Patterns of Therapy Switching, Augmentation, and Discontinuation After Initiation of Treatment With Select Medications in Patients With Osteoarthritis

Background: Osteoarthritis (OA) is a debilitating condition characterized by chronic pain. Several pain medications are recommended, and patients frequently alternate among these medications. Objectives: The purpose of this study was to examine the use of pain medications in clinical practice with respect to recommended guidelines. This objective was accomplished by evaluating patterns of switching, augmentation, and discontinuation after treatment initiation with select medications in patients with OA. Methods: The LifeLink Health Plan Claims Database was used to select patients with OA (International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] code 715.XX) who were newly prescribed (index event) nonsteroidal antiinflammatory drugs (NSAIDs), cyclooxygenase (COX)-2 inhibitors, acetaminophen, tramadol, weak opioids, and strong opioids. Descriptive statistics, Kaplan-Meier analyses, and the COX proportional hazards model were used to assess therapy switching, augmentation, and discontinuation during the 12-month postindex period. Patterns of intraarticular injections and joint replacement surgeries among the cohorts were also evaluated. Results: Substantial proportions of OA patients switched, augmented, or discontinued therapy during the postindex period. Rates of therapy switching, augmentation, and discontinuation, respectively, were significantly different (all P < 0.0001 for overall effects using log-rank tests) across the evaluated medication cohorts: NSAIDs, 22.3%, 6.7%, 93.2%; COX-2 inhibitors, 27.5%, 10%, 87.4%; acetaminophen, 46.0%, 6.5%, 98.7%; tramadol, 44.5%, 8.4%, 95.6%; weak opioids, 27.2%, 4.1%, 98.3%; and strong opioids, 41.1%, 3.3%, 97%. Therapy switching, augmentation, and discontinuation occurred within 2 months after treatment initiation in two thirds of patients and within 6 months in >90% of patients. The patterns of intraarticular injections were significantly different across treatment cohorts, as were the patterns of joint replacement surgeries (both P < 0.0001 for overall effects using log-rank tests), with average times to surgery that appeared longer with acetaminophen, NSAIDs, and COX-2 inhibitor initiators (416-447 days) than with tramadol and opioids (354-385 days). Conclusions: Results indicate that therapy switching and discontinuation were frequent among OA patients initiating treatment with the currently recommended medication classes and might suggest suboptimal pain relief or potentially intolerable therapy-related side effects.