SAT0557 10-year trajectories of pain in early knee and hip osteoarthritis; the check study

Background Osteoarthritis (OA) is one of the leading causes of chronic pain worldwide. Yet, relatively little is known about the trajectories of pain in early knee and hip OA. Objectives To determine subgroups of patients with homogenous patterns of pain over 10 years follow-up, among patients with early hip and/or knee complaints in primary care. Methods Data from the CHECK cohort (Cohort Hip and Cohort Knee) were used. For this cohort, 1002 patients between 45 and 65 years at or within 6 months from their first consultation at a general practitioner for symptoms of their hips and/or knees, were included and followed for 10 years. A numeric rating scale (NRS) for perceived pain was obtained at baseline and after 2, 5, 8 and 10 years, or until total joint replacement. Using these longitudinal data, subgroups of patients with comparable trajectories over time were identified using Latent Class Growth Analysis (LCGA). Models with 3 to 6 classes when using linear, cubic and quadratic trajectories were evaluated using Mplus software. Results At baseline, the 998 subjects (79% women) with ≥2 NRS data points available had a mean age of 55.9±5.2 years and a mean BMI of 26.2±4.0 kg/m2. In total, 410 patients reported knee pain only (41%), 173 hip pain only (17%), and 415 reported both knee and hip pain (42%) at baseline. On a joint level, 156 knees and 160 hips had KL-grade ≥2 at baseline. The by LCGA derived models of 3 groups with a linear trajectory and of 6 groups with a cubic trajectory resulted in comparable goodness of fit indicators (Bayesian Information Criteria 17 991 vs. 17927, Akaike Information Criteria 17 927 vs. 17761, and entropy 0.694 vs. 0.683 for the 3 and 6 group models, respectively). Both models and the corresponding trajectories are presented in the figure 1. In the 3 group model, the ‘high pain trajectory’ (group 2, blue line, n=206) contained most patients with knee and hip complaints (53%), females (84%), TJR (21%), and the highest mean BMI (27.6±4.6). The ‘low pain trajectory’ (group 3, green line, n=441) contained most patients with only hip (19%) and only knee (48%) complaints, and the lowest mean BMI (25.3±3.6). In the 6 group model, the ‘always high pain trajectory’ (group 1, red line, n=176) contained most females (86%) and TJR (24%), and the highest mean BMI (27.7±4.7). The ‘always low pain trajectory’ (group 5, brown line, n=289) contained most patients with only knee complaints (51%) and the lowest mean BMI (25.0±3.4). The ‘decreasing pain trajectory’ (group 2, blue line, n=37) contained most patients with only hip complaints (22%). The ‘fluctuating high pain trajectories’ (groups 3 (n=88) and 4 (n=142), green and pink lines) contained most females (3:83% and 4:86%), and patients with knee and hip complaints (3:61% and 4:44%). Conclusions The 6 group model identified more extreme groups with lower minimal and higher maximal prevalence of the presented clinical characteristics. In the end, the conclusion drawn from the two models appear similar; Patients presenting with both knee and hip complaints had less favourable pain trajectories over the following 10 years and the number of total joint replacements was the highest in the groups always reporting high pain scores. Disclosure of Interest None declared