AES Guideline: Treatment of Convulsive Status Epilepticus (2016)
Evidence-based guideline from the American Epilepsy Society (AES), published in Epilepsy Currents (2016), reviewing 38 RCTs on pharmacologic treatment of convulsive status epilepticus in children and adults.
🔹 Bottom Line
- Guideline: AES 2016 evidence-based guideline — 38 RCTs reviewed for CSE treatment in children and adults
- First-line (Level A): BZDs are the initial therapy of choice — IM midazolam, IV lorazepam, or IV diazepam; all equivalently efficacious
- No IV access (Level A): IM midazolam is superior to IV lorazepam (RAMPART: 73% vs 63%, p < 0.001)
- Second-line (Level U): No evidence-based preferred agent — fosphenytoin, valproic acid, or levetiracetam as single dose
- Efficacy drops sharply: VA Cooperative — 1st drug 55.5% → 2nd drug 7.0% → 3rd drug 2.3%
- Fosphenytoin > phenytoin on tolerability (Level B); phenytoin is an acceptable alternative
- Respiratory depression is LOWER with BZDs than with placebo — untreated SE is more dangerous (Level A)
- Children: Non-IV routes (rectal diazepam, buccal/intranasal midazolam) are probably effective (Level B)
Guideline Overview
Source & Scope
- Organization: American Epilepsy Society (AES) Guideline Committee
- Published: Epilepsy Currents, Vol. 16, No. 1, January/February 2016
- Authors: Glauser T, Shinnar S, Gloss D, Alldredge B, Arya R, Bainbridge J, et al.
- Evidence base: 38 RCTs identified — 4 class I, 2 class II, 32 class III
- Scope: Pharmacologic treatment of convulsive SE in children and adults; seizures ≥5 min
- NOT addressed: Refractory SE, neonatal SE, NCSE, etiology-specific therapy, surgery, ketogenic diet, neurostimulation
Key Definitions
- Status epilepticus: Seizure lasting ≥5 min (operational definition for treatment initiation)
- Traditional SE: ≥30 min continuous seizure activity → risk of permanent neuronal injury
- Efficacy: Ability of drug to stop CSE
- Tolerability: Incidence, severity, and impact of anticonvulsant-related adverse effects
- Effectiveness: Encompasses both efficacy and tolerability
- Safety: Life-threatening adverse events
Evidence Classification
| Level | Evidence Required | Conclusion | Recommendation |
|---|
| A | ≥1 class I study or ≥2 consistent class II | Established as effective/ineffective/harmful | Should be done / should not be done |
| B | ≥1 class II study or ≥3 consistent class III | Probably effective/ineffective/harmful | Should be considered / should not be considered |
| C | ≥2 consistent class III studies | Possibly effective/ineffective/harmful | May be considered / may not be considered |
| U | Insufficient evidence | Data inadequate or unproven | None |
Q1: Which Anticonvulsants Are Efficacious?
Adults — Initial Therapy
| Agent | Route | Evidence Level | Key Data |
|---|
| Midazolam | IM | Level A | RAMPART: 73% seizure cessation; superior to IV LZP without IV access |
| Lorazepam | IV | Level A | VA Cooperative: 59.1% vs placebo 21.1% (OR 4.8); may repeat ×1 |
| Diazepam | IV | Level A | VA Cooperative: 42.6% vs placebo 21.1% (OR 2.3); may repeat ×1 |
| Phenobarbital | IV | Level A | VA Cooperative: efficacious but slower administration rate |
- No difference in efficacy between IV lorazepam and IV diazepam (Level A)
- IM midazolam has superior effectiveness compared with IV lorazepam in adults without established IV access (Level A)
- IV lorazepam is more effective than IV phenytoin at stopping seizures ≥10 min (Level A)
