Last update:


Arch Hellen Med, 32(1), January-February 2015, 92-95


Linear regression models in predicting acute myocardial infarction progression from serum levels of TNI, CRP and AST

K. Poulakis,1 M. Chouzouris,2 A. Konstantopoulos,3 K. Daskalopoulou,4 P. Xenos,3 N. Polyzos5
1Department of Computer and Information Science, Linköping University, Linköping, Sweden,
2Barcelona Graduate School of Economics (GSE), Barcelona, Spain,
3Department of Statistics and Insurance Science, University of Pireus, Pireus,
4Department of Microbiology, "Elpis" General Hospital of Athens, Athens,
5Department of Social Management, University of Thrace, Komotini, Greece

Acute myocardial infarction (AMI) is often followed by complications, such as a second infarction, chronic arrhythmia or heart failure. The ability to predict the progression of an AMI is therefore of importance. An attempt was made to create statistical models using the serum levels of C-reactive protein (CRP), troponin I (TnI) and aspartate aminotransferase (AST), and to explore the contribution of the sex of the patient. A sample of 268 patients diagnosed with AMI in the Cardiology Clinic of the Athens General Hospital "Elpis" (152 women and 116 men) were tested on admission and again after 24 hours. No significant correlation was demonstrated between serum levels of CRP, TnI and AST and sex. A transmitted linear regression model for TnI was shown to be an acceptable predictor of AMI progression (Rsquare= 0.767). Bearing in mind the significance of TnI as a cardiac-specific factor, the ability to predict its serum levels offers a rough estimate of the damage caused by the AMI.

Key words: AST, CRP, Forecasting, Myocardial infarction, TnI.

© Αρχεία Ελληνικής Ιατρικής