Risk prediction of chronic heart failure in patients with type 2 diabetes

Abstract

<b>Introduction.</b> Diabetes mellitus and chronic heart failure (CHF) occur concomitantly, and each disease independently increases the risk for the other. In this connection, we studied the factors associated with CHF progression and proposed a mathematical model for assessing the prediction of the risk of CHF development, which can be applied in real clinical practice. <br><b>Patients and methods.</b> A total of 316 patients with DT2 were examined. The median age was 68 (63; 74) years. The levels of cystatin C, homocysteine, interleukin-6, and albumin/creatinine ratio were assessed in all patients with DT2. To evaluate the efficacy of this model, a retrospective analysis of patients’ medical records was carried out ("test sample"). The discriminant power (ROC) of the model in new cases of CHF was 0.91 (0.87; 1.00). Thus, it indicates a high predictive power of this method. <br><b>Results.</b> A model has been built to predict the likelihood of developing CHF in patients with DT2 (sensitivity 76.2%, specificity 73.3%). <br><b>Conclusions.</b> This model will allow for accurate prediction of the likelihood of developing CHF in patients with DT2. Further studies are needed to use it for choosing the optimal tactics for managing these patients.