Incidence and risk factors for complications following cervical epidural steroid injections.

Background context: Cervical epidural steroid injections (CESIs) are commonly used in management of cervical pathology. While severe complications have been reported, they are rare, and thus their incidence and risk factors have not been well characterized.

Purpose: To determine the incidence of complications 7, 30, and 90 days after CESI and to identify associated risk factors.

Study design: Retrospective cohort study of a large, national administrative claims database (Meritive^TM Marketscan®).

Patient sample: The study included 229,412 patients (mean age 52.2 ± 11.3 years, 57.1% female) undergoing 448,209 CESIs from 2008 to 2015.

Outcome measures: Complication rates following CESI, and risk factors for complication.

Methods: Adults undergoing a CESI with active enrollment for one year prior to and 90 days after CESI were included. Patients with neoplasm, fracture, infection, inflammatory spondyloarthropathy or prior cervical surgery within one year preceding CESI were excluded. Multivariable repeated measures logistic regression models were used to identify risk factors for complications at 7, 30, and 90 days. Complications were stratified into procedural and medical complications. Each time point was modeled separately with Bonferroni adjustment for the multiple time points.

Results: Complications increased from 1.7% to 4.1%, to 8.0% within 7, 30, and 90 days, respectively. Procedural complications occurred in 0.6% of cases within 7 days, with neurologic complications occurring in 0.5% of cases. Medical complications occurred in 1.1%, 3.6%, and 7.2% within 7, 30 and 90 days, respectively. After multivariable risk adjustment, a transforaminal approach was associated with greater odd for procedural complications within 7 days relative to an interlaminar approach (p<0.001), with an odds ratio (95% confidence interval) of 1.55 (1.35, 1.77). The transforaminal approach also was associated with greater odds of neurologic complications at 7 days [1.69 (1.45, 1.95); p<0.001]. A myelopathy diagnosis was not associated with risk of overall, procedural, medical, or neurological complications. The most influential factors independently associated with greater odds of any complication within 30 days included anticoagulant prescription [5.40 (4.55, 6.41); p<0.001] and history of venous thromboembolism [3.01 (2.40, 3.77); p<0.001]. The factors most strongly associated with procedural complications at 7 days included history of neurologic deficit [3.04 (2.65, 3.50); p<0.001] and antiplatelet medication prescription [2.21 (1.49, 3.27); p=0.003].

Conclusions: Complications following CESIs are infrequent. Neurologic complications were most common among procedural complications at 0.5% at 7 days. The transforaminal approach was associated with higher procedural complication rate compared to the interlaminar approach. Procedural and neurological complications were also associated with multiple factors including obesity, existing neurologic deficit, anxiety, and antiplatelet medication prescription. Attention should be given to these factors to mitigate procedural complications.