Clinical and toxicological factors associated with pediatric intensive care unit admission in pediatric acute poisoning: a 10-year retrospective cohort study

INTRODUCTION: Acute poisoning is an important cause of emergency department visits among children and adolescents. Early identification of patients requiring escalation-of-care or pediatric intensive care unit-level monitoring remains clinically challenging. The purpose of this study is to evaluate clinical and toxicological factors associated with pediatric intensive care unit admission in pediatric acute poisoning. METHODS: This retrospective cohort study included 652 patients younger than 18 years managed for acute poisoning between 2015 and 2024 at a tertiary university hospital. Demographic, exposure-related, toxicological, clinical, and laboratory variables were analyzed. Pediatric intensive care unit admission was evaluated as an escalation-of-care and healthcare utilization outcome. Three multivariable logistic regression models were constructed, including models excluding variables closely related to admission decisions to reduce potential incorporation bias. RESULTS: Of 652 patients, 99 (15.2%) were admitted to the pediatric intensive care unit. Across the primary and sensitivity models, delayed presentation, psychiatric disorder history, and multiple-drug ingestion remained independently associated with admission. Lower Glasgow Coma Scale score and cardiovascular findings were associated with admission in models including clinical severity variables. Exploratory receiver operating characteristic analyses demonstrated poor discriminatory performance of individual parameters, indicating that none were suitable as stand-alone triage tools. No deaths occurred, invasive mechanical ventilation was required in one patient, and advanced organ-support interventions were uncommon. DISCUSSION: Pediatric intensive care unit admission appeared to reflect both observed clinical instability and anticipated toxicological risk rather than poisoning severity alone. The poor discriminatory performance of individual parameters suggests that admission decisions are multifactorial and should not rely on a single clinical or laboratory variable. CONCLUSION: Delayed presentation, multiple-drug ingestion, psychiatric disorder history, and clinical severity indicators were associated with pediatric intensive care unit admission. These findings should be interpreted as observational evidence regarding escalation-of-care decisions rather than as a validated prediction model or triage tool in this single-center pediatric poisoning cohort.