Epidural injections and selective nerve blocks are widely used for pain relief, but steroid usage is controversial due to safety concerns. We carried out this retrospective cohort study to estimate the incidence rates of neurological complications associated with epidural and selective nerve blocks, in relation to steroid use patterns.
Materials and Methods:
Using a national insurance claims database, we identified patients who received at least one epidural injection or nerve block from 2009 to 2013. We estimated incidence rates and hazard ratios in propensity score-matched cohorts stratified by steroids, using the Charlson comorbidity index, age, sex, anesthetics, and antithrombotics as variables. We included cases attending hospital within 24 hours after injection and treated for neurological complications.
Incidence rates of neurological complications per 100,000 person-days for injections with and without steroids were 1.48 (95% confidence interval [CI]: 1.25-1.65) and 0.86 (95% CI: 0.66-1.30), respectively; rates for particulate steroid injections and nonparticulate steroid injections were 1.73 (95% CI: 1.41-1.95) and 0.90 (95% CI: 0.43-1.47), respectively. The adjusted hazard ratio (aHR) of neurological complications for injections with versus without steroids was 1.71 (95% CI: 0.96-2.49). The aHR of particulate versus nonparticulate steroid injections was 4.98 (95% CI: 1.01-262.35), at the cervicothoracic level. The aHR of neurological complications for nonparticulate steroids compared with nonsteroidal injections was 0.97 (95% CI: 0.46-3.01).
At the cervicothoracic level, the incidence rate of neurological complications with particulate steroid injections was higher than that with nonparticulate steroid injections. Injections with nonparticulate steroids and without steroids were equally safe.