Predictive factors for severe outcomes in substance abuse-related emergency visits: A 5-year retrospective analysis at a medical center in Taiwan

Background: Substance abuse is a considerable medical issue worldwide, yet current surveillance systems in Taiwan offer limited insights into the clinical characteristics and outcomes of substance abuse patients. This study aimed to delineate the epidemiology of emergency department visits related to substance abuse at a hospital in Taiwan and to identify factors predictive of severe complications or mortality. Methods: A retrospective analysis was conducted on substance abuse-related emergency department visits at a medical center in Taiwan between 2009 and 2013. Eligible participants were individuals aged 20 or older who had confirmed substance abuse through urinalysis. Variables such as patient demographics, substances abused, clinical characteristics, and outcomes were collected. Severe outcomes were defined as admission to the intensive care unit, requirement for endotracheal intubation, or in-hospital death. Logistic regression models were employed to identify factors contributing to severe outcomes. Results: The cohort consisted of 623 patients, of whom 64.0% were female and 67.1% were aged between 20 and 49 years. Benzodiazepines were detected in 75.3% of patients, while z-drugs (specifically zopiclone, zolpidem, or zaleplon) were found in 27.8%. Depressants, stimulants, and hallucinogens were present in 14.9%, 10.6%, and 0.6% of the cases, respectively. Of the patient, 121 (19.4%) experienced severe outcomes, including 116 (18.6%) intensive care unit admissions, 73 (11.7%) intubations, and 11 (1.8%) in-hospital deaths. Multivariable logistic regression analysis revealed multiple predictors of severe outcomes, such as emergency department triage level, aspiration pneumonia, leukocytosis, abnormal hepatic function, abnormal renal function, hypernatremia, and hypocalcemia. Conclusion: In Taiwan, benzodiazepines emerged as the most prevalent substance of abuse among emergency department visitors, and a significant proportion of these patients experienced severe outcomes. Continuous monitoring of severe outcome predictors is essential for enhanced understanding and management.