Review: lorazepam provides the best control for status epilepticus.

and commentary also appear in ACP Journal Club and a modified version of the abstract appears in Evidence-Based Nursing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . For correspondence: Dr K Prasad, All India Institute of Medical Sciences, New Delhi, India. [email protected] Source of funding: no external funding. Anticonvulsant drugs for status epilepticus to hospital discharge* Outcomes Number of trials (n) Comparisons Event rates RRR (95% CI) NNT (CI) Non-cessation of seizures 3 (264) Lorazepam v diazepam 24% v 38% 36% (10 to 55) 8 (5 to 25) 1 (137) Lorazepam v placebo 41% v 79% 48% (29 to 62) 3 (2 to 5) 1 (198) Lorazepam v phenytoin 35% v 56% 38% (14 to 55) 5 (3 to 13) 1 (139) Diazepam v placebo 57% v 79% 27% (8 to 43) 5 (3 to 17) 2 (165) Intrarectal diazepam gel v placebo 32% v 72% 57% (38 to 70) 3 (2 to 4) 1 (39) Intrarectal diazepam gel 30 mg v 20 mg 28% v 71% 61% (14 to 82) 3 (2 to 7) Continuation of status epilepticus requiring a different drug 3 (264) Lorazepam v diazepam 24% v 39% 37% (12 to 55) 7 (4 to 25) 1 (137) Lorazepam v placebo 41% v 79% 48% (29 to 62) 3 (2 to 5) 1 (139) Diazepam v placebo 57% v 79% 27% (8 to 43) 5 (3 to 17) Death 1 (139) Diazepam v placebo 4.4% v 15% 72% (2 to 92) 10 (5 to 100) Ventilatory support 1 (139) Diazepam v placebo 8.8% v 23% 61% (6 to 84) 8 (4 to 50) *Abbreviations defined in glossary; weighted event rates, RRR, NNT, and CI calculated from data in article using a fixed effects model. All drugs given intravenously unless otherwise noted. Event rates with 1 trial are unweighted. Commentary S E is a neurological emergency with a 30 day mortality rate of about 22%, contingent on duration before treatment, underlying cause, and patient age. Prasad et al have attempted to determine which initial pharmacological treatment for SE is best in terms of rapidity of action, maintenance of efficacy, and incidence of adverse events. Most of the studies enrolled patients with ‘‘premonitory SE,’’ which, while not meeting the criteria for ‘‘established SE,’’ is generally thought to be a condition best addressed early and aggressively. Their results affirm the consensus of standard clinical practice, but underscore the diversity that exists among investigator definitions of SE and outcome measures. Their strongest conclusion, that lorazepam is more effective than diazepam or phenytoin, reinforces guidelines published .10 years ago, matches the preferences of surveyed neurologists, and is in turn buttressed by the theoretical pharmacokinetic advantages of lorazepam. The review shows that any of the agents investigated perform better than placebo regardless of administration route, although routes were not a focus of study. Despite this lack of comparative data, we recommend IV formulations when available, and rectal formulations when IV is not feasible—reserving the intramuscular route as a last resort. This review also does not address what to do when initial treatments fail, but a related review concludes that continuous IV pentobarbital, titrated to electroencephalographic background suppression, produces the most favourable results. Prasad et al highlight the need for further RCTs that use a standardised approach to the classification of SE, the dosing and route of compared agents, and common outcome measures. J Craig Henry, MD Robert Holloway, MD, MPH University of Rochester Medical Center Rochester, New York, USA 1 DeLorenzo RJ, Hauser WA, et al. Neurology 1996;46:1029–35. 2 Treatment of convulsive status epilepticus. Recommendations of the Epilepsy Foundation of America’s Working Group on Status Epilepticus. JAMA 1993;270:854–9. 3 Claassen J, Hirsch LJ, Mayer SA. J Neurol Sci 2003;211:37–41. 4 Claassen J, Hirsch LJ, Emerson RG, et al. Epilepsia 2002;43:146–53. 54 THERAPEUTICS www.evidence-basedmedicine.com EBM Volume 11 April 2006