Construct validation and correlates of preoperative expectations of postsurgical recovery in persons undergoing knee replacement: baseline findings from a randomized clinical trial

Background: A patient's recovery expectations prior to knee arthroplasty influence postsurgical outcome and satisfaction but a unidimensional measure of expectation has not been reported in the literature. Our primary purpose was to determine the extent to which a patient expectations scale reflects a unidimensional construct. Our second purpose was to identify pre-operative variables associated with patients' expectations. We hypothesized that previously identified predictors of the latent expectation scale score would be associated with expectations and that previously unexplored variables of pain catastrophizing, depressive and anxiety symptoms, self-efficacy and number of painful body regions would also associate with pre-operative expectations. Methods: Our randomized clinical trial had 384 patients assessed prior to knee replacement surgery. The expectations scale along with several predictor variables including WOMAC, psychological distress, and sociodemographic variables were obtained. Confirmatory factor analysis tested the unidimensionality of the measure and structural equation modeling identified predictors of the latent expectations measure. Results: The expectations scale was found to be unidimensional with superior model fit (chi(2) = 1.481; df = 2; p = 0.224; RMSEA = 0.035; 90% CI = [0-0.146]; CFI = 0.999; TLI = 0.993). The only variable significantly associated with expectations in the multivariate model was self-efficacy. Conclusions: The expectations scale used in our study demonstrated unidimensionality and has strong potential for clinical application. Poor self-efficacy is a potential target for intervention given its independent association with expectation. Addressing expectations directly and indirectly through self-efficacy assessment may assist in better aligning patient's expectations with likely outcome.