Multimodal prediction of pain and functional outcomes 6 months following total knee replacement: a prospective cohort study

Background Knee osteoarthritis (OA) is among the most common and disabling persistent pain conditions, with increasing prevalence and impact around the globe. In the U.S., the rising prevalence of knee OA has been paralleled by an increase in annual rates of total knee arthroplasty (TKA), a surgical treatment option for late-stage knee OA. While TKA outcomes are generally good, post-operative trajectories of pain and functional status vary substantially; a significant minority of patients report ongoing pain and impaired function following TKA. A number of studies have identified sets of biopsychosocial risk factors for poor post-TKA outcomes (e.g., comorbidities, negative affect, sensory sensitivity), but few prospective studies have systematically evaluated the unique and combined influence of a broad array of factors. Methods This multi-site longitudinal cohort study investigated predictors of 6-month pain and functional outcomes following TKA. A wide spectrum of relevant biopsychosocial predictors was assessed preoperatively by medical history, patient-reported questionnaire, functional testing, and quantitative sensory testing in 248 patients undergoing TKA, and subsequently examined for their predictive capacity. Results The majority of patients had mild or no pain at 6 months, and minimal pain-related impairment, but approximately 30% reported pain intensity ratings of 3/10 or higher. Reporting greater pain severity and dysfunction at 6 months post-TKA was predicted by higher preoperative levels of negative affect, prior pain history, opioid use, and disrupted sleep. Interestingly, lower levels of resilience-related "positive" psychosocial characteristics (i.e., lower agreeableness, lower social support) were among the strongest, most consistent predictors of poor outcomes in multivariable linear regression models. Maladaptive profiles of pain modulation (e.g., elevated temporal summation of pain), while not robust unique predictors, interacted with psychosocial risk factors such that the TKA patients with the most pain and dysfunction exhibited lower resilience and enhanced temporal summation of pain. Conclusions This study underscores the importance of considering psychosocial (particularly positively-oriented resilience variables) and sensory profiles, as well as their interaction, in understanding post-surgical pain trajectories.