Heterogeneity of Clinical Characteristics and Treatment Use among Patients with Osteoarthritis in Primary Care Clinics

Purpose: Although guidelines exist for the management of hip and knee osteoarthritis (OA), little is known about how treatment use may vary across health care settings, particularly in the United States (U.S.). This study compared clinical characteristics and treatment use patterns among patients with hip and knee OA in ten Family and Internal Medicine clinics in one healthcare system, which varied in patient panel size and represented both urban and rural settings. Methods: Baseline data were obtained from an ongoing randomized clinical trial of Patient and Provider Interventions for Managing Osteoarthritis in Primary Care in the Duke University Healthcare System. Participants (n = 488; 21–57 per clinic) had hip and / or knee OA, were overweight, and were not meeting physical activity recommendations. Clinical characteristics included the Western Ontario and McMaster University Osteoarthritis Index (WOMAC) scale, self-reported duration of arthritis symptoms, and diagnoses of hip and / or knee OA. Self-reported OA treatment use included: current use of pain medications for OA (yes /no), ever having received physical therapy for knee / hip OA (yes / no), ever having used a knee brace (yes / no), ever having a knee joint injection (yes / no); the latter two questions were only asked for patients with diagnoses of knee OA. For all measures we computed means and standard deviations or proportions for each clinic. Results: Clinical characteristics of patients varied considerably across study sites. Mean WOMAC scores ranged from 35.0 (SD = 16.1) to 44.1 (SD = 15.0) and the mean duration of arthritis symptoms ranged from 8.2 (SD = 9.3) to 13.9 (SD = 12.7) years. Most patients at each clinic had diagnoses of knee OA (90%-100%), but the proportion with hip OA diagnoses varied from 26%-68% across clinics; 23%-61% had both diagnoses. Most patients at each clinic were currently using pain medications to treat their OA (73%-88%). However, use of other treatments was lower and varied more substantially across clinics. In seven of the clinics, less than half of patients had ever received physical therapy for knee OA; proportions ranged from 25%-64% across clinics. Physical therapy for hip OA was even less common (0%-60%). Knee brace use (any type) ranged from 40%-64%, but use of braces with metal supports was much lesscommon (0%-40%). Knee injections were reported by 43%-74% of patients. Conclusions: The heterogeneity of patient characteristics across clinics in this study illustrates the importance of including multiple and diverse sites in clinical and health services trials for knee and hip OA. These data on OA treatment use across different clinics are some of the first to be reported in the U.S. and show considerable variation, particularly in the use of non-pharmacological therapies. Despite the relatively long average duration of symptoms in these patients, physical therapy use was very low in many of the clinics. This represents one important area for improvement and standardization in the practice of OA management, particularly knee OA, since physical therapy is an evidence-based and recommended treatment component.