Alldredge SE
(2001)Objective
To compare intravenous lorazepam, diazepam, and placebo administered by paramedics for out-of-hospital status epilepticus.
Study Summary
• Lorazepam vs diazepam was not statistically different (p=0.08)
• Respiratory/circulatory complications were paradoxically fewer with benzodiazepines (10.6% lorazepam, 10.3% diazepam) than placebo (22.5%), and ICU admission was lower with treatment
Intervention
IV lorazepam 2 mg vs IV diazepam 5 mg vs IV placebo, each repeatable once after 4 min; N=205
Inclusion Criteria
Adults ≥18 years with convulsive seizures ≥5 min (or ≥3 seizures without recovery) and obtainable field IV access.
Study Design
Arms: Lorazepam vs Diazepam vs Placebo (control)
Patients per Arm: Lorazepam: 66, Diazepam: 68, Placebo: 71
Outcome
• Respiratory/circulatory complications: 10.6% vs 10.3% vs 22.5%
• ICU admission: 10.6% vs 9.6% vs 27.5% (p<0.05)
• Lorazepam NNT ~3; diazepam NNT ~5 vs placebo
Bottom Line
Both IV lorazepam and diazepam were significantly more effective than placebo for prehospital SE: seizure termination 59.1% (lorazepam) vs 42.6% (diazepam) vs 21.1% (placebo; P<0.001 for both vs placebo). Lorazepam was the most effective. Paradoxically, benzodiazepine-treated patients had better outcomes on ED arrival — shorter SE duration, lower intubation rate, and shorter ICU stay — disproving concerns about prehospital respiratory depression.
Major Points
- Seizure termination before ED: lorazepam 59.1% vs diazepam 42.6% vs placebo 21.1% (P<0.001 for both drugs vs placebo).
- Lorazepam vs diazepam: 59.1% vs 42.6% (P=0.08, NS) — lorazepam numerically but not significantly superior.
- Respiratory/circulatory complications: lorazepam 10.6%, diazepam 10.3%, placebo 22.5% — paradoxically FEWER with treatment.
- ICU admission: lorazepam 10.6%, diazepam 9.6%, placebo 27.5% — treated patients had better ED disposition.
- SE duration from treatment to termination: lorazepam/diazepam ~5 min vs placebo ~25 min in ED.
- Double-blind, randomized, placebo-controlled. 205 patients. San Francisco paramedics, 1994-1998.
- Paramedic administration was safe and feasible — established paradigm for prehospital SE treatment.
- Lorazepam dose: 2mg IV; diazepam dose: 5mg IV. Repeated once if seizure continued after 4 min.
- Only included adults ≥18 years with convulsive seizures >5 min. Excluded known pregnancy.
- Landmark trial: led to RAMPART (IM midazolam) and established prehospital benzodiazepine as standard of care.
Study Design
- Study Type
- Randomized, double-blind, placebo-controlled trial
- Randomization
- Yes
- Blinding
- Double-blind. Identical prefilled syringes in sequentially numbered kits.
- Sample Size
- 205
- Follow-up
- Hospital discharge
- Centers
- 1
- Countries
- United States
Primary Outcome
Definition: Termination of SE before ED arrival
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| 15/71 (21.1%) | Lorazepam 39/66 (59.1%); Diazepam 29/68 (42.6%) | - (Lorazepam vs placebo P<0.001; Diazepam vs placebo P=0.003) | <0.001 overall |
Limitations & Criticisms
- Single-center (San Francisco) — may not generalize to all EMS systems.
- Only IV route studied — IM route (later studied in RAMPART) may be more practical.
- Low single doses (LZP 2mg, DZP 5mg) — may underestimate efficacy of higher doses.
- Placebo arm: ethical concerns about withholding treatment in SE (justified by equipoise at time).
- Small sample (205) — underpowered for lorazepam vs diazepam comparison.
- Adults only — pediatric prehospital SE not addressed.
- Short follow-up (hospital discharge only).
Citation
N Engl J Med 2001;345:631-637.