Author + information
Accurate risk stratification of patients with acute myocardial infarction is important due to great variability in mortality risk. However, no risk prediction models up to date included a large cohort of Chinese patients, accounting for approximately one fifth of the world population.
Data were extracted from the China Acute Myocardial Infarction (CAMI) registry between January 2013 and September 2014. A total number of 23417 patients with AMI were identified and enrolled in our study (STEMI: 17622, NSTEMI: 5795). Primary outcome was in-hospital mortality. Patients were divided into a derivation cohort of 17563 patients for the development of a multivariate logistic regression model to predict in-hospital mortality risk, and a validation cohort of 5854 patients for validation of the model. A simplified integer risk score was also developed to allow for practical clinical use.
A total of 16 variables were independently associated with in-hospital mortality and were included in our model: age, female, hypertension, hyperlipidemia, BMI, smoking status, systolic blood pressure, heart rate, serum creatinine level, white blood cell count, serum potassium level, serum sodium level, Killip classification, heart arrest, ST-segment elevation of ECG, anterior wall involvement. Within the derivation cohort, the C-statistic for our model was 0.83 (95% confidence interval [CI]: 0.82 to 0.84). The C-statistic for the simplified score was 0.80 (95% CI 0.79-0.82) with excellent calibration (Hosmer-Lemeshow p=0.10). Within the validation cohort, the score also showed good discrimination (C-statistic 0.80, 95% CI: 0.78 to 0.83, HL P = 0.34). Event rates increased significantly as the CAMI score, classified by tertile, increased in the derivation, validation as well as the whole cohort.
We developed and validated a novel risk prediction tool to accurately predict the risk of in-hospital mortality of patients with AMI in China.
CORONARY: Acute Myocardial Infarction