Abstracts

Rationale: The ketogenic diet (KD) is a dietary therapy frequently used for refractory epilepsy. Unintentional carbohydrate exposure could lead to a worsening of seizure control and possibly status epilepticus, ultimately increasing overall morbidity and mortality. At Phoenix Children’s Hospital (PCH) we sought to safe guard our patients by utilizing our electronic health record (EHR). Methods: The information technology (IT) department pioneered a multilayer alert system for patients admitted to the hospital with on the KD. The EHR system is triggered by one of the following orders:  entering KD as a significant event or entering KD or ketogenic formula into diet orders. The EHR will then send alerts if a new medication or intravenous fluids (IVF) containing carbohydrates order is entered. For any such instance, an acknowledgement message is sent to the ordering provider reminding that the patient is on KD. The provider must then acknowledge this and provide a reason to proceed.  An email alert is also sent to all the epileptologists, head of pediatric hospitalist service, and the ketogenic dieticians (ketogenic diet team). The system then generates a pharmacy consult following a KD diet order or KD significant event. The medications containing carbohydrates are from a pre-formulated list continuously updated by the ketogenic dieticians/pharmacy.  If the KD is not entered as a significant event after 24 hours of it being entered as a diet order, another email alert is sent to the ketogenic diet team.Additionally, in the event of a new admission, with a prior history of documented ketogenic diet in the PCH system, an email alert is sent to the ketogenic team indicating the admission. If after 4 hours, a ketogenic diet order still not placed, a follow up email is sent. Furthermore, if a discharge order is placed it will alert the ordering provider if a ketogenic diet significant event is missing and patient was on ketogenic diet. Finally, each patient on the KD has a dashboard report of admission stating KD orders, KD significant event, contraindicated orders, all medications and ketogenic formula (if applicable). Results: Over the course of 2 years, 68 patients were admitted on the ketogenic diet on a total of 182 separate occasions, an average of 2.8 admissions per patient. Average length of admission was 14 days, median of 4 days and a range of 1-116 day. In that group, there were 151 generated alerts for possible unintended carbohydrate exposure related to medications. Of those, 63 (42%) resulted in the order being cancelled. Eighty-eight patients (58%) had the order submitted for a variety of appropriate reasons (i.e. patient given a D10 bolus for hypoglycemia with a history of GLUT1).  Conclusions: Our analysis suggests that with our alert system, 42% of unintentional carbohydrate exposures were prevented. With an increasing data base of carbohydrate containing medications, we suspect our alert system to further prevent morbidity in our ketogenic diet patients. Next, we hope to actively classify our carbohydrate alerts into unintentional and intentional carbohydrate exposure to further confirm our findings and determine if inappropriate orders are being incorrectly continued.  Funding: None