Phenobarbital for Acute Alcohol Withdrawal: Prospective Randomized Double-blind Placebo-controlled Study.

Rosenson et al. J Emerg Med. 2013 Mar;44(3):592-598

Objective	To investigate if a single dose of intravenous (i.v.) phenobarbital combined with a standardized lorazepam-based alcohol withdrawal protocol decreases intensive care unit (ICU) admission in ED patients with acute alcohol withdrawal
Design	Prospective, randomized, double blind, placebo-controlled study.
Population	198 patients with suspected acute alcohol withdrawal syndrome

Design

Inclusion/Exclusion	I: > 18 year old with suspected acute alcohol withdrawal syndrome E: Allergy to study drugs, hepatic impairment, no IV access, and other primary diagnosis
Intervention	IV phenobarbital (10 mg/kg) in 100 mL normal saline over 30 mins
Outcomes	Primary: Initial level of hospital admission from the ED Secondary: Use of continuous lorazepam infusion, hospital length of stay, total amount of lorazepam used, and incidence of adverse events

Baseline Characteristics

	Phenobarbital (n = 51)	Placebo (n = 51)
Male	46 (90)	45 (88)
Age, years: median (IQR)	46 (40–52)	48 (37–54)
Initial AWCA score: median (IQR)	6 (4–10)	7 (4–10)
Initial heart rate: median (IQR)	106 (100–123)	112 (108–120)
Initial tremor: n (%)	48 (95)	48 (95)
Initial sweats: n (%)	25 (49)	32 (63)
Initial agitation: n (%)	20 (40)	21 (41)
Initial anxiety: n (%)	35 (68)	43 (84)
Altered level of consciousness: n (%)	30 (58)	35 (68)
Auditory/visual disturbances: n (%)	20 (40)	21 (41)
Time to initial lorazepam administration, minutes: median (IQR)	84 (48–146)	84 (40–312)
Time to study medication administration, minutes: median (IQR)	144 (103–263)	150 (100–26
Patients with prior alcohol withdrawal admissions to study institution: n (%)	21 (41)	25 (49)

Results

Primary Outcome	ICU admission rate: Phenobarbital vs Placebo8% vs. 25% Difference 17% [95% confidence interval (CI) 4–32%])
Secondary Outcomes	Use of continuous lorazepam infusion4% vs. 31%Difference 27% [95% CI 14–41%]Total lorazepam required26 vs. 49 mg Difference 23 mg [95% CI 7–40] There were no differences in telemetry admission or floor ward admissionTrend toward lower median ICU or total hospitalHospital LOS: 76 hr (54–114) vs 118 hr (47–190) ICU LOS: 34 hr (30-276) vs 94 hr (43–134)
Adverse Effects	No differences in incidence of intubation, seizure, mechanical restraints, and bedside sitter.There were no falls or mortality reported in either group.

Discussion

---

Alcohol withdrawal syndrome in critically ill patients: Protocolized versus Nonprotocolized management.

Duby JJ et al. J Trauma Acute Care Surg. 2014 Dec;77(6):938-43

Objective	to compare patient outcomes in critically ill patients with AWS, regardless of their admission ICU diagnosis, that were treated with this protocolized approach versus a non-protocolized approach.
Design	Retrospective pre-post study.
Population	135 patients with suspected acute alcohol withdrawal syndrome admitted to the ICU

Design

Inclusion/Exclusion	I: > 18 year old with suspected acute alcohol withdrawal syndrome admitted to ICU E: Patients with severe brain injury—defined as persistent Glasgow Coma Score < 8
Intervention	Pre-Protocol:Typically received continuous infusions or scheduled doses of BZDs per physician preference Post-ProtocolEscalating doses of diazepam and phenobarbital according to an AWS protocol
Outcomes	Primary: ICU length of stay Secondary: Mean and median BZD use, mean and median phenobarbital use, duration of sedation, requirement for mechanical ventilation (MV), ventilator-free days, and requirement for MV due to AWS

Baseline Characteristics

	Pre  (n = 60)	Post (n = 75)	P Value
Age	55.7 ± 8.7	50.7 ± 13.8	0.03
Male	81.6%	81.3%	1.0
History of Alcohol Withdrawal	40%	30.6%	0.28
History of Psychosis	10%	12%	0.78
History of Delirium Tremens	10%	4%	0.19
History of Seizure	18.3%	21%	0.83
Mean SOFA score on admit	6.1 ± 3.7	3.9 ± 2.9	0.0004
Mean blood alcohol level on admit (mg/dL)	135 ± 156	134 ± 140	0.56

Results

Primary Outcome	ICU LOS: Pre vs Post Protocol9.6 ± 10.5 vs 5.2 ± 6.4   (P-value 0.0004)
Secondary Outcomes	Time on Ventilator (days)5.6 ± 13.9   vs 1.31 ± 5.6    (P-value < 0.0001) Ventilator-free days21.3 ± 9.5  vs  26.3 ± 5.6   (P-value 0.0004) Intubation due to AWS 13 (22%)  vs 4 (5%) (P-value < 0.001) Need for continuous sedation33 (55%) vs 18 (24%)    (P-value < 0.001) Duration of sedation (days)10.8 ± 8.9  vs 3.5 ± 3.5    (P-value < 0.001)
Adverse Effects	Death7 (12%)  vs 2 (3%) (P-value 0.07)

Discussion

---

Tidwell WP, Thomas TL, Pouliot JD, Canonico AE, Webber AJ. Treatment of Alcohol Withdrawal Syndrome: Phenobarbital vs CIWA-Ar Protocol. Am J Crit Care. 2018;27(6):454-460.

