METHODS: This observational cohort study included data from all facilities owned by the 121 nonfederal hospitals in the State of Iowa, July 2012 through September 2017. One end point was the percentage of all lumbar or sacral transforaminal or interlaminar epidural injections where the patient had received ≥4 such injections during the preceding 364 days. Comparisons also were made among hospitals’ percentages of injections that were the fifth or greater (ie, patient had already received ≥4 during preceding 364 days) using Bonferroni-adjusted conservative 95% confidence intervals.
RESULTS: There were 48,270 unique patients who underwent at least 1 lumbosacral epidural steroid injection. The patients received care at 112 hospitals’ facilities. Most patients received no additional steroid injections within 364 subsequent calendar days after the first steroid injection (54.1%). There were ≥5 steroid injections for 1.27% of patients (ie, the injection was the fifth or greater). Among the 39 hospitals in Iowa that performed overall at least 1 steroid injection every 4 days, there were 6 hospitals at which the percentages of injections that were the fifth or greater significantly exceeded the overall prevalence of 1.91% (range: 3.0%–6.4%). There were 14 of the 39 hospitals with prevalences significantly less.
CONCLUSIONS: Although most patients received only 1 lumbosacral steroid injection within 1 year, 1.27% of patients received 5 or more, and 1.91% of injections were the fifth or greater. Several hospitals had significantly greater than the overall average percent of steroid injections which were fifth or more. This heterogeneity warrants study of whether annual steroid injections per patient should be a clinical quality measure for the care received by patients with lower back pain or whether payment should be greater when injections are in accordance with guidelines.