Several statistical models are used in medical science applications. In the present study, the binary logistic regression model is used to pin point the significant risk factors affecting the occurrence of a second myocardial infarction (MI). The model is applied to 1500 patients who were initially treated for first MI and have been followed up after at least two years from the first MI treatment, and occurrence of a second MI was observed. Several demographic and medical covariates were used in the analysis. The probability of occurrence, odds ratio for having or not having a second MI is obtained. Results show that three risk factors affect the second occurrence of myocardial infarction, these factors are family history, congestive heart failure, and smoking. Analysis was performed also for males and for females. Same risk factors were reached for males, but only the later two factors were reached for females. Odds ratios for males with “congestive heart failure” are susceptible to a second MI from 14 times to 72 times, while females susceptibility ranges from 14 to 247 times compared to those who did not experience a second MI. The interaction between gender and family history proved to be a significance risk factor. Classification tables show that correct classification is approximately 93% for the three analyses, and ROC curves exhibit acceptable classification. The study recommends the study of more factors to the analysis such as treatment types at first MI, building up a database for MI patients, and conducting an awareness health program for patients who have had a first MI.

logistic regression model, Wald test, odds ratio, dichotomous covariates, score test, likelihood ratio (LR) test, ROC curve, myocardial infarction, MI.