Rationale:
Epilepsy care comprises a significant amount of health care expenditure within the US, with costs attributed to both high rates of emergency department (ED) and inpatient services. Little is known about factors associated with seizure-related ED revisits for patients following an initial seizure-related ED visit. We sought to characterize the timing of ED revisits for adult patients following an initial seizure-related ED visit and associated factors.
Methods:
We conducted a retrospective cohort study of ED visits from 2018-2020 with a primary diagnosis of status epilepticus (SE), seizure, or convulsions, using the State Emergency Department Databases (SEDD) from California, New York, and Florida. We used International Classification of Diseases, 10th Revision, Clinical Modification codes G40.X for epilepsy/seizure (including SE) and R56.X for convulsions. We examined the relationship between various visit factors and primary diagnosis, and factors associated with initial visit vs revisit, identified using linkage variables within the SEDD. We used Cox proportional hazard models and cumulative incidence functions to calculate the risk of ED revisit.
Results:
In total, 364,972 patients were included with 216,567 initial presentations and 148,375 revisits. Probability of ED revisit was 8.98% (8.88-9.07) at day 7, 19.3% (19.2-19.5) at day 30, and 40.7% (40.5-40.8) by day 365. The probability of revisit was highest at all timepoints for patients with a primary diagnosis of seizure without status (10.1% [9.99-10.33] at day 7, 22.4% [22.2-22.6] at day 30, 47.2% [47-47.4] at day 365).
For ED revisits compared to initial ED visit, patients were more often younger (18-54 years old, 79% vs 71%), Black (31% vs 21%), insured by Medicaid (46% vs 33%), and had a primary diagnosis of seizure (67% vs 51%). Average age of patients who presented with a primary diagnosis of SE was 42 (IQR 29-60), seizure without status was 40 (28-55), and convulsions was 39 (28-55). Patients with a primary diagnosis of SE had higher Elixhauser Comorbidity Index (median 2, IQR [1-3] vs 1 [1-2]) as well, though there was no notable difference among initial presentations compared to revisits (1, [1-2]). Disposition of patients at initial presentation and revisit was similar, with patients predominantly being discharged home (90% of initial visits, 91% of revisits).
Cox proportional hazard ratios demonstrated a decreased rate of revisits with increasing age (≥85, HR 0.36, 95% CI [0.34-0.38]), decreased rate of revisits with private insurance compared to Medicare (0.59 [0.58-0.60]), and increased rate of revisits with primary diagnosis of seizure with or without status compared to convulsions (without status 1.58 [1.57-1.60], with status 1.46 [1.39-1.53]).
Conclusions:
Seizure-related ED utilization is a common occurrence, with nearly 41% of patients returning within one year. Younger age, Black race, and Medicaid insurance were all associated with higher revisit rates. These findings suggest this patient population may benefit from improved access and quality of epilepsy-related care to reduce ED utilization
Funding: N/A