Analysis of the Cost and Efficacy of Intra-Articular Knee Injections

INTRODUCTION: Intra-articular joint injections have been used with the goal of providing patients with symptomatic relief. Recently, however, the efficacy of corticosteroid (CS) and hyaluronic acid (HA) injections in the management of knee osteoarthritis (OA) is questionable. In this analysis, we investigated the costs associated with injections by assessing overall use, conversion and average time to total knee arthroplasty (TKA), and reimbursement. METHODS: Patients aged 50 to 70 years with a knee-related diagnosis of pain, effusion, or OA were identified in the Humana insurance national data set. Patients who received intra-articular injections were stratified by the type and number of injection(s) received. The subsequent rates of TKA were compared with Kaplan-Meier curves for patients who underwent CS injections, HA injections, and a benchmarking cohort of patients with OA and no history of knee injections in the medical record. Average time to TKA was determined from index diagnosis, and total cost was compared using Wilcoxon rank sum analyses. RESULTS: A total of 778,686 patients were identified. Of these, 637,112 had no knee injection history, while 124,129 received CS and 17,445 received HA injections. The 10-year conversion to TKA was highest in HA cohort (31.6%), followed by the CS cohort (24.0%) and the noninjection cohort (7.3%) (P < 0.001). Time to TKA increased with number of injections for both injection types. For patients who underwent TKA, median cost was greater in HA ($16,687) and CS ($15,563) cohorts relative to noninjection cohort ($14,733) (P < 0.001). DISCUSSION: Compared with the noninjection cohort, both HA and CS cohorts experienced increased costs and increased time to TKA. However, while the cost incurred in HA cohort was greater than that in CS cohort, no appreciable benefit was demonstrated for conversion or time to TKA. Therefore, if intra-articular knee injections are indicated for the nonsurgical management of knee OA, the results of this study support CS over HA.