Adults — Second-Line Therapy
| Agent | Dose | Max | Evidence Level |
|---|
| Fosphenytoin | 20 mg PE/kg | 1500 mg PE | Level U |
| Valproic acid | 40 mg/kg | 3000 mg | Level B |
| Levetiracetam | 60 mg/kg | 4500 mg | Level U |
| Phenobarbital | 15 mg/kg | — | Level B (alternative if above unavailable) |
- No evidence that any second-line option is better than the others
- IV valproic acid has similar efficacy to IV phenytoin (Level B) with better tolerability
- Insufficient data on levetiracetam as second therapy in adults (Level U)
- All second-line agents should be given as a single full dose
Children — Initial Therapy
| Agent | Route | Evidence Level | Notes |
|---|
| Lorazepam | IV | Level A | 0.1 mg/kg/dose, max 4 mg; may repeat ×1 |
| Diazepam | IV | Level A | 0.15–0.2 mg/kg/dose, max 10 mg; may repeat ×1 |
| Diazepam | Rectal | Level B | 0.2–0.5 mg/kg, max 20 mg |
| Midazolam | IM | Level B | 10 mg for >40 kg, 5 mg for 13–40 kg |
| Midazolam | Intranasal | Level B | Non-IV midazolam probably more effective than diazepam (IV/rectal) |
| Midazolam | Buccal | Level B | Faster time to seizure cessation vs rectal diazepam |
- No difference between IV lorazepam and IV diazepam in children (Level A)
- Insufficient data on intranasal lorazepam, sublingual lorazepam, rectal lorazepam, valproic acid, levetiracetam, and phenytoin as initial therapy in children (Level U)
🔹 Clinical Pearl
The #1 treatment error in SE is BZD underdosing. Give the full guideline-recommended dose as a single administration. Initial therapy should NOT be broken into multiple smaller doses (except IV lorazepam and diazepam which can be repeated once).
Q2: Adverse Events
Adults (Level A)
- Most common adverse events: Respiratory and cardiac symptoms with IV anticonvulsant administration
- VA Cooperative data:
| Agent | Hypoventilation | Hypotension | Cardiac Rhythm |
|---|
| Lorazepam | 10.3% | 25.8% | 7.2% |
| Diazepam | 16.8% | 31.6% | 2.1% |
| Phenobarbital | 13.2% | 34.1% | 3.3% |
| Phenytoin | 9.9% | 27.0% | 6.9% |
| Placebo (untreated) | 22.5% treatment-emergent adverse events |
- Critical finding: Respiratory depression is LOWER in BZD-treated patients than placebo-treated patients (Level A)
- This confirms that respiratory problems are a consequence of untreated SE, not of BZD treatment
- No substantial difference in cardiorespiratory adverse events between BZDs and phenobarbital (Level A)
RAMPART Adverse Events (IM Midazolam vs IV Lorazepam)
- Treatment-emergent AEs: IM midazolam 26.7% vs IV lorazepam 30.6%
- Decreased consciousness: MDZ 9.5% vs LZP 8.8%
- Respiratory depression: MDZ 6.4% vs LZP 10%
- Hypotension: only 1.2% overall
- No significant differences in adverse events between the two agents
Children
- Respiratory depression is the most common clinically significant adverse event (Level B)
- No substantial difference between midazolam, lorazepam, and diazepam by any route for respiratory depression rates (Level B)
- Adverse events with BZDs reported less frequently in children than adults (Level B)
- Respiratory depression after rectal diazepam in children: 1.2–6.4% across class III studies
- Buccal midazolam: no respiratory depression reported in 2 class III pediatric studies
🔹 Clinical Pearl
Untreated SE causes MORE respiratory depression than BZD treatment. The VA Cooperative trial showed placebo-treated patients had a 22.5% adverse event rate vs ~10% with lorazepam. Never withhold BZDs due to fear of respiratory depression.
