Open population approach to identify risk factors and detection of cases of heart failure in Mexico

Abstract

Abstract Heart failure (HF) is one of the cardiovascular diseases that in recent years has been a cause of high morbidity and mortality, worldwide it is estimated that 2% of the population lives with this heart condition. However, Mexico, despite having a high incidence of diseases associated with the development of HF, there are no clinical-epidemiological studies of the pathology. The present project is an epidemiological study on HF carried out by the civil association with the aim of identifying risk factors (RF) associated with the clinic that allow early diagnosis as well as transmitting guidelines that can prevent the development of the disease and/or its complications in the open population (OP) of Mexico. Cardiological assessments were conducted in some states of Mexico as a pilot test of the model designed by SL called Latidos Fest (LF). Each site had the following activities like vital signs, cardiac assessment, electrocardiogram and if indicated, echocardiogram. Those who attended the assessment had to sign an informed consent form and at the end received a summary with indications according to the clinical-epidemiological findings. With the data obtained, an Excel database was created in which descriptive analyses were also performed. The statistical model Factor Analysis of Mixed Data (FAMD) was generated to identify quantitative and qualitative variables associated with the presence of HF and a classification tree model to establish the hierarchical value of the variables and their predictive value for the risk and detection of HF cases. For both statistical models RStudio software was used. A total of 949 cardiac evaluations in OP were performed thanks to the NGO of the LF in 7 different states of Mexico during 2022. In 68.1% of the population, at least 1 RF for developing HF was detected, 16.33% were identified as having HF with no previous diagnosis (PD), 0.74% already had a PD and 14.12% were apparently healthy. The demographic variables with the highest percentage are shown in table 1. From the FAMD model, the variables in which the individuals coincided most are shown in figure 1-A. When performing the classification trees, the overall prediction associated (OPA) to cases with and without HF was 75% confidence of which, the accuracy of classifying a case with HF was 67.4% and without HF was 82.2%; for the OPA for the risk of developing HF and not having HF the confidence was 82%, in this the accuracy of classifying without HF was 15%, however, the accuracy of classifying someone at risk of developing HF was 95.8% (Figure 1-B). Thanks to the cardiology assessments in OP that were performed in each LF, risk factors and new cases of HF were identified. With the findings in the models, it is possible to intentionally search for people with the factors found, however, it is necessary to expand the cardiological screening in other populations to check if the clinical-epidemiological parameters match or are different.Table 1.Percentage of demographic variaFigure 1.Results of the models. A. FAMD