Establishment of a nomogram model for acute chest pain triage in the chest pain center

BackgroundAcute myocardial infarction (AMI) is the leading life-threatening disease in the emergency department (ED), so rapid chest pain triage is important. This study aimed to establish a clinical prediction model for the risk stratification of acute chest pain patients based on the Point-of-care (POC) cardiac troponin (cTn) level and other clinical variables.MethodsWe conducted a post-hoc analysis of the database from 6,019 consecutive patients (excluding prehospital-diagnosed non-cardiac chest pain patients) attending a local chest pain center (CPC) in China between October 2016 and January 2019. The plasma concentration of cardiac troponin I (cTnI) was measured using a POC cTnI (Cardio Triage, Alere) assay. All the eligible patients were randomly divided into training and validation cohorts by a 7:3 ratio. We performed multivariable logistic regression to select variables and build a nomogram based on the significant predictive factors. We evaluated the model's generalization ability of diagnostic accuracy in the validation cohort.ResultsWe analyzed data from 5,397 patients that were included in this research. The median turnaround time (TAT) of POC cTnI was 16 min. The model was constructed with 6 variables: ECG ischemia, POC cTnI level, hypotension, chest pain symptom, Killip class, and sex. The area under the ROC curve (AUC) in the training and validation cohorts was 0.924 and 0.894, respectively. The diagnostic performance was superior to the GRACE score (AUC: 0.737).ConclusionA practical predictive model was created and could be used for rapid and effective triage of acute chest pain patients in the CPC.

Automated summary

This study aimed to establish a clinical prediction model for risk stratification of acute chest pain patients based on rapid cardiac troponin testing and other clinical variables. The model, which included variables such as ECG results, troponin levels, hypotension, chest pain symptoms, Killip class, and gender, showed superior diagnostic accuracy compared to the GRACE score.