Knee replacement outcome predicted by physiotherapists: a prospective cohort study

Area of Science:

• Orthopaedic Surgery
• Patient Outcomes Research
• Rehabilitation Medicine

Background:

• Knee arthroplasty (KA) is a common treatment for knee osteoarthritis, yet 20% of patients report dissatisfaction.
• Identifying factors that predict patient satisfaction and recovery is crucial for improving outcomes.

Purpose of the Study:

• To evaluate if physiotherapists' early post-operative prognoses can predict 6- and 12-month outcomes for knee arthroplasty patients.
• To assess the predictive accuracy of these prognoses using patient-reported outcome measures.

Main Methods:

• Physiotherapists provided 0-10 scale prognoses for 361 KA patients.
• Patient outcomes were assessed at 6 and 12 months using the Oxford Knee Score (OKS), EuroQol 5D-3L, and Patient Acceptable Symptom State (PASS).
• Multivariable logistic regression analyzed the prediction of PASS and treatment success, with calibration and discrimination measures.

Main Results:

• Predictive models for PASS and treatment success demonstrated poor to acceptable discriminative values (AUC 0.62-0.73).
• Calibration plots revealed significant uncertainties in the predictions made by physiotherapists.
• The models showed odds ratios between 1.47 and 1.92.

Conclusions:

• Physiotherapists' prognoses are associated with patient-reported outcomes and satisfaction after knee arthroplasty but possess weak predictive power.
• These early prognoses may serve as a valuable supplementary tool for identifying patients requiring additional post-operative care.
• Further research could refine prognostic models to enhance prediction accuracy.