Can the intensive care requirement score improve triage at the emergency department for acutely intoxicated patients?

The intensive care requirement score was developed to predict the need for intensive care unit admission in acutely intoxicated patients. However, as the score was originally derived from a cohort already admitted to intensive care, its performance in emergency department populations remains uncertain. This study evaluates its performance in that setting. This retrospective study included patients who presented to the emergency department of the University Medical Centre Utrecht between 2015 and 2020. Eligibility criteria were similar to those used in the original derivation study, with modifications appropriate for the emergency department setting. Patients were excluded if they were younger than 14 years; had no toxic effects (e.g., ingestion of batteries, razor blades or spoons); or if intoxication was not the primary reason for presentation. The primary outcome of intensive care requirement was defined as receiving intensive treatment within 24 h of hospital admission or in-hospital death. Missing data were imputed, and model updating techniques, including calibration intercept adjustment and recalibration, were applied to improve predictive accuracy. Model performance was assessed using discrimination, calibration, and decision curve analysis. The final cohort comprised 1,146 patients, of whom 21 (1.8%) required intensive care unit admission. Discrimination was excellent, with an area under the receiver operating characteristic curve of 0.91 (95% CI: 0.84 − 0.98). Prediction accuracy improved most following recalibration, yielding a calibration intercept of 0 (95% CI: −0.47 to 0.47) and a calibration slope of 1.04 (95% CI: 0.75 − 1.33). Clinical decisions based on the recalibrated model reduced unnecessary intensive care unit admissions by 15% at a 5% risk threshold. In this study, the recalibrated intensive care requirement score accurately predicted the need for intensive care unit admission in intoxicated patients presenting to the emergency department. However, large-scale, multi-centre studies are needed to validate the generalizability of these findings across diverse healthcare settings. Following adjustment, the intensive care requirement score can assist in reducing unnecessary intensive care unit admissions among intoxicated patients presenting to the emergency department.