Análise hierarquizada dos fatores associados à readmissão hospitalar por doenças cardiovasculares

Abstract

This study aims to analyze factors associated with readmissions for cardiovascular disease in a hierarchical perspective. It is a correlational case-control study. It was conducted from November 2015 to April 2016 at a cardiovascular diseases specialized hospital. It included 306 participants, 153 were cases with previous history of hospital admissions for cardiovascular disease, and 153 were controls without that history. Those following data were collected: sociodemographic as distal elements to the outcome; living habits and use of health services as intermediate level; and physical measures as proximal characteristics. It was approved in Ethical Committee by the number 1.269.232. The statistical model of hierarchical analysis was constructed using the backward method, initially including the variables from the distal level with p-value below 0.1. Then it were included variables of the intermediate level with the addition of those significant ones of step 1, keeping them in the model. It was eliminated not significant ones, considering 20% as entry criterion and 10% of significance to stay. Step 3 was similar to 2, considering the variable proximal level and maintaining significant ones in the previous step. Finally, after the elimination of variables, it was estimated the final hierarchical model, with a significance level of 5%. The final model included female gender as a risk factor (OR: 1.06) and age as protective (OR: 0.96) in the distal level. The intermediate level was composed of the smoking cessation time (OR: 0.53), physical activity (OR: 0.07; 95% CI: 0.008 to 0.061), stress (OR: 11.9; 95% CI: 3.7 to 38.5), admissions for other causes (OR: 26; 95% CI: 9.9 to 67.8) and limited mobility to basic unit of primary care (OR: 4.3; 95% CI: 1.3 to 13.9). Finally, the proximal level amounted hypertension (OR: 21.4; 95% CI 1.34 to 34.2), glucose levels (OR: 1.008; 95% CI: 1.00 to 1.017) and hospitalization for unstable angina (OR: 114.6; 95% CI 4.7 to 287.9). It was concluded that the study achieved the goal of create a hierarchical model adjusted for the Brazilian reality. Those causal factors require interventions especially in primary care, supported by public policies in the medium to long term. The identification of the risk of hospital readmission for cardiovascular disease may be a milestone for the development of guidelines for application of resources in specific aspects of care, with the power to prevent or delay the occurrence.