Prospective identification of risk factors for patients with severe acute poisoning in emergency departments in northern Taiwan.

@article{liao2026prospective,
  title = {{Prospective identification of risk factors for patients with severe acute poisoning in emergency departments in northern Taiwan.}},
  author = {Po-Hsiang Liao and Chih-Chuan Lin and Chia-Ming Chang and Ying-Tse Yeh and Tse-Yao Wang and Hsien-Yi Chen and Chorng-Kuang How and Yen-Chia Chen},
  journal = {BMC emergency medicine},
  year = {2026},
  month = jul,
  doi = {10.1186/s12873-026-01709-2},
  url = {https://europepmc.org/article/MED/42972223},
  note = {CC BY-NC-ND},
}

TY  - JOUR
TI  - Prospective identification of risk factors for patients with severe acute poisoning in emergency departments in northern Taiwan.
AU  - Po-Hsiang Liao
AU  - Chih-Chuan Lin
AU  - Chia-Ming Chang
AU  - Ying-Tse Yeh
AU  - Tse-Yao Wang
AU  - Hsien-Yi Chen
AU  - Chorng-Kuang How
AU  - Yen-Chia Chen
PY  - 2026
DA  - 2026/07/31
JO  - BMC emergency medicine
DO  - 10.1186/s12873-026-01709-2
UR  - https://europepmc.org/article/MED/42972223
AB  - Background  Poisoning remains a major public health issue and a common reason for emergency department (ED) visits in northern Taiwan. Identifying predictors of severe poisoning is crucial for risk stratification and improved patient outcomes.Methods  This prospective, multicenter, cross-sectional cohort study enrolled 2,470 patients with acute poisoning over a consecutive two-year period across two tertiary medical centers in northern Taiwan. Data on demographics, clinical features, toxic agents, and vital signs were collected at ED triage. The primary outcome was severe poisoning, defined as a Poisoning Severity Score (PSS) of 3-4. For descriptive purposes, patients were categorized into non-severe (PSS 0-2) and severe groups. Multivariable logistic regression was performed to identify independent predictors of severe poisoning, reporting adjusted odds ratios (aOR) with 95% confidence intervals.Results  Poisoning accounted for 4.7 cases per 1,000 ED visits during the study period. Ingestion was the most common exposure route (55.2%), followed by inhalation (16.6%). Prescribed medication represented the largest toxicological category (32.5%), whereas pesticides accounted for 9.8% of exposures but contributed disproportionately to mortality. Overall, 77 patients (3.1%) died, with paraquat being responsible for 59.7% of all fatalities, making it the most lethal agent observed. Other substances associated with fatal outcomes included warfarin, amphetamines, hydrochloric acid, benzodiazepines, digoxin, carbon monoxide, and methotrexate. Most patients (88.5%) experienced non-severe poisoning (PSS 0-2), while 11.5% experienced severe poisoning (PSS 3-4). In multivariable analyses, significant predictors of severe poisoning at ED presentation included increasing age, abnormal vital signs at triage, suicidal intent, and pesticide ingestion. These associations remained robust in the supplementary model, which excluded all ED triage vital signs, where age, intentional exposure, illicit drugs and pesticide ingestion continued to independently predict severe poisoning.Conclusions  Advanced age, abnormal vital signs, suicidal intent, and pesticide exposure, particularly paraquat, are strong predictors of severe poisoning and mortality in this cohort from northern Taiwan. Early identification of high-risk patients based on clinical features available at ED triage may facilitate timely risk stratification and resource allocation for acute poisoning, particularly in regions with similar poisoning patterns to northern Taiwan.
SN  - 1471-227X
ER  -