Abstracts

Rationale: Ketogenic diet (KD) therapy is an effective treatment for medication-resistant epilepsy. In a recent review1, dropout rates of 10-20% for the Classic KD (CKD) and 8-50% for the Modified Atkins Diet (MAD) were identified. We present an initial investigation of our patient population to better understand dropout rates among children treated with the KD. The number of registered dietitian (RD) interactions and other related variables were investigated to determine if there is a difference between children who followed a diet for 12 months and those who discontinued a diet before 12 months. Methods: A prospective observational cohort study of all consented children starting a KD for the treatment of medication-resistant epilepsy from September 2013 was performed at the Hospital for Sick Children, Toronto, Canada. Families received RD support outside regular clinic visits by phone, email or in person, including diet adjustment for seizures, feeding method, growth, and management of adverse events. Demographic, diet and seizure data were collected at diet initiation. Caregiver reports of adverse events, frequency of RD interactions, and seizure data were collected at follow-up. Reason for KD discontinuation was also recorded. Data were compared between children continuing KD at 12 months and those who discontinued the KD earlier. Results: Fifty children were consented for study; 9 were excluded from analysis because they did not start a KD, withdrew consent, died or experienced a prolonged hospital admission confounding data collection. The mean age was 5.39 years (SD=5.00), 24 (59%) were male, and 15 (37%) fed via gastrostomy tube. Twenty-four (59%) initiated CKD, 14 (34%) Medium-Chain Triglyceride KD, and 3 (7%) MAD. At follow-up, 24 (69%) reported ≥50% reduction in baseline seizure frequency. Twenty (49%) reported adverse events. The mean frequency of RD interactions per month was 3.00 (SD=1.62). Twenty-nine children (71%) followed a KD for 12-months, and 12 (29%) discontinued prior to 12-months. Reasons for discontinuation were medical (ineffectiveness or adverse events attributed to KD [n=6]) and non-medical (difficulty with meal preparation, food refusal, financial, and managing KD during illness [n=6]). There was no significant difference in the frequency of RD interactions per month for children who followed a KD for 12-months (M=3.06, SD=1.80) and those who discontinued (M=2.87, SD=1.15); p=0.91. Groups did not differ in age, gender, spoken language, diet type, method of initiation, feeding, adverse events or in number who reported ≥50% reduction in baseline seizure frequency. Within the group of children who discontinued KD, there was no significant difference in the frequency of RD interactions for those that stopped for medical reasons (M=3.48, SD=1.31) and non-medical reasons (M=2.25, SD=0.54); p=0.11. Conclusions: No factors were identified to distinguish those children that continued the diet for 12 months compared to those who did not. Other factors may contribute to drop out rates. While there was no statistical difference in the frequency of RD interactions between groups, there were fewer interactions in those who discontinue early due to non-medical reasons than those of medical reasons. Increased support for families managing the KD may impact drop-out rates. Further investigation is needed to better understand these findings. Funding: Ontario Brain Institute