LAUSR

We read the article entitled “Value of CHA2DS2-VASc Score for Prediction and Ruling Out of Acute Stent Thrombosis After Primary Percutaneous Coronary Intervention” by Açikgöz et al with interest. They reported that the CHA2DS2-VASc score, stent length, stent diameter, and serum magnesium level were independent predictors of acute stent thrombosis in patients undergoing primary percutaneous intervention for ST-segment elevated myocardial infarction. Although these findings are interesting, we have some concerns about the study. Normally distributed continuous variables are expressed as mean + standard deviation and non-normally distributed continuous variables are expressed as median and interquartile range. However, in this study, all continuous variables including triglycerides and creatine kinase myocardial band values, which are obviously not compatible with the normal distribution, are presented as mean + standard deviation. In addition, the multivariate logistic regression model in Table 3 is misconstructed. Parameters determined as P < .25 as a result of univariate analyzes should be included in the regression model, considering that they may be significant in multivariate logistic regression analysis. With this information, platelet count and triglycerides levels presented in Table 1 are expected to be included in the regression model. Only including P < .1 determined parameters in the regression model as a result of univariate analysis may cause some variables to be overlooked. The most determinant parameters in predicting acute stent thrombosis could be determined using the “Backward LR” method, since there are many possible risk factors according to the sample size. Again, in the multivariate logistic regression analysis, only one of the variables with a high correlation should be included in the model to prevent multicollinearity. However, there is no information about this issue and correlation analysis was not performed in this study. Since the present study has a retrospective and crosssectional design, obtaining information regarding hypertension, diabetes mellitus, and stroke from medical records may have caused errors in the calculation of the CHA2DS2-VASc score. Similarly, the presence of undiagnosed peripheral artery disease at the time of coronary angiography in some patients may have affected the CHA2DS2-VASc score. Furthermore, magnesium levels were only measured once. Conditions that could affect magnesium levels could not be analyzed. The results of this hypothesis-generating retrospective, cross-sectional study should be confirmed by prospective cohort studies.