Diagnostic Model of in-Hospital Mortality in Patients with Acute ST-Segment Elevation Myocardial Infarction Used Artificial Intelligence Methods

Abstract

Objective: To develop and externally validate a diagnostic model of in-hospital mortality in the population of unselected real-world patients with acute ST elevation myocardial infarction (STEMI ). Design: Multivariable logistic regression of a cohort of hospitalized patients with acute STEMI . Setting: Emergency department ward of a university hospital. Participants: Diagnostic model development: Totally 2183 hospitalized patients with acute STEMI from January 2002 to December 2011. External validation: Totally 7485 hospitalized patients with acute STEMI from January 2012 to August 2019. Outcomes: In-hospital mortality. Results: Totally 61 (2.8%) patient died in the development dataset and 127(1.7%) patient died in the validation dataset. The strongest predictors of in-hospital mortality were age and Killip classification. We developed a diagnostic model of in-hospital mortality .The area under the receiver operating characteristic ( ROC ) curve ( AUC ) was 0.9126 0.0166, 95% confidence interval( CI )= 0.88015 ~ 0.94504 in the development set .We constructed a nomograms using the development database based on age and Killip classification. The AUC was 0.9305 0.0113,95% CI= 0.90827 ~ 0.95264 in the validation set . Discrimination, calibration ,and decision curve analysis were satisfactory. Conclusions: We developed and externally validated a strong diagnostic model of in-hospital mortality in patient with acute STEMI . We registered this study with WHO International Clinical Trials Registry Platform (ICTRP) (registration number: ChiCTR1900027129; registered date: 1 November 2019). http://www.chictr.org.cn/edit.aspx?pid=44888&htm=4. Key Words: coronary disease ST elevation myocardial infarction in hospital mortality nomograms