Leaving Hospital Early Linked With Higher Drug Overdose Risk

Patients who leave the hospital against medical advice have a 10-fold higher risk for subsequent drug overdose in the following month, a new study indicates.

Overall, 1 in 30 patients with an early discharge — and 1 in 12 patients with opioid use disorder who left early — experienced an overdose within 30 days of leaving the hospital.

“In the last 20 years, I’ve provided care for many patients who ultimately decided to leave hospital before their inpatient medical care was complete. I worry about these patients,” senior author John Staples, MD, a clinical associate professor in internal medicine at the University of British Columbia in Vancouver, British Columbia, told Medscape Medical News.

“Often, they have ongoing medical problems — like infections — that haven’t been fully treated. But I also worry because I know that out-of-control substance use disorder sometimes contributes to the decision to leave hospital,” he said. “Patients can be in pain related to whatever landed them in hospital, and sometimes the treatment of opioid withdrawal is suboptimal. Plus, being sick is stressful. These forces can make drug cravings worse.”

Before discharge, for instance, hospital care teams must arrange outpatient prescriptions, restart medications for opioid use disorder, such as methadone and suboxone, and coordinate follow-up care, which can be difficult to do in the chaotic moments before patients leave, Staples said. Many times, systems aren’t in place to ensure proper follow-up after discharge, either.

“Reflecting on this, I wondered what happened to these patients after they leave hospital. I particularly wondered about the risk of overdose,” he said. “If we understand what’s happening to them after they leave the hospital, we might be able to come up with ways to keep them safe.”

The study was published online on September 23 in the Canadian Medical Association Journal.

Against Medical Advice

Staples and colleagues conducted a retrospective cohort study to understand the link between drug overdose and “before medically advised” hospital discharge, which is also known as “patient-initiated” or “against medical advice” discharge. They looked at administrative health data for 189,808 adult nonelective hospital admissions between 2015 and 2019 in British Columbia.

Overall, 6440 hospital admissions (3.4%) ended with patients leaving the hospital before medically advised. Among patients with opioid use disorder, one in five hospital stays ended in a discharge against medical advice.

In general, patients with an early discharge were more likely to be younger men with fewer medical comorbidities and more psychiatric illnesses. They were also more likely to have a history of opioid use disorder, intravenous drug use, multiple previous drug overdoses, alcohol use disorder, a hospital admission for a psychiatric condition in the past year, and a history of homelessness.

Among 979 overdoses, 239 occurred among patients with a discharge against medical advice and 740 occurred among those with a physician-advised discharge.

Among 820 overdoses in the first 30 days after any hospital discharge, 755 (92.1%) involved patients with a history of substance use disorder.

The rate of fatal or nonfatal illicit drug overdose in the 30 days after hospital admission was 10 times higher after an early departure than after a physician-advised discharge, at 2.8% vs 0.3%. The unadjusted risk for fatal overdose was six times higher.

Early discharge was associated with a 58% relative increase in the risk for subsequent overdose, even after adjusting for other risk factors. Before medically advised discharge was also associated with a twofold higher risk for subsequent emergency department visits and unplanned hospital readmissions.

“This implies that the before medically advised discharge itself causes some of the increase in overdose risk,” Staples said. “This is another signal telling clinicians, hospitals, and health systems that we could be doing a better job of smoothing the transition from the hospital to the community, even — or especially — if the patient is leaving hospital before it’s medically advised to do so.”

Developing Safe Solutions

Hospitals could help patients with a history of substance use disorder by developing evidence-based protocols to prevent early discharges and conduct postdeparture outreach, the study authors wrote.

For instance, evidence-based interventions could include adequate treatment of pain and withdrawal, specialized addiction medicine teams, efforts to reduce stigmatization of people who use drugs, and less restrictive temporary pass and visitor policies, Staples said.

“These patients are sometimes saying they don’t want the hospital care they were being offered, but that doesn’t mean they don’t want any medical care,” he said. “They might flourish under less restrictive inpatient conditions.”

Patients with substance use disorder may want low-barrier addiction care, prescriptions for opioid use disorder, and access to harm-reduction supplies such as take-home naloxone kits, Staples added.

“With nearly 45,000 opioid overdose deaths recorded between 2016 and 2023, Canada continues to grapple with a drug-poisoning epidemic,” Sarah Larney, an associate professor of family medicine and emergency medicine at the University of Montreal, Montreal, Quebec, Canada, and an investigator at the Canadian Research Initiative in Substance Misuse, told Medscape Medical News.

Larney, who wasn’t involved with this study, has researched overdose trends from the past decade in Quebec, finding that deaths have increased over time, particularly in 2020, and remained high.

“Interventions are essential,” she said. “Naloxone, safer supply, and accessible supervised consumption sites are part of the answer, but they are not enough today, even though we know harm reduction works.”

With the lethal nature of today’s drug supply, she said, broader actions around drug law reform are needed, such as drug decriminalization or legalization.

“The literature tells us that criminalization and punishment do not work,” Larney said. “Instead, we should invest in social policy reforms that address the root causes of problematic drug use — poverty, income inequality, and the housing crisis.”

The study was supported by the Canadian Institutes of Health Research, Vancouver Coastal Health Research Institute’s Innovation and Translational Research Award, and the University of British Columbia (UBC) Division of General Internal Medicine, British Columbia, Canada, which were paid to Staples as the principal investigator. Staples also reported receiving salary support from a Michael Smith Health Research BC Health Professional-Investigator Award and the UBC Division of General Internal Medicine’s Academic Investment Fund. Larney reported no relevant financial relationships.

Carolyn Crist is a health and medical journalist who reports on the latest studies for Medscape Medical News, MDedge, and WebMD.