24,643 Patients Later, Phenobarbital and Diazepam Look Like a Tie. Don't Rip Up the Protocol Yet.

In most hospitals, a person coming off alcohol gets a sedative, because the body that has adapted to alcohol can seize, hallucinate or stop when the alcohol stops. The old standby is a benzodiazepine like diazepam. The challenger is phenobarbital, a barbiturate older than most of the doctors prescribing it, whose use for alcohol withdrawal has risen by more than 200% in a decade on thin evidence.

Here is the take: a tie is an answer, and the answer is that nobody needs to panic-swap their protocol.

On October 12, a team led by Tess Bradley, with Boston University’s Nicholas Bosch as corresponding author, published the largest head-to-head yet in JAMA Internal Medicine: 24,643 adults hospitalized for alcohol withdrawal across 754 U.S. hospitals, split by what they were given on the first day.

Why a sedative at all

Heavy, long-term drinking teaches the brain to compensate: it turns its own brakes down and its accelerator up so you can still function with alcohol on board. Take the alcohol away and nothing is pushing back against the accelerator. That is the shaking, the racing heart, and in the worst cases the seizure. Diazepam and phenobarbital both work by putting a brake back on, slowly enough that the brain can reset. The question for hospitals is which one does it more safely, because patients on either can end up too sedated, and because the answer decides what a nurse watches for all night.

What they actually did

This was a retrospective look-back at hospital records, not a trial where patients are randomly assigned. That matters. In a trial, a coin decides who gets which drug. In records, a doctor decided, usually for a reason (sicker patient, local habit, what the pharmacy stocked), and those reasons can quietly tilt results. The team used statistical matching to balance the two groups on age, other illnesses and overall health, which narrows the problem and cannot erase it.

What they found

Patients on phenobarbital and patients on diazepam had no clear difference in length of stay or in how often they needed a breathing machine. Bosch’s summary of the result: the two drugs are “almost equal in how well they work.” The authors wrote that both “appear to offer nearly similar benefits and risks.”

The one separation was about people leaving. Patients given phenobarbital were less likely to walk out of the hospital against medical advice. Bosch himself calls that “a finding that will require more research to confirm,” and he is right to. It could be the drug, or it could be that the people and units choosing phenobarbital differ in ways the records don’t show. If it holds, it matters more than a day shaved off a stay, because a person who leaves mid-withdrawal takes the danger of withdrawal out the door with them.

What the study doesn’t settle

Two limits for anyone about to quote this in a pharmacy committee. First, the comparison is with diazepam, a long-acting benzodiazepine. Earlier work that favored phenobarbital mostly compared it with lorazepam, a shorter-acting one, so the tie says nothing about the older fights. Second, a recent emergency department study in the Journal of Emergency Medicine compared whole treatment strategies for severe withdrawal and is the one to read for sicker patients. An accompanying JAMA Internal Medicine editorial, by Dempsey and Anderson, flags the open questions: whether the drugs perform differently depending on how severe the withdrawal is, and what the right phenobarbital dosing is.

What to do with it this week

If you work in a hospital or detox, the authors’ own read is that you need not rapidly replace diazepam protocols with phenobarbital. Put one question to your team: do we track, by drug, how many patients leave against medical advice during withdrawal? If you don’t, you can’t tell whether a switch would help you.

Put one question to your team: do we track, by drug, how many patients leave against medical advice during withdrawal?

If you are the person on the bed, or you love one, the practical point is older than this study. Alcohol withdrawal is a medical event, not a character test, and the sedative used is a clinician’s call that today’s evidence says is close to a coin flip. What isn’t a coin flip is staying, because the hospital is the one place that can catch the seizure. The drug is the smaller decision. Being there is the bigger one.