Q3: Most Effective Benzodiazepine
Adults (Level A)
- No significant difference in effectiveness between IV lorazepam and IV diazepam
- IM midazolam is established as more effective than IV lorazepam when IV access is not available (Level A)
- Pharmacokinetic advantage: lorazepam has a longer duration of action (but not longer half-life) than diazepam
RAMPART Trial (2012) — Class I
| IM Midazolam | IV Lorazepam | Difference |
|---|
| n | 448 | 445 | — |
| Seizure cessation | 73.4% | 63.4% | 10% (95% CI: 4.0–16.1; p < 0.001) |
| Median time to cessation | 1.6 min from active Tx | 4.8 min from active Tx | IM faster due to quicker delivery |
| Result | IM MDZ met noninferiority AND demonstrated superiority |
Children (Level A & B)
- No significant difference between IV lorazepam and IV diazepam (Level A)
- Non-IV midazolam (IM/intranasal/buccal) is probably more effective than diazepam (IV/rectal) in children (Level B)
- Meta-analysis of 6 pediatric studies: non-IV midazolam vs IV/rectal diazepam → RR for seizure cessation = 1.52 (95% CI: 1.27–1.82) favoring midazolam
- Similar respiratory complication rates (RR 1.49; 95% CI: 0.25–8.72)
- Time to seizure cessation: shorter for intranasal midazolam vs IV diazepam in 2 studies
🔹 Clinical Pearl
IM midazolam is preferred first-line when IV access is not established (prehospital, seizing patient). Its superiority in RAMPART was driven by faster drug delivery, not greater pharmacologic potency. IV lorazepam and IV diazepam remain equivalent when IV access is available.
Q4: IV Fosphenytoin vs IV Phenytoin
- Insufficient data to compare efficacy of phenytoin vs fosphenytoin (Level U)
- Fosphenytoin is better tolerated than phenytoin (Level B)
- Phenytoin is an acceptable alternative when fosphenytoin is unavailable (Level B)
Tolerability Comparison (3 Class III RCTs)
| Adverse Effect | Fosphenytoin | Phenytoin |
|---|
| Pain at infusion site | 17% | Greater (not quantified) |
| Pruritus at infusion site | 48.6% | Lower |
| Cardiac arrhythmias | None reported | Present |
| BP changes | 13.7 mmHg decrease | 5.9 mmHg decrease |
| Infusion slowed/discontinued | Less frequent | 63.6% experienced issues |
- No fosphenytoin-related cardiac arrhythmias, heart rate changes, respiratory changes, or BP changes in tolerability study
- Fosphenytoin can be infused faster (150 mg PE/min vs 50 mg/min for phenytoin)
Q5: When Does Efficacy Drop? (Refractory SE)
VA Cooperative Trial (1998) — Class I
- Only class I RCT to address sequential therapy efficacy
- 570 adults with overt CSE; 4-arm, double-blind, sequential design
- If 1st drug failed → randomized to 2nd drug; if 2nd failed → 3rd drug
| Therapy Phase | Success Rate | Interpretation |
|---|
| 1st anticonvulsant | 55.5% | Initial BZD therapy — best chance of success |
| 2nd anticonvulsant | 7.0% | Dramatic drop — highlights urgency of adequate 1st-line dosing |
| 3rd anticonvulsant | 2.3% | Virtually ineffective; consider anesthetics at this stage |
- Second-line RCTs (non-VA) reported higher success: 77–90% (class II) and 50–88% (class III)
- However, these studies were not blinded and initial therapy was not part of the RCT
- No clear evidence to guide third-therapy phase (Level U)
- Third therapy options: repeat second-line agent, anesthetic doses of thiopental/midazolam/pentobarbital/propofol (all with cEEG)
🔹 Clinical Pearl
The VA Cooperative trial is the only class I study showing sequential efficacy decline: 55.5% → 7.0% → 2.3%. This dramatically demonstrates why adequate first-line BZD dosing is critical — each subsequent drug is substantially less effective. The study also showed that lorazepam was the only agent superior to phenytoin head-to-head (p = 0.001).
