Pediatric opioid and cannabis poisoning hospitalizations in the United States, 2016-2022: a national age-matched inpatient analysis.

Fournier L.; Shehadeh S.; Saied M.; Maman D.; Pillar G. · Frontiers in pediatrics · 2026

**Background & Methods** This retrospective cohort study analyzed National Inpatient Sample data from 2016-2022 to compare hospitalization outcomes between children and adolescents (0-18 years) with principal diagnoses of opioid versus cannabis poisoning in the United States. The analysis included 17,555 weighted pediatric poisoning hospitalizations (10,780 opioid-related, 6,775 cannabis-related), with a 1:1 nearest-neighbor age-matched cohort design adjusted for demographics, socioeconomic status, and admission characteristics. **Key Findings** - Cannabis poisoning hospitalizations increased significantly from 24.0% of all pediatric opioid and cannabis poisonings in 2016 to 47.2% in 2022 (survey-weighted trend OR 1.18 per year; P<0.001), representing a doubling of relative burden over the study period. - Opioid poisoning was associated with substantially greater morbidity compared to age-matched cannabis poisoning cases: mechanical ventilation for >24 hours (adjusted OR 4.65), acute kidney injury (adjusted OR 2.55), and respiratory failure (relative risk 2.33). - Opioid poisoning resulted in prolonged hospitalizations (1.06 additional days) and significantly elevated costs ($17,286 higher hospital charges) compared to cannabis poisoning cases. **Dosage & Administration** Not reported. The study analyzed poisoning cases without characterizing doses or routes of exposure. **Safety & Adverse Effects** Documented complications in opioid poisoning cases included respiratory failure requiring mechanical ventilation, acute kidney injury, and respiratory depression necessitating prolonged ventilatory support. Cannabis poisoning cases experienced fewer severe complications requiring mechanical ventilation or resulting in acute organ dysfunction. Comparative mortality data were not reported due to limited events in one exposure group. **Evidence Quality** This is a large, nationally representative retrospective cohort study with appropriate survey-weighted statistical methods and multivariable adjustment. Limitations include reliance on administrative ICD-10-CM coding for exposure classification without clinical confirmation, inability to assess actual doses or toxin amounts, incomplete mortality reporting due to small event numbers, and inherent coding biases. The age-matched design strengthens causal inference but does not eliminate confounding. Findings represent observational associations rather than causal relationships and apply to hospitalized populations only, limiting generalizability to all poisoning exposures.

More recent cannabis research