Intern year, they say, is about learning to distinguish “sick” from “not sick.”
As the intern on-call overnight, you call the Emergency Department physician for pass off on Mr. Jones; when you hear “46-year-old man with heavy alcohol use disorder presenting with new jaundice and mild confusion,” it takes you only a split second to recognize that your patient is “sick.”
You have flipped through the labs, and as you walk down to the Emergency Department, you plug Mr. J’s bilirubin and prothrombin time into the medical calculator app on your phone. Maddrey’s discriminant function of 60. Knowing that anything above 32 indicates severe alcoholic hepatitis, your pace quickens.
You take a history and examine Mr. J, confirming the time course of his symptoms and his alcohol use history with his brother. After reviewing the work-up already underway, you become confident in your clinical diagnosis of alcoholic hepatitis, and you consider treatment options. Prednisolone and pentoxifylline are the two you know of, but even your senior resident hems and haws when you ask how effective each of these treatments is. She is not alone – data from prior studies have been mixed, leaving doubts about how best to treat this clinical syndrome.
To address this uncertainty, Thursz and colleagues conducted the STOPAH trial, now published in NEJM. Using a 2×2 factorial design, 1100 adults with a clinical diagnosis of severe alcoholic hepatitis from 65 hospitals in the United Kingdom were randomized to receive either prednisolone, pentoxifylline, both, or placebo.
Similar to your patient, trial participants were on average in their late 40s, majority male, and had a mean discriminant function in the 60s. The assigned treatment was continued for 28 days.
The results: for pentoxifylline, there was no improvement in 28-day mortality. 16% of patients who were treated with pentoxifylline died – the same percentage as those who did not receive it. For prednisolone, mortality was 14% among those who received glucocorticoids versus 18% for those who did not. With a p-value of 0.06, this trend toward benefit fell just short of statistical significance. Neither treatment improved longer-term survival at 90 days or 1 year; serious infections were nearly twice as common among those receiving prednisolone (13% vs 7%, P=0.002).
For pentoxifylline, then, these data seem conclusive. However, for prednisolone, these data don’t offer a clear directive. Deputy Editor Dr. Mary Beth Hamel addresses this gray zone: “For prednisolone, some uncertainty persists about a possible short-term benefit. But it is clear that any benefit is far from a durable, long-term solution for those suffering from alcoholic hepatitis.”
So to your patient: do you administer either of these treatments? Pentoxifylline – probably not. Prednisolone is a tougher decision. You know there is no magic to the P value of 0.05, and you think the possibility of even a short-term mortality benefit is worth the increased risk of infection. But before you write the order, you want to see what your senior resident thinks.
As you walk back to the work room, you scroll through your phone. You close the discriminant function calculator and then Google helps you find that famous Scottish proverb that has been swirling in your head: “They speak of my drinking, but never think of my thirst.” You think about how that would make a good opening for an article about how we need better ways to have helped Mr. Jones before it came to this – before he became “sick.” But you don’t have time to think about that now – it’s almost morning. You put your phone back in your white coat pocket, sit down at a computer, and start your note.
View the NEJM Quick Take video summary of the results of this article on NEJM.org.