• Objective
  • To compare a symptom-triggered benzodiazepine protocol to a phenobarbital protocol for the treatment of alcohol withdrawal

• Design
  • Retrospective cohort study of a 42-bed medical ICU in a private teaching hospital in Nashville Tennessee
  • Inclusion: Medical ICU patients admitted from January 1, 2016 through June 30, 2017 and treated for the onset or prevention of AWS 
  • Exclusion: Received CIWA-Ar-based treatment for >24 hours before starting phenobarbital protocol; received no doses of either protocol; pregnant;  had phenobarbital as a documented outpatient maintenance medication

• Intervention and Comparator
  • Phenobarbital taper protocol including lorazepam prn for agitation – initial phenobarbital dose 64.8 mg PO TID-260 mg IV based on history of delirium tremens, n=60
  • CIWA-Ar-based lorazepam protocol, n=60

• Outcomes (Phenobarbital protocol vs CIWA-Ar protocol)
  • Use of phenobarbital protocol was associated with a statistically significant ↓ in ICU LOS (2.5 vs 4.4, p <0.001)
  • Use of phenobarbital protocol was associated with a significantly ↓ total hospital LOS (4.3 vs 6.9, p=0.004)
  • Use of phenobarbital significantly ↓ the total lorazepam equivalents (11.3 mg vs 35.2 mg, p <0.001)
  • Use of adjunctive medications, specifically dexmedetomidine, was much lower in the phenobarbital group (7% vs 17%, p=0.002)

• Limitations
  • Small sample size, single-center, retrospective design
  • High incidence of comorbid conditions may have affected need for mechanical ventilation

• Conclusions
  • Phenobarbital-based treatment of AWS seems to provide comparable benefit to traditional benzodiazepine-based treatment

---

Overview of Evidence

Studies evaluating the use of Phenobarbital in Benzodiazepine in alcohol withdrawal

Author and Year	Research Design and Sample size	Intervention	Results
Nisavic, 2019	Retrospective observational/ n=562	Benzodiazepine only fixed dosingPhenobarbital- Based Protocol (IM load + PO taper)	No difference in AWS-related seizures, ICU admission, over-sedation, LOS, and hallucinations ↑ Delirium in BZD group In BZD→PB crossover pts, PB led to rapid improvement of BZD resistant AWS symptoms
Sullivan, 2018	Retrospective observational/ n=209	Benzodiazepine only CIWA- Protocol PB + Benzodiazepine CIWA Protocol	No difference in ICU admission, intubation, hypotension, ED LOS, CIWA score at ED discharge PB group had ↓ hospital LOS and Max CIWA score at 24 hrs
Rosenson, 2013	Retrospective cohort analysisn=102	PB 10 mg/kg IV x1 + PRN benzodiazepines Placebo + PRN benzodiazepines	PB had ↓ ICU admission PB had ↓continuous infusion lorazepam PB had ↓ total lorazepam requirements No difference in ICU or hospital LOS
Duby et al (2014)	Retrospective, pre-post studyn = 135	Nonprotocolized BZD versus symptom- triggered, protocolized dose escalation of diazepam and phenobarbital	ICU LOS was significantly lower with protocolized delivery (5.2 ± 6.4 days vs  9.6 ± 10.5 days, P = 0.0004); significantly  fewer intubations for AWS with protocolized delivery (5% vs 22%, P < 0.001). Protocolized delivery resulted in significantly less time on the ventilator, more ventilator free days, less need for continuous sedation, and shorter duration of sedation
Ibarra, 2019	Retrospective observational/ n=78	Lorazepam protocol only (LZP)   PB x 1 + LZP protocol (PB+LZP)	No difference in daily lorazepam requirements or hospital LOS   PB+LZP group had ↑ pts d/c within 72 hrs   No patient in PB group experienced intubation or hypotension
Nelson, 2019	Pre-post observational/ n=300	IV diazepam alone (DZP)   IV LZP + IV PB (LZP + PB)   IV PB alone (PB)	No difference in ICU admission, ICU LOS, and need for intubation.   PB associated with ↑ ED LOS but ↓ BZD requirements
Sullivan, 2018	Retrospective observational/ n=209	BZD only CIWA- Protocol   PB + BZD CIWA Protocol	No difference in ICU admission, intubation, hypotension, ED LOS, CIWA score at ED discharge   PB group had ↓ hospital LOS and Max CIWA score at 24 hrs
Hendey, 2011	Prospective, randomized, double-blind trial n=44	PB 260 mg IV ×1, 130 mg IV PRN   Lorazepam 2 mg IV PRN + PRN chlordiazepoxide	PB and LZ both reduced the average CIWA-Ar score from baseline to discharge   No difference in  ED LOS and hospital LOS
Young, 1987	Prospective, uncontrolled trial n=62	260 mg IV ×1 then 130 mg IV until clinical end point of light sedation	Safe discharge from ED was achieved in 92% of patients   Average ED LOS was 3 h, 47 min   No discharged patients returned to ED during the following week   Adverse effect in 6% of patients (none were admitted to hospital): 1 hypotension, 1 ataxia, 2 lethargy after final bolus doses

---

Phenobarbital Protocols