[Development and validation of risk score model for acute myocardial infarction in China: prognostic value thereof for in hospital major adverse cardiac events and evaluation of revascularization].

Objective: To develop a simple risk score model of in-hospital major adverse cardiac events (MACE) including all-cause mortality, new or recurrent myocardial infarction (MI), and evaluate the efficacy about revascularization on patients with different risk.
Methods: The basic characteristics, diagnosis, therapy, and in-hospital outcomes of 1512 ACS patients from Global Registry of Acute Coronary Events (GRACE) study of China were collected to develop a risk score model by multivariable stepwise logistic regression. The goodness-of-fit test and discriminative power of the final model were assessed respectively. The best cut-off value for the risk score was used to assess the impact of revascularization for ST-elevation MI (STEMI) and non-ST elevation acute coronary artery syndrome (NSTEACS) on in-hospital outcomes.
Results: (1) The following 6 independent risk factors accounted for about 92.5% of the prognostic information: age > or =80 years (4 points), SBP < or =90 mm Hg (6 points), DBP > or =90 mm Hg (2 points), Killip II (3 points), Killip III or IV (9 points), cardiac arrest during presentation (4 points), ST-segment elevation (3 points) or depression (5 points) or combination of elevation and depression (4 points) on electrocardiogram at presentation. (2) CHIEF risk model was excellent with Hosmer-Lemeshow goodness-of-fit test of 0.673 and c statistics of 0.776. (3)1301 ACS patients previously enrolled in GRACE study were divided into 2 groups with the best cut-off value of 5.5 points. The impact of revascularization on the in-hospital MACE of the higher risk subsets was stronger than that of the lower risk subsets both in STEMI [OR (95% CI) = 0.32 (0.11, 0.94), chi2 = 5.39, P = 0.02] and NSTEACS [OR (95% CI) = 0.32 (0.06, 0.94), chi2 =4.17, P = 0.04] population. However, both STEMI (61.7% vs. 78.3%, P = 0.000) and NSTEACS (42.0% vs 62.3%, P = 0.000) patients with the risk scores more than 5.5 points had lower revascularization rates.
Conclusion: The risk score provides excellent ability to predict in-hospital death or (re) MI quantitatively and accurately. The patients undergoing revascularization with risk score greater than 5.5 have lower incidence rates of endpoint.