Featured

Hospitalized Patients Who Receive Alcohol Use Disorder Treatment Can Substantially Reduce Heavy Drinking

A new study found that the oral and extended-release injectable forms of naltrexone are equally effective in helping patients consume less alcohol, suggesting that clinicians should integrate this medication into routine hospital care.

Nearly 30 million adults in the United States experience alcohol use disorder (AUD), but the vast majority of people with this condition do not receive treatment. A new study led by the School of Public Health, BU Chobanian & Avedisian School of Medicine, and Boston Medical Center researchers indicates that hospitals may be an ideal setting to close this gap in care. 

Published in JAMA Internal Medicine, the study found that hospitalized patients with alcohol use disorder who began taking the AUD medication naltrexone before they were discharged were able to reduce heavy drinking in the three-month period following their hospital stay.

The findings specifically showed that both forms of naltrexone, which is offered as a pill or as an extended-release injectable, were similarly effective at reducing patients’ alcohol consumption. After three months of treatment, heavy drinking within the last 30 days decreased by approximately 38 percentage points among patients who took the oral version of naltrexone, compared to about a 46 percentage-point decrease among patients who received the injectable version of the medication.

The study is the first randomized clinical trial—the gold standard for effectiveness research—to compare the efficacy of both formulations of naltrexone, and the researchers hope this insight encourages clinicians to integrate this medication into routine hospital care.

“The hospital is a busy place with many competing priorities and one’s chronic drinking problems can be missed or prioritized lower than the acute problems that led to the patient’s hospitalization,” says study corresponding author Jeffrey Samet, professor of community health sciences at SPH, John Noble Professor of Medicine at BU Chobanian & Avedisian School of Medicine, and a primary care physician at BMC. “Treating alcohol use disorder among hospitalized patients is a great opportunity to better care for individuals with this problem, as our findings clearly show that outcomes for these patients improve.”

The research team utilized data from the Alcohol Disorder Hospital Treatment (ADOPT) study to compare the effectiveness of each type of naltrexone administered to 248 patients with alcohol use disorder who completed a hospital stay at an urban hospital between June 2016 and March 2020. Patients received the medication on the day of discharge, taking either a daily pill or receiving a monthly injection for three months. The researchers based medication effectiveness on the reduction in heavy drinking and healthcare utilization within the last 30 days of the three-month follow-up period. Heavy drinking was defined as five or more drinks for men and four or more drinks for women per day.

Although both forms of naltrexone were roughly equal, the researchers observed a slightly higher adherence to the injectable form of the medication, consistent with prior research. But they note that adherence levels could improve in real-world situations where patients would have flexibility in determining with their doctor the type of naltrexone that best suits their individual needs and preferences in formulation, dosing frequency, and cost. The injectable version of the drug eliminates the need for daily adherence, but requires additional time and effort for injection visits and higher upfront costs, at $1,064 per injection versus $38.10 for 30 days of pills. 

The team did not observe a difference in acute or AUD-related hospital visits based on consumption of either medication in the three-month period after patients were discharged from the hospital. 

“I hope the takeaway for policymakers and health systems is clear: effective, evidence-based medication treatment for alcohol use disorder exists, and it is time to make this treatment more accessible at hospital discharge,” says study lead author Kara Magane, senior director of research operations in the Department of Health Law, Policy & Management at SPH, adding that future research should explore how to optimize this delivery.

“It would also be valuable to understand patient perspectives on treatment preferences and adherence after discharge, and how factors such as housing or follow-up access influence outcomes,” Magane says. “Ultimately, the goal is to make AUD treatment initiation routine and patient-centered in hospital settings.”

This publication also marks a deeply personal milestone: it completes the work begun by Richard Saitz, the study’s principal investigator and former chair and professor of community health sciences at SPH who passed away in 2022 before the study’s findings could be shared. “Rich Saitz, a pioneering voice in addiction medicine, envisioned and led this research with characteristic rigor, compassion, and clarity of purpose,” Magane says. “The study is dedicated to him and reflects his enduring influence on the field and on those of us fortunate enough to work alongside him.”

The study was funded by the National Institute on Alcohol Abuse and Alcoholism, with additional support by Alkermes and the United States Drug Testing Laboratories, Inc. At SPH, the study was coauthored by Kimberly Dukes, research associate professor of biostatistics; Sarah Fielman (SPH’23), research project manager in the Department of Community Health Sciences; Debbie Cheng, assistant dean of data science, executive director of the Center for Health Data Science, and professor of biostatistics; Matthew Bullard, senior project manager for the Biostatistics and Epidemiology Data Analytics Center (BEDAC); and Clara Chen, assistant director of operations for BEDAC. Additional coauthors include researchers BU Chobanian & Avedisian School of Medicine, BMC, BU College of Arts & Sciences, South Shore Hospital, and the University of Pittsburgh School of Medicine.