Introduction: Drug-resistant epilepsy, characterized by the failure of two or more appropriate anti-epileptic medications to achieve seizure freedom, poses significant treatment challenges. Among the neuro-modulating devices used in current epilepsy treatments, vagal nerve stimulation (VNS) is thought to affect cortical activity via projections from the nucleus tractus solitarius to other brainstem nuclei and the cortex. VNS has been FDA-approved since 1997.

Objective: This study aims to analyze the impact of VNS implantation on the frequency of seizure-related emergency department (ED) visits at 1, 2, and 3 years post-implantation.

Materials & Methods: A single-center retrospective analysis was conducted on patient charts diagnosed with epilepsy and implanted with a VNS device between 2010 and 2023. Of the 128 charts reviewed, 70 were excluded due to missing data, lack of follow-up, or concurrent treatments such as temporal lobectomy. The final analysis included 58 patients. The study compared the number of ED visits one year pre-implantation to those at one, two, and three years post-implantation using the Wilcoxon signed-rank test (p<0.05). Descriptive analyses of quantitative and categorical data were also performed.

Results: The average age of epilepsy onset was 16.5 ± 12.4 years, with females constituting 50% of the patient population. The median age at VNS implantation was 36 ± 12.2 years, and the median number of medications at implantation was three. Generalized onset seizures were observed in 41% of patients, focal onset in 22.4%, and both in 6.9%. No significant difference in ED visits was noted between one year pre-implantation and the first year post-implantation (p=0.24) or the second year post-implantation (p=0.06). However, a significant reduction in ED visits was observed in the third year post-implantation compared to baseline (p=0.049). Of the patients, 39 had no ED visits in the year prior to VNS implantation, and most continued to have no ED visits in the subsequent year. Among the 11 patients with at least two ED visits in the year before VNS, most experienced an improvement post-implantation.

Conclusion: While no statistical difference in ED visits was observed at one and two years post-VNS implantation, a significant decrease was noted at three years. Patients with two or more ED visits prior to implantation showed significant reductions in ED visits one year post-implantation. The absence of ED visits in a subset of patients raises questions about potential data voids due to follow-ups in other health systems.

The study faced limitations due to a lack of documentation on patients seeking care in other health systems and a high loss to follow-up, reducing the study's power. Future research should consider factors such as VNS parameters, age of epilepsy onset, and long-term follow-up data, especially after the introduction of the auto-stimulation function in VNS devices.