Characterizing Pain and Relief after Joint Replacement in Hip and Knee Osteoarthritis

Purpose: Osteoarthritis (OA) is one of the most prevalent causes of chronic pain. In advanced knee and hip disease, total joint replacement surgery (TJR) is regarded as a safe and successful procedure, providing pain relief and good functional outcomes. Still, a significant proportion of patients continues to experience moderate to severe pain in their replaced joint. Risk factors predictive of pain persistence have been repeatedly studied, using unitary outcome measures. Here, we assess pain using multiple measures, searching for models predictive of pain and pain relief, for both hip and knee OA. Methods: In this prospective observational study, we followed knee OA (n=94) and hip OA (n=25) patients before and up to six months post-surgery. The main inclusion criterion was primary OA, with surgical indication for TJR. Socio-demographic information and medical history were retrieved. Several questionnaires were used to assess multiple domains (pain quality, mood and affect, health and quality of life). We also measured radiographic OA severity (Kellgreen-Lawrence scale and revised OARSI Atlas) and performed a functional evaluation using performance-based tasks (timed up and go test, 6-minute walking test). Four different pain intensity-related scales were used as outcome measures. Our study protocol was approved by the local ethics committee (Centro Hospitalar de São João Ethics Committee), and all patients provided informed written consent prior to participation. Results: Consistent with the literature, hip surgery provided better pain relief than knee surgery. Principal component analysis (PCA) on questionnaires scores and functional evaluation measures revealed five dimensions, accounting for 70% of the variance (image 1). Using a hierarchical regression approach, pain intensity-related scales were modeled based on the previous PCA dimensions, radiographic severity of OA and relevant demographic data (age, gender, BMI). Distinct models were obtained for each pain intensity outcome, both for baseline pain and post-operation pain. The same models failed to generalize to hip OA. Next, an aggregate pain intensity measure was built from all four measures, but in this case, no behavioral measures prior to surgery would predict post-surgical pain, for either knee or hip OA. As a final step we modeled the correlations of the PCA dimensions as a network and examined how its network properties change after surgery, uncovering distinct recovery profiles for hip and knee OA patients (image 2). Conclusions: Our approach to model pain and its persistence in OA provides a different and comprehensive methodology. These results indicate that measures of OA pain remain inadequate and that clinical and behavioral measures prior to surgery don't seem to provide reliable predictive information regarding pain persistence.View Large Image Figure ViewerDownload Hi-res image Download (PPT)