Abstracts

Rationale: RSE in children often requires medical induction of coma for management. Pentobarbital is widely used for this purpose, but has a long half-life, and can cause cardiac depression. Several recent reports suggest ketamine is a safe and effective treatment for RSE. It is unknown if use of ketamine for RSE has increased. Furthermore, outcomes after treatment for RSE with ketamine are understudied. We conducted a retrospective cohort study of children with RSE, comparing treatment with ketamine versus treatment with pentobarbital but not ketamine.Methods: The PHIS database includes demographics, diagnostic codes, and daily billing information for procedures, laboratory tests, and medications for children at 45 children’s hospitals in North America. From this database, we created a list of candidate visits representing children who may have been treated with either ketamine or pentobarbital (or both) for the treatment of RSE. The criteria were: age 0–21 years, any diagnosis of 345.x (epilepsy) OR 780.39 (convulsions), at least 2 day stay in the pediatric ICU, at least two days with a charge for pentobarbital OR ketamine, at least one day on a ventilator, and discharge date from Jan 2010 through Sept 2014. From these, we manually abstracted charts from two institutions to determine which children 1) had RSE, 2) were treated with pentobarbital for RSE, and 3) were treated with ketamine for RSE. Through discussion and review of visualizations of the PHIS data, we created and validated preliminary definitions for two groups: 1) “RSE ketamine” is RSE treated with ketamine and/or pentobarbital and 2) “RSE pentobarbital” is RSE treated with pentobarbital but not ketamine. We applied these definitions to the remaining candidate patients.Results: At our two centers, the definition of “RSE ketamine” (two consecutive days of ketamine use, EEG, and endotracheal ventilation) had a positive predictive value (PPV) of 75% and sensitivity of 100%; and “RSE pentobarbital” (two consecutive days of pentobarbital use, EEG, and endotracheal ventilation) had a PPV of 82% and sensitivity of 93%. These definitions identified 630 “RSE pentobarbital” cases and 81 “RSE ketamine” cases. Demographics were similar between the groups. Children in the “RSE ketamine” group required more days in the ICU, on EEG, and on the ventilator. In-hospital mortality was similar between the groups. (Table 1) In 2014, significantly more of the 45 PHIS hospitals used ketamine for RSE (5 of 45 in 2010 vs 14 of 45 through Sept 2014). Furthermore, the number of cases per year in the RSE ketamine group grew from 6 to 35 (2011 to 2014 estimated), while the number in the RSE pentobarbital group dropped from 159 to 121. (Table 2)Conclusions: Children treated with ketamine have longer, more intensive hospital stays, which is consistent with our own practice to reserve ketamine for the most challenging cases. From 2010–2014, ketamine was used for RSE at more hospitals and for more children. Further research is indicated to understand safety and clinical outcomes of ketamine for RSE.