New Tools Identify Patients at High Risk of Overdose Post-Hospital Discharge

Research published in the Canadian Medical Association Journal reveals that risk prediction tools can significantly aid in identifying patients most vulnerable to overdose and death after leaving the hospital against medical advice. Patients who opt for a “before medically advised” (BMA) discharge are approximately twice as likely to die and ten times more likely to experience an illicit drug overdose within the first 30 days of leaving.

In the United States, around 500,000 patients initiate BMA discharges each year, while Canada sees about 30,000 similar cases. This alarming trend raises critical concerns regarding patient safety and the need for effective assessment tools.

Dr. Hiten Naik from the University of British Columbia and his co-authors emphasize the importance of calculating individual risk factors. They argue that combining these assessments with clinical judgment enables healthcare professionals to engage in meaningful discussions with patients regarding the implications of BMA discharges. This approach can help evaluate a patient’s decision-making capacity and explore strategies to mitigate potential risks following discharge.

Development of Risk Prediction Models

The researchers created two risk prediction models aimed at estimating the likelihood of death and overdose among patients with varying health backgrounds. One model focuses on the risk of death from any cause within 30 days following a BMA discharge, while the other specifically targets those with a history of substance use to predict the risk of an illicit drug overdose.

Analyzing data from British Columbia, the study examined two cohorts: Cohort A comprised 6,440 adults from the general population who underwent BMA discharges, and Cohort B included 4,466 individuals with substance use histories. Findings indicated that, contrary to expectations, death was less common in Cohort A, with one death occurring for every 63 BMA discharges. Significant predictors of mortality included multimorbidity, heart disease, and cancer.

In contrast, Cohort B revealed concerning trends. High-risk factors for overdose included homelessness, reliance on income assistance, opioid use disorder, non-alcohol substance use disorder, previous overdoses within the past year, and discharges from surgical services. Notably, the study found that illicit drug overdoses occurred at a rate of around one for every 19 BMA discharges among those with a history of substance use, highlighting a critical window for overdose prevention strategies.

Implications for Healthcare Systems

The authors propose that healthcare systems could integrate these risk prediction models to enhance their approach to managing high-risk BMA discharges. Implementing alerts and automatic enrollment in support programs for identified patients could provide essential assistance and reduce the likelihood of adverse outcomes.

“These models offer a starting point for identifying patients who are high risk and may benefit from greater support,” the researchers conclude. The findings underscore the need for more comprehensive strategies to protect vulnerable individuals during post-discharge periods.

For further details, refer to the original study: “Predicting drug overdose and death after ‘before medically advised’ hospital discharge,” published in the Canadian Medical Association Journal in 2025. DOI: 10.1503/cmaj.250492.