Jordan Starr1, Irene Rozet1, Alon Ben-Ari2. 1. Department of Anesthesiology and Pain Medicine, University of Washington, Seattle, WA. 2. Northern California VA Health Care System, Sacramento, CA.
Abstract
OBJECTIVES: Total knee arthroplasty (TKA) is a procedure to improve quality of life. However, some patients require early total knee revision (TKR). Chronic opioid use before TKA is associated with TKR. No risk calculator including opioid use or other risk factors is currently available for predicting TKR. MATERIALS AND METHODS: We retrospectively analyzed medical records of Veterans Affairs patients who underwent TKA from January 1, 2006 to January 1, 2012. Patients were followed until January 1, 2013. Chronic opioid use was defined as opioid use for ≥3 months preoperatively. A cross-validated Cox proportional hazards model was created to predict TKR before initial TKA. Model performance was evaluated by the mean absolute error at 1 and 5 years. RESULTS: Totally, 32,297 patients were included. A risk calculator was generated with a mean absolute error of 0.1% at 1 year and 3.6% at 5 years. Chronic opioid use was a significant predictor of TKR (hazard ratio [HR], 1.27; 95% confidence interval, 1.13-1.43; P<0.001). Other model variables were age (HR, 0.95; P<0.001), female sex (HR, 0.77; P=0.020), body mass index (HR, 0.99; P=0.022), diabetes (HR, 1.20; P=0.001), chronic kidney disease (HR, 1.48; P<0.001), and nonchronic opioid use (HR, 1.07; P=0.313). DISCUSSION: Preoperative chronic opioid use is a predictor of TKR. Using this association and others, a TKA revision risk calculator was generated at http://www.bit.do/tka.
OBJECTIVES:Total knee arthroplasty (TKA) is a procedure to improve quality of life. However, some patients require early total knee revision (TKR). Chronic opioid use before TKA is associated with TKR. No risk calculator including opioid use or other risk factors is currently available for predicting TKR. MATERIALS AND METHODS: We retrospectively analyzed medical records of Veterans Affairs patients who underwent TKA from January 1, 2006 to January 1, 2012. Patients were followed until January 1, 2013. Chronic opioid use was defined as opioid use for ≥3 months preoperatively. A cross-validated Cox proportional hazards model was created to predict TKR before initial TKA. Model performance was evaluated by the mean absolute error at 1 and 5 years. RESULTS: Totally, 32,297 patients were included. A risk calculator was generated with a mean absolute error of 0.1% at 1 year and 3.6% at 5 years. Chronic opioid use was a significant predictor of TKR (hazard ratio [HR], 1.27; 95% confidence interval, 1.13-1.43; P<0.001). Other model variables were age (HR, 0.95; P<0.001), female sex (HR, 0.77; P=0.020), body mass index (HR, 0.99; P=0.022), diabetes (HR, 1.20; P=0.001), chronic kidney disease (HR, 1.48; P<0.001), and nonchronic opioid use (HR, 1.07; P=0.313). DISCUSSION: Preoperative chronic opioid use is a predictor of TKR. Using this association and others, a TKA revision risk calculator was generated at http://www.bit.do/tka.