Comparison of long-term survival with continued medical therapy, vagus nerve stimulation, and cranial epilepsy surgery in paediatric patients with drug-resistant epilepsy in the USA: an observational cohort study

Background Long-term survival in paediatric epilepsy is incompletely characterised. A better understanding of treatment effects on mortality in paediatric patients with drug-resistant epilepsy is needed for health-care decision making. We aimed to compare the long-term survival rates associated with antiseizure medications only, antiseizure medications plus vagus nerve stimulation (VNS), and antiseizure medications plus cranial epilepsy surgery in paediatric patients with drug-resistant epilepsy using a large national administrative database in the USA. Methods In this observational cohort study, patients aged 0–17 years who were diagnosed with drug-resistant epilepsy using International Classificiaton of Diseases codes between Jan 1, 2004, and Dec 31, 2020, were identified from the Pediatric Health Information System, an administrative database that contains inpatient, emergency department, ambulatory, and observation unit encounter-level data from more than 49 children's hospitals in the USA. Patients treated with at least three types of antiseizure medications were included in the medical therapy cohort, those treated with antiseizure medications plus VNS were included in the VNS cohort, and those treated with antiseizure medications plus cranial epilepsy surgery were included in the surgery cohort. Participants were followed up until the date of their last clinical encounter, in-hospital death, or Dec 31, 2020. Inverse probability of treatment weighting (IPTW) was used to balance baseline demographics and clinical characteristics between treatment groups. The unconditional probabilities of survival were estimated by weighted Kaplan-Meier analysis. A weighted Cox proportional hazards model was used to investigate the association between risk of overall death and age, sex, geographical region, race and ethnicity, comorbidity, primary diagnosis, insurance, and treatment. Findings This study included 10 240 patients treated with antiseizure medications only, 5019 patients treated with antiseizure medications plus VNS, and 3033 patients treated with antiseizure medications plus cranial epilepsy surgery. The median age of paediatric patients was 7 years (IQR 4–12) in the medical therapy cohort, 9 years (6–13) in the VNS cohort, and 9 years (5–13) in the surgery cohort. The IPTW-adjusted probabilities of surviving beyond 10 years were 89·27% (95% CI 87·71–90·85) for the medical therapy cohort, 92·65% (90·62–94·72) for the VNS cohort, and 98·45% (97·53–99·38) for the surgery cohort. The difference in survival probabilities was significant (log-rank p<0·0001). Compared with the medical therapy cohort, the IPTW-adjusted hazard ratio for overall death was 0·60 (95% CI 0·50–0·74) for the VNS cohort and 0·19 (0·10–0·33) for the surgery cohort. Interpretation Paediatric patients with drug-resistant epilepsy who underwent cranial epilepsy surgery or VNS had a higher survival rate than those who received only medical treatment. These findings highlight the importance of a multidisciplinary comprehensive team approach to the treatment of epilepsy, which includes tailored evaluation and deployment of medical and surgical treatment options for patients with this challenging disease. Funding None.