Mortality and repeated poisoning after self-discharge during treatment for acute poisoning by substances of abuse: a prospective observational cohort study

Though substance use is a known risk factor for self-discharge, patients self-discharging during treatment for acute poisoning have not previously been described. We charted characteristics of patients self-discharging during treatment for acute poisoning by substances of abuse looking for associations between self-discharge, repeated poisoning, and death.

All patients 12 years and older treated for acute poisoning by substances of abuse at an emergency outpatient clinic in Oslo, Norway, were included consecutively from October 2011 through September 2012. We collected data on gender, age, main toxic agent, suicidal intention, homelessness, history of severe mental illness, and self-discharge. Information on deaths was retrieved from the National Cause of Death Register. We did a multiple logistic regression analysis to look for associations between self-discharge and repeated poisoning and a Cox regression analysis for associations between self-discharge and death.

During one year, 1731 patients were treated for 2343 episodes of acute poisoning by substances of abuse. Two-hundred-and-sixty-six (15%) patients self-discharged during at least one poisoning episode. Self-discharging patients were older, median age 39 years vs 32 years (p <  0.001), more frequently homeless, 20/266 (8%) vs 63/1465 (4%) (p = 0.035), and the main toxic agent more frequently was an opioid, 82/266 (31%) vs 282/1465 (19%) (p <  0.001). Self-discharge was an independent risk factor for repeated poisoning. The adjusted odds ratio for two or more poisoning episodes during one year among self-dischargers was 3.0 (95% CI 2.2–4.1). The association was even stronger for three or more poisoning episodes, adjusted odds ratio 5.0 (3.3–7.5). In total, there were 34 deaths, 9/266 (3.4%) among self-discharging patients and 25/1465 (1.7%) among patients not self-discharging (p = 0.12). The adjusted hazard ratio for death among self-discharging patients was 1.6 (0.75–3.6).

Self-discharge was associated with frequent poisonings by substances of abuse. Short-term mortality was doubled among self-discharging patients, though this increase was not statistically significant. Still, the increased risk of repeated poisoning marks self-discharging patients as a vulnerable group who might benefit from targeted post-discharge follow-up measures.