Can Preoperative Patient-Reported Outcome Measures Predict Clinical Outcomes Following Total Knee Arthroplasty?

Background: The Centers for Medicare and Medicaid Services now mandates the collection of patient-reported outcome measures (PROMs) before and after total knee arthroplasty (TKA), though their utility in predicting clinical outcomes remains unclear. This study compared the power of preoperative PROMs to predict clinical outcomes after TKA to established indices, including the Charlson Comorbidity Index and the Risk Assessment and Prediction Tool (RAPT).

Methods: We retrospectively reviewed 2,923 patients undergoing elective, primary, unilateral TKA who completed the Knee Injury and Osteoarthritis Outcome Score for Joint Replacement and four Patient-Reported Outcomes Measurement Information System (PROMIS) domains within 90 days preoperatively. Outcomes included same-day discharge (SDD) rate, prolonged length of stay (LOS ≥ 48 hours), nonhome discharge, 90-day readmission, and revision surgery. Predictive utility was assessed using receiver operating characteristic curves (area under the curve [AUC]) and multivariable logistic regressions.

Results: The RAPT was the strongest predictor of SDD (AUC = 0.697), extended LOS (AUC = 0.703), and nonhome discharge (AUC = 0.877). The PROMIS Physical Health (AUC = 0.609 for SDD; 0.607 for LOS; 0.696 for discharge) and PROMIS Mental Health (AUC = 0.613; 0.610; 0.655) demonstrated modest predictive value. In adjusted models, higher PROMIS Mental Health was associated with greater odds of SDD (odds ratio (OR) = 1.027, P = 0.003) and lower odds of extended LOS (OR = 0.975, P < 0.001). Higher PROMIS Physical Health was associated with decreased odds of nonhome discharge (OR = 0.935, P = 0.017). The Charlson Comorbidity Index was the strongest predictor of 90-day readmission (AUC = 0.604) and the only significant variable in adjusted models. There were no PROMs that were significant predictors of readmission or revision.

Conclusions: Preoperative PROMs are associated with some clinical outcomes following TKA, but demonstrate limited predictive power compared to established tools like the RAPT score. These findings underscore the need for more robust, outcome-aligned PROMs to improve surgical planning in a value-based care model.