Development and validation of a model to predict post-intubation hypotension in the emergency department: a retrospective study

Endotracheal intubation is a critical procedure in the emergency department (ED), and one of the most common complications is hypotension. This study aims to develop and internally validate a nomogram-based predictive model to identify ED patients at high risk of post-intubation hypotension (PIH). Consecutive ED patients who underwent emergent oral endotracheal intubation were enrolled in the study. PIH was defined as mean arterial pressure (MAP) < 65 mmHg or required initiation of any vasopressor within 30 min following intubation. Patients were randomly assigned in a 7:3 ratio to a training cohort and a validation cohort. The Least Absolute Shrinkage and Selection Operator (LASSO) and multivariable logistic regression were used to identify risk factors from clinical information and laboratory data performed peri-intubation. Subsequently, a nomogram model was constructed to predict the occurrence of PIH. Model performance was assessed using the area under the receiver operating characteristic curve (AUC-ROC), calibration curve, and decision curve analysis. Furthermore, the nomogram-predicted risks were used to stratify risk. A total of 748 patients were included; the incidence of PIH was 19.5% (146/748). Three independent predictors were included in the nomogram: shock index, pulse pressure, and partial pressure of carbon dioxide. The AUC was 0.790 in the training cohort and 0.805 in the validation cohort. The calibration curve was well-fitted. Decision curve analysis indicated that the model demonstrates significantly better net benefit within the probability range of 1% to 51%. Using X-tile software, 2 cutoff values were identified; patients were categorized as low risk (≤ 0.17), intermediate risk (0.17–0.4), and high risk (> 0.4), with incidence of PIH of 5.8%, 28.4%, and 52%, respectively. A nomogram provides an accurate and practical tool for predicting the risk of PIH in the ED. The model may help identify high-risk patients and timely interventions to decrease the incidence and duration of hypotensive episodes. To the best of our knowledge, this is the first study on a predictive nomogram model for PIH in the ED. The Least Absolute Shrinkage and Selection Operator (LASSO) regression and multivariable logistic regression were used to identify risk factors from hemodynamic parameters and arterial blood gas analysis data performed peri-intubation. Patients who were already hypotensive or receiving vasopressors prior to intubation were excluded from the analysis. We rarely use neuromuscular blockers during induction and lacked external dataset validation.