Determinants of Intentional Pharmaceutical Poisoning: A Cross-Sectional Study From a Referral Poisoning Center in Iran Publisher

Summary: Who faces greater intentional medication poisoning risk? An Iranian study linked age, sex and psychiatric history with intentional cases; an AI model predicted them well. #Toxicology #PublicHealth

Eizadi Mood N ; Rajaeizadeh M ; Feizi A ; Yazdi R ; Dorostkar A ; Mirmoghtadaee P ; Meamar R
Source: Journal of Applied Toxicology Published:2026

Abstract

Pharmaceutical poisoning is a major health concern and a leading method of intentional self-poisoning. In Iran, broad medication access has increased the burden, yet factors distinguishing intentional from accidental exposures remain underexplored. This study aimed to identify the determinants of intentional pharmaceutical poisoning using both traditional statistical methods and machine learning techniques. We conducted a cross-sectional study at a referral poisoning center in Isfahan, Iran, between January 2015 and December 2019. Medical records of patients admitted with pharmaceutical poisoning were reviewed for demographics, toxicological characteristics, and past history. Univariate logistic regression assessed determinants of intentional poisoning. A multilayer perceptron (MLP) neural network evaluated variable importance and discrimination. Among 3096 eligible patients, 94.4% were intentional cases. Mean (SD) age of intentional cases was 32.14 (10.74) years, predominantly 20–40 years. Logistic regression identified female sex (OR: 2.01; 95% CI: 1.47–2.73), age 20–40 (OR: 9.85; 95% CI: 6.79–14.29), age 41–60 (OR: 4.80; 95% CI: 2.94–7.85), and history of psychiatric disease (OR: 1.54; 95% CI: 1.03–2.30) as significant determinants. The MLP showed strong discrimination (AUC: 0.859); age, number of children, education, and history of psychiatric disease were top predictors. Efforts should prioritize family and social support, mitigation of educational/occupational stressors, early psychiatric screening, timely psychological care, and community-based programs to identify and protect high-risk groups and reduce recurrence. © 2026 John Wiley & Sons Ltd.