IJCS | Volume 31, Nº3, May/ June 2018

259 Kaufman et al. Mortality in Elderly Individuals Submitted to Coronary Artery Bypass Grafting International Journal of Cardiovascular Sciences. 2018;31(3)258-263 Original Article of the surgical procedure, in the Brazilian population of patients aged 70 years or older. The aim of our study was to evaluate the epidemiological characteristics of patients undergoing coronary artery bypass grafting at a specialized hospital of the Brazilian Unified Health System ( Sistema Único de Saúde - SUS), and to analyze the risk predictors of surgical death and complications inherent to the procedure. Methods Retrospective cohort of the cardiac surgery database of Instituto Nacional de Cardiologia, in Rio de Janeiro (RJ), of patients operated fromDecember 2004 to March 2012. This database included demographic, clinical, laboratory, echocardiographic and angiographic variables of the preoperative, intraoperative and immediate postoperative periods at the Intensive Care Unit (ICU) and the late postoperative period at the infirmary, totaling 327 variables, in addition to vital status at 30 days. In the present study, we assessed the demographic, clinical, laboratory, echocardiographic and angiographic data of the patient’s preoperative period and the need for combined valve surgery, considering the type of evolution (discharge vs. death within 30 days of the procedure). Angina pectoris was classified according to the Canadian Class Society (CCS) criteria. 9 The institutional Research Ethics Committee authorized the performance of this study, approval number 0117/110906. Statistical analysis Statistical analysis was performed by recording the frequencies, the means and their respective standard deviations, or themedianandquartiles,whenappropriate. Student’s t test was used to compare the outcome groups, when the variables showed a normal distribution, or Mann-Whitney U test, when normalitywas not observed. For dichotomous variables, chi-squared or Fisher’s exact test were applied, as appropriate. To study the association between the independent variables and the outcome, two-phase multivariate logistic regression was used and the variables that had p < 0.20 in the first stage were included in the final stage. The StataCorp 14 program (by StataCorp LP) was used. It was considered an alpha value of 0.05. All tests were two-tailed. Results The epidemiological characteristics are shown in table 1. There were 372 patients, most of them males, with a median age of 74.26 years and a mean body mass index of 26.2 kg/m². Regarding the risk factors, the patients had systemic arterial hypertension (93.2%), diabetes (29.0%) and were current smokers (4.3%). Even with a mean creatinine level within the normal range, patients had decreased creatinine clearance. The mortality rate of the study population was 19.35% (72 patients) at 30 days. Table 1 - Demographic and clinical characteristics of the study population Characteristic Age, years (median and interquartile range) 74.26 (71.77-77.08) Men 67.2 Body mass index 26.2 ± 4.1 Systemic arterial hypertension 93.2 Diabetes Mellitus 29.0 Smoking 4.3 Previous acute myocardial infarction 19.4 Previous stroke 5.3 Creatinine clearance 51.77 (42.38-61.86) Creatinine 1.1 (1.0-1.38) Previous coronary artery bypass grafting 2.1 Previous percutaneous coronary angioplasty 9.1 Angina 80.9 Unstable angina 27.6 Stable angina 72.4 Left main coronary artery lesion 41.9 Proximal anterior descending artery lesion 62.0 Trivascular 75.0 Ejection fraction 62.0 (49.0-69.0) NYHA functional class III or IV 9.2 Death at 30 days 19.3 Results expressed in median (interquartile range), percentage or mean ± median. NYHA: New York Heart Association.