‘Overall pain’ is strongly associated with patient satisfaction: An item‐level longitudinal analysis of the Oxford Knee Score in total and unicompartmental knee arthroplasty

PURPOSE: While the Oxford Knee Score (OKS) is linked with patient satisfaction following knee arthroplasty, item-level associations are unknown. The primary aim of this study was to identify which of the 12 OKS questions were most strongly associated with satisfaction from early to 15-year follow-up to guide clinical practice towards improved patient-perceived outcomes. A secondary aim was to calculate Patient Acceptable Symptom State (PASS) thresholds to provide clinically meaningful benchmarks. METHODS: A total of 9384 total knee (TKA) and 2424 unicompartmental knee arthroplasty (UKA) with OKS and dichotomised satisfaction scores in the New Zealand Joint Registry were included. Score associations and PASS thresholds up to 15 years were assessed using logistic regression and receiver operating characteristic (ROC) analyses. RESULTS: Q1 'overall pain' demonstrated the strongest and most consistent association with satisfaction, with longitudinal consistency for both TKA (6 months: odds ratio (OR) 3.49; 5 years: OR 4.25; 10 years: OR 3.77; 15 years: OR 3.67; all p < 0.001 against the null hypothesis OR = 1) and UKA (6 months: OR 3.88; 5 years: OR 4.25; 10 years: OR 4.31; 15 years: OR 2.89, all p < 0.001). This question demonstrated good discriminatory performance, with AUC values approaching those of the full OKS and a composite score with the four strongest predictors (area under the curve [AUC] 0.87-0.93). PASS thresholds were 35.5 and 36.5 at 6 months, 34.5 and 39.5 at 5 years, 33.5 and 37.5 at 10 years, and 35.5 and 28.5 at 15 years for TKA and UKA, respectively. The PASS threshold for Q1 was 2.5 across procedures at all follow-up. CONCLUSIONS: Despite OKS being a multidimensional instrument, overall pain was the item most strongly associated with satisfaction after knee arthroplasty, suggesting that pain-related symptoms may be more closely aligned with satisfaction than the other domains. PASS thresholds provide real-world benchmarks for clinical practice. LEVEL OF EVIDENCE: Level IV, retrospective case series.