Background: Osteoarthritis (OA) severity as demonstrated by preoperative radiographs and preoperative pain play an important role in the indication for total knee arthroplasty (TKA). We investigated whether preoperative radiographic evidence of OA severity modified the effect of preoperative self-reported pain on postoperative pain and function 1 and 2 years after TKA for OA.
Methods: Data from the Longitudinal Leiden Orthopaedics Outcomes of Osteoarthritis Study (LOAS), a multicenter cohort study on outcomes after TKA, were used. OA severity was assessed radiographically with the Kellgren and Lawrence (KL) score (range, 0 to 4). Pain and function were evaluated with the Knee injury and Osteoarthritis Outcome Score (KOOS). After adjustment for body mass index (BMI), age, sex, and the Mental Component Summary scores from the Short Form-12, multivariate linear regression analyses with an interaction term between the preoperative KL score and preoperative pain were performed.
Results: The study included 559 patients. The preoperative KL score was independently associated with 1-year postoperative pain and function (β = 5.4, 95% confidence interval [CI] = 1.4 to 9.4, and β = 7.7, 95% CI = 3.2 to 12.2), while preoperative pain was associated only with postoperative pain (β = 0.3, 95% CI = 0.1 to 0.6) and not with postoperative function (β = 0.2, 95% CI = -0.2 to 0.5). Comparable associations were found between 2-year postoperative pain and KL score (β = 8.0, 95% CI = 3.2 to 12.7) and preoperative pain (β = 0.5, 95% CI = 0.1 to 0.8) and between 2-year postoperative function and KL score (β = 7.7, 95% CI = 3.2 to 12.2). The study showed a trend toward the KL score modifying the effect of preoperative pain on 1-year postoperative pain (β = -0.1, 95% CI = -0.1 to 0.0) and 2-year postoperative pain (β = -0.1, 95% CI = -0.2 to 0.0) and on 1 and 2-year function (β = -0.1, 95% CI = -0.2 to 0.0 for both), with the effect of preoperative pain on postoperative pain and function seeming to become less important when there was radiographic evidence of greater preoperative OA severity.
Conclusions: Patients with less pain and higher KL grades preoperatively had better function and pain outcomes 1 and 2 years after TKA. However, the effect of preoperative pain on the postoperative outcomes seems to become less important when the patient has radiographic evidence of more severe OA. We believe that analysis of the severity of preoperative pain is an important proxy for optimal postoperative patient outcome.
Level of evidence: Prognostic Level II. See Instructions for Authors for a complete description of levels of evidence.