Abstracts

Rationale: Epilepsy affects approximately 15,000 children in the province of Ontario, Canada. Approximately 20-30% of the patients are intractable and the high-fat, low-carbohydrate ketogenic diet (KD) has been increasingly used in this group. Studies have shown that the diet is effective in decreasing the number of seizures in children1, however, little is known about the effect of KD on health care utilization patterns. The purpose of this study was to compare health care utilization patterns in Ontario children before and after initiation of KD between 2000 and 2010. 3 tertiary care centres (The Children’s Hospital of Eastern Ontario (CHEO, Ottawa), Sick Kids (Toronto) and McMaster Children’s Hospital (Hamilton)) utilized the diet at that time. Therefore the data represents a snapshot for the majority of Ontario pediatric patients initiated on diet between 2000 and 2010.Methods: Following Research Ethics Board approval, retrospective patient data was collected from patient charts for patients initiated on KD between January 1, 2000 and December 31, 2010. Demographic variables are represented in Table 1. Data was entered into a secure, web-based REDCap (Research Electronic Data Capture) database by study coordinators at each of the three sites. Once cleaned, data was linked to health administrative data from several Ontario-wide databases available to ICES (Institute for Clinical Evaluative Sciences), capturing the patients’ contact with the Ontario healthcare system across time. The ICES data that was of particular interest involved Emergency and Inpatient admissions. The following databases were used for the analysis: Discharge Abstract Database (DAD), Registered Persons Database (RPDB), and ERCLAIM. In the case that patients initiated diet more than once in the study time frame, only the earliest initiation was counted (Figure 1).Results: 205 patients were entered in REDcap but 35 patients did not meet eligibility criteria leaving 170 for the analysis. For patients with at least one ER record in the observation period, there was a statistically significant decrease in the mean number of ER records per day post diet initiation. More specifically, there was an average decrease of 78 ER records/10,000 person days (2 sided t-test: p<0.0001, 95% CI: 40– 100) post KD compared to the pre diet time period. There was also a statistically significant decrease in number of inpatient episodes per day following diet initiation. On average, there were 33 fewer episodes per 10,000 person days (2 sided t-test: p<0.0001, 95% CI:20 – 50) after KD start. In addition, there was a significant decrease in the average length of stay (LOS) per episode post diet. The average LOS decreased by 1.9 days post diet (2 sided t-test: p<0.04, 95% CI:0.081 –3.7).Conclusions: In the province of Ontario, the KD is effective in decreasing the overall health care burden, as measured by ER records, hospital admissions and LOS.