AES Treatment Algorithm
Phase 1: Stabilization (0–5 min)
- ABCs: airway, breathing, circulation, disability — neurologic exam
- Time seizure from its onset; monitor vital signs
- Oxygen via nasal cannula/mask; consider intubation if respiratory assistance needed
- Initiate ECG monitoring
- Finger-stick blood glucose: if <60 mg/dL →
- Adults: thiamine 100 mg IV then D50W 50 mL
- Children ≥2 years: D25W 2 mL/kg IV
- Children <2 years: D12.5W 4 mL/kg
- Attempt IV access; collect electrolytes, hematology, toxicology, AED levels
Phase 2: Initial Therapy (5–20 min)
A benzodiazepine is the initial therapy of choice (Level A). Choose ONE of 3 equivalent first-line options:
| Agent | Route | Dose | Level | Repeat? |
|---|
| Midazolam | IM | 10 mg (>40 kg); 5 mg (13–40 kg) | A | Single dose |
| Lorazepam | IV | 0.1 mg/kg/dose, max 4 mg | A | May repeat ×1 |
| Diazepam | IV | 0.15–0.2 mg/kg/dose, max 10 mg | A | May repeat ×1 |
If the 3 above options are not available, choose one of:
| Agent | Route | Dose | Level |
|---|
| Phenobarbital | IV | 15 mg/kg, single dose | A |
| Diazepam | Rectal | 0.2–0.5 mg/kg, max 20 mg, single dose | B |
| Midazolam | Intranasal | Level B | B |
| Midazolam | Buccal | Level B | B |
Phase 3: Second Therapy (20–40 min)
No evidence-based preferred second therapy of choice (Level U). Choose ONE, give as a single dose:
| Agent | Dose | Max | Level |
|---|
| Fosphenytoin | 20 mg PE/kg | 1500 mg PE | U |
| Valproic acid | 40 mg/kg | 3000 mg | B (one class II study) |
| Levetiracetam | 60 mg/kg | 4500 mg | U |
- If none of the above are available: IV phenobarbital 15 mg/kg (if not already given) — Level B
- ESETT trial (published 2019, after this guideline) later confirmed all three are equivalent (~46–47% efficacy)
Phase 4: Third Therapy (40–60 min)
- No clear evidence to guide therapy in this phase (Level U)
- Options include:
- Repeat second-line therapy
- Anesthetic doses of thiopental, midazolam, pentobarbital, or propofol
- All with continuous EEG monitoring
- Consider escalating directly to anesthetics for severe/prolonged cases
🔹 Clinical Pearl
Key algorithm principles: (1) Give BZDs as a single adequate dose, not multiple smaller doses. (2) All second-line agents are dosed by weight with a maximum cap. (3) The ESETT trial (2019, post-guideline) validated the equivalent efficacy of fosphenytoin, VPA, and LEV as second-line agents at ~46–47% success rate each. (4) There is no evidence to support one third-line agent over another.
Landmark Trials Referenced
| Trial | Year | Class | Key Finding |
|---|
| VA Cooperative | 1998 | I | 4-arm RCT: LZP best initial agent; sequential efficacy drops 55.5% → 7.0% → 2.3% |
| Prehospital SE (Alldredge) | 2001 | I | IV LZP 59.1% vs IV DZP 42.6% vs placebo 21.1%; both superior to placebo |
| RAMPART | 2012 | I | IM MDZ 73.4% vs IV LZP 63.4%; noninferior + superior; prehospital |
| PECARN (Chamberlain) | 2014 | I | Pediatric: IV LZP 72.1% vs IV DZP 72.9%; no difference |
| LZP vs DZP Comparative | 1983 | II | LZP 78% vs DZP 58% after single dose; DZP 76% after two doses — NS difference |
Summary of Recommendations by Evidence Level
Level A Recommendations (Established)
- In adults: IM midazolam, IV lorazepam, IV diazepam, and IV phenobarbital are efficacious for CSE ≥5 min
- IM midazolam > IV lorazepam when no IV access (adults)
- IV lorazepam > IV phenytoin for stopping seizures ≥10 min
- No difference between IV lorazepam and IV diazepam (adults and children)
- Respiratory and cardiac symptoms are the most common adverse events with IV anticonvulsants
- BZD-treated patients have LESS respiratory depression than placebo (untreated SE)
- In children: IV lorazepam and IV diazepam are efficacious for CSE ≥5 min
Level B Recommendations (Probable)
- In children: rectal diazepam, IM midazolam, intranasal midazolam, buccal midazolam are probably effective
- IV valproic acid has similar efficacy to IV phenytoin as second-line with better tolerability
- Fosphenytoin is better tolerated than phenytoin; phenytoin is an acceptable alternative
- Adverse events with BZDs are less frequent in children than adults
Level U (Insufficient Evidence)
- Levetiracetam as initial or second therapy
- Intranasal/sublingual/rectal lorazepam, valproic acid, phenytoin as initial therapy in children
- Comparative efficacy of fosphenytoin vs phenytoin
- Third-therapy phase — no evidence to guide agent selection
