’ ranging from 0.75 to 0.84. For construct validity, physical dimension scores of the QoR-40 (physical comfort, physical independence, pain) were weakly or not significantly related to the physical component scale of the SF-12 (−0.09 to 0.34), except for long-term where the correlations were moderate (0.35 to 0.62). The emotional state score of the QoR-40 was moderately to substantially related to the mental component scale of the SF-12 for all periods (0.40 to 0.78). Effect sizes were higher for the QoR-40 when compared to the SF-12 in the intervals of the first month, while they were higher for the SF-12 in the intervals above one month. The QoR-40 demonstrated ceiling effects for the physical independence dimension pre-operatively and three months post-operatively, and for all periods for the psychological support dimension. As for predictive validity, short-term post-operative QoR-40 scores were weakly or not significantly predictive of long-term postoperative SF-12 scores (0.01 to 0.41).
Conclusion: The QoR-40 appears to be a reliable tool assessing a quality of life construct different from the SF-12 and more responsive to change during the short-term follow-up to surgery. The QoR-40 could be used to assess short-term quality of life following surgery. The QoR-40 should not be used to predict long-term quality of life. To further improve the tool, the psychological support dimension of the QoR-40 should be reconceptualized because of ceiling effects.
Correspondence should be addressed to: COA, 4150 Ste. Catherine St. West Suite 360, Westmount, QC H3Z 2Y5, Canada. Email: meetings@canorth.org
