Efficacy and Risk of Respiratory Depression with Rectal Diazepam Use in Children with Epilepsy

Rationale: Rectal diazepam (0.5 mg/kg-maximum single dose 20 mg) is our standard practice for terminating prolonged (seizure > 5 minutes and continuing) or acute repetitive seizures in children, both inpatient and at home. Recently, clinically significant respiratory depression, requiring mechanical support has been reported in 8.8% of children receiving rectal diazepam (Norris et al, Dev. Med. Child Neurol 1999;41:340-343). This is a major concern, especially when prescribed for home use. We analyzed all uses of rectal diazepam on our pediatric epilepsy unit since 1998 for efficacy and risk of respiratory depression. Methods: We follow a strict hospital protocol. For at least 20 minutes following the administration of rectal diazepam, there is continuous observation and respiratory rate is recorded every five minutes. Charts were reviewed for age, weight, dose, seizure response, respiratory rate. Results: 532 doses of rectal diazepam were administered to 78 children ages 3 mos. to 20 years (mean age 6.75 years). The average dose was 0.6 mg/kg. Single doses ranged from 0.3 to 1.3 mg/kg. Those receiving less than 0.5 mg/kg weighed over 40 kg with our maximum single dose 20 mg. Seven children had repeat doses within 10 minutes, 8 had 3 or more doses in < 24 hours. Rectal diazepam terminated prolonged seizures or acute repetitive seizures in 96% (511/532) of administrations. There were no (0/532) episodes of clinical respiratory depression. Conclusions: We reviewed only inpatients because we had protocol documentation of effectiveness and respiratory status on every use of rectal diazepam. This eliminates the possibility of an unreported case of respiratory depression. Rectal diazepam was 96% effective. No clinically significant respiratory depression was seen in 532 uses. We have prescribed rectal diazepam for over 10 years, giving an estimated 1500-2000 doses to over 300 children without a reported case of respiratory depression. Our experience is quite different from that reported by Norris et al. Possible explanations for these differences will be discussed.