C. Stillman, K. Knupp, J. Oliver, A. Conley, K. Kennedy, L. Silveira, et al. (2025). A retrospective analysis of pediatric patients on a ketogenic diet: A comparison of inpatient versus outpatient diet initiations. Epilepsy research, 214, 107556. https://doi.org/10.1016/j.eplepsyres.2025.107556

PubMed 40315550

[Abstract]
INTRODUCTION: Ketogenic diet therapies are effective therapies for drug-resistant epilepsy. Conventional initiation of the ketogenic diet occurs via inpatient (IP) admission to a hospital. The COVID19 pandemic forced changes to practices allowing for comparison between inpatient (IP) and outpatient (OP) initiations. Our aim was to evaluate differences between IP and OP initiations including laboratory results, seizure reduction and communications with patients.

METHODS: This is a retrospective chart review of patients initiated on a ketogenic diet (modified Atkins [MAD] or classic ketogenic [CKD]) between 2007 and 2022. We compared variables such as demographic data, communications, lab values, seizure counts, IP or OP initiation, presence of a gastrostomy tube (g-tube), and diet type.

RESULTS: Of the 157 total subjects, 139 subjects initiated CKD and 18 subjects initiated MAD. 39 initiated OP and 118 initiated IP. The odds of a 50 % reduction in seizures at 65 days post initiation increased four times for IP initiation after adjusting for the impact of serum beta hydroxybutyrate (BHB). This difference was no longer present at 196 days post initiation. Number of communications between diet initiation and the first visit post initiation were similar for IP and OP. G-tube presence or absence did not impact outcomes.

CONCLUSION: IP initiation resulted in better seizure control at the first visit post initiation. CKD was the only variable associated with increased communications. Since seizure improvement rates were similar at 196 days, a gradual approach with lower CKD ratios may be considered. G-tube presence had no impact on outcomes and should be weighted less when considering admission.