Summary

In this study, we demonstrate that CSIs do not appear to uniquely increase the risk of osteonecrosis when assessed in reference to patients with similar stages of hip pathology.

Abstract

Introduction

Intraarticular corticosteroid injections (CSI) are a key component in the conservative management of many painful hip conditions. Importantly, recent reports have raised questions about their safety, demonstrating one-year rates of osteonecrosis after CSI as high as 17%-27%. However, it is likely these results may have been related to the specific patient samples, and whether CSIs—as opposed to the underlying disease—may lead to osteonecrosis is unknown. Therefore, the purpose of this study was to assess the association between CSIs and osteonecrosis compared to a similar patient population undergoing non-CSI injections.

Methods

This was a retrospective, propensity-matched cohort study of patients undergoing intraarticular hip injection in the MarketScan database from 2007-2017. Patients receiving hip CSI were matched to patients receiving hip hyaluronic acid (HA) injections based on age, sex, hip pathology (osteoarthritis [OA], unspecified pain, other), year, baseline time in database prior to CSI, follow-up time, and geographic region using propensity-scores. The patients’ first injections were identified and time to development of osteonecrosis was analyzed using Kaplan-Meier curves. Hazard ratios (HR) were estimated using Cox-proportional hazard models. Analyses were censored on loss to follow-up and hip surgery for a non- osteonecrosis indication. Patients with a history of osteonecrosis at the time of their first hip CSI or who received both types of injections were excluded.

Results

This study included 1,650 patients undergoing intraarticular hip injections (825 HA, 825 CSI). All baseline factors were matched between groups, including mean (95% CI) age (52.9 [52.3-53.5] years vs. 53.1 [52.5-53.8] years, p=0.59), sex (55.5% female for both, p=1.0), and diagnosis (59.0% OA vs. 58.1% OA, p=0.75) (Table 1). Estimated cumulative one- year incidence of osteonecrosis was 2.6% (1.5%-4.2%) for patients receiving CSI and 2.7% (1.7%-4.2%) for patients receiving HA (p=0.82) (Figure 1). Similar results were seen across the entire study period (HR 1.07, 95% CI 0.60-1.91), and results persisted in a range of sensitivity analyses.

Discussion

AND CONCLUSION:While chronic oral steroid use is well known to be associated with non-traumatic osteonecrosis of the femoral head, whether CSIs were associated with osteonecrosis was unclear. In this study, we demonstrate that CSIs do not appear to uniquely increase the risk of osteonecrosis when assessed in reference to patients with similar stages of hip pathology. Obtaining a proper control group can be challenging in non-randomized studies due to varying degrees of baseline disease; by leveraging HA injections and matching on patient factors, we were able to minimize the impact of residual confounding. Although this study cannot determine whether intraarticular injections themselves (regardless of drug) lead to osteonecrosis, these results suggest that rates of osteonecrosis after CSI may often be due in part to the natural course of the underlying disease.