Phenobarbital or diazepam for alcohol withdrawal: no difference in stay or ventilation

Design

Target-trial emulation using a multicentre hospital database

Population

24,643 US adults hospitalised with alcohol withdrawal syndrome

Primary outcome

Hospital length of stay; invasive mechanical ventilation

Effect

4.1 vs 4.3 days (difference −0.21, 95% CI −0.42 to 0.00); ventilation 3.1% vs 2.7%

Phenobarbital is increasingly used for inpatient alcohol withdrawal, but most comparisons have been against lorazepam, a short-acting benzodiazepine. This study compared it with diazepam, a long-acting, self-tapering benzodiazepine closer to phenobarbital in how it behaves. Using a database covering 754 US hospitals, researchers emulated a trial in 24,643 adults admitted with alcohol withdrawal who started either drug on the day of admission, adjusting for differences with machine-learning-based methods.

Length of stay was 4.1 days with phenobarbital and 4.3 with diazepam (difference −0.21 days, 95% CI −0.42 to 0.00). Invasive ventilation was needed in 3.1% and 2.7% (difference 0.4%, 95% CI −0.7% to 1.4%). Other complications and escalation did not differ. Self-directed discharge was less common after phenobarbital (9.5% vs 12.6%).

The practical conclusion is that the choice between the two can rest on local protocols, nursing familiarity and monitoring capacity rather than on expected superiority. Phenobarbital's narrow therapeutic window and long half-life still demand a clear protocol, and diazepam remains a sound standard. This is observational; residual confounding by who chose which drug is possible.

• Use whichever of phenobarbital or diazepam your unit has a clear, practised protocol for.
• Do not switch to phenobarbital expecting shorter stays or fewer ventilations.
• With phenobarbital, use weight-based loading under a protocol and monitor sedation closely.
• Give parenteral thiamine early, ideally before glucose, in patients admitted with alcohol withdrawal.
• Plan follow-up for alcohol use disorder before discharge; self-discharge was common with both drugs.

Why it matters

Phenobarbital has been adopted on the strength of comparisons with lorazepam; against diazepam, the expected advantage did not appear.

Don't overread it

The lower self-discharge rate with phenobarbital is a secondary, observational finding and may reflect how sick or settled patients were.

The statistics, in plain English

The length-of-stay difference of −0.21 days has an interval reaching 0, so no difference is plausible; even at best it is about five hours. The ventilation interval runs from 0.7% lower to 1.4% higher, which rules out a large difference either way.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.