ABC | Volume 110, Nº4, April 2018

Original Article Cardoso et al Infection in patients with heart failure Arq Bras Cardiol. 2018; 110(4):364-370 color doppler. The comorbidities were identified based on the description of the attending physicians. Renal failure was confirmed by the presence of high levels of urea and creatinine, while diabetes mellitus, by the prescription of hypoglycemic agents on admission. Hypothyroidism was identified in the presence of a prescription of levothyroxine or increased levels of TSH. Atrial fibrillation was diagnosed based on the electrocardiographic tracing, while pulmonary infection was diagnosed based on signs and symptoms, in addition to chest X-ray, blood cell count and C-reactive protein (CRP). Urinary infection was diagnosed based on signs and symptoms, in addition to blood cell count, urinalysis and urine culture. We assessed the characteristics of the patients with infection, and compared them with those of the patients without infection. Statistical analysis The Kolmogorov-Smirnov test was used to test data normality (p > 0.05 = normal distribution). Regarding the characteristics of the population, continuous variables with normal distribution were presented as mean ± standard deviation. Continuous variables without normal distribution were presented as median (interquartile range 25%-75%). Categorical variables were presented as absolute number and percentage. When comparing the groups, the continuous variables were presented as mean ± standard deviation. Unpaired Student t test was used for variables with normal distribution, and Mann-Whitney U test for variables with non-normal distribution. Chi-square test of association or Fisher exact test was used to compare the categorical variables. All tests performed are two-tailed and a p value < 0.05 was considered statistically significant. All statistical analyses were performed with the Statistical Package for the Social Sciences (SPSS) software. Results This study included 260 patients, with a mean age of 66.1 ± 12.7 years, 54.2% of the male sex. The patients were followed up during hospitalization and after discharge. The length of follow-up was 240.05 days (standard error = 10.47, 95% confidence interval = 219.52 – 260.57 days). During hospitalization, 119 patients (45.8%) had infection, 88 (33.8%) being diagnosed with pulmonary infection and 39 patients (15.0%), with urinary infection. Eight patients had both pulmonary 7 and urinary infections concomitantly. Renal failure was present in 142 patients (54,6%), chronic obstructive pulmonary disease (COPD) in 34 patients (13.1%), 7 hypothyroidism in 47 patients (18.1%), diabetes mellitus in 95 patients (36.5%), and atrial fibrillation in 119 patients (45.8%). Table 1 shows the major characteristics of the population studied, of which, 170 patients (65.4%) had HF with left ventricular ejection fraction (LVEF) < 40%, 37 (14.2%) had LVEF of 40%-49%, and 53 (20.4%) had LVEF ≥ 50%. The mean length of stay was 28.6 days (20.52). During hospitalization, 56 patients (21.5%) died. Within 30 days from discharge, 58 patients (28.43%) required a visit to the emergency department, and 28 (13.73%), a new admission. When comparing the groups with and without infection, the following characteristics were similar: age, sex, hemoglobin, blood pressure, heart rate and LVEF (Table 2). Renal failure was present in 73 patients (61.3%) with infection vs 69 patients (48.9%) without infection (p = 0.045). The mean dose of furosemide was similar in both groups: 68.06 mg/day (37.58) for the infected patients vs 71.84 mg/day (39.23) for non‑infected ones (p = 0.568). In the group with infection, 42 patients (35.3%) died during the total follow-up vs 50 patients (35.5%) of the group without infection (p = 0.977). During hospitalization, 32 patients with infection (26.9%) died vs 24 patients without infection (17%) (p = 0.054). When assessing only the discharged patients, 10 of the group with infection (11.5%) died during follow-up vs 26 (22.2%) of the group without infection (p = 0.047). Table 3 shows the characteristics related to in-hospital mortality, and Table 4 shows mortality during total follow-up. Renal failure was observed in 54.6% of the patients, more frequently in those who died during hospitalization (Table 3) or during the total study period (Table 4). Discussion Infection was associated with decompensated HF in 45.8% of the patients, and in that group of infected patients, an increase in mortality was observed during hospitalization. However, after hospital discharge, the group with infection showed better outcome as compared to those without infection. The most frequent comorbidity in our study was renal failure, affecting 54.6% of the patients, and relating to in-hospital mortality during follow-up after discharge. The causes of heart decompensation varied according to the population studied. Acute coronary syndrome, arrhythmias and acute respiratory disease are the most frequent precipitating factors of heart decompensation. 2 At the emergency department of our hospital, the most common cause of hospitalization was non-adherence to treatment, and infections were considered the cause of hospitalization in 8% of the cases. 5 In the BREATHE Registry, poor adherence was also the most frequent cause, and infection was the second one, contributing to decompensation in 22.9% of the cases. 6 The association between infection and decompensation and worse prognosis is well known. An investigation performed at the InCor, via the statistics department, showed that, in the last 10 years, 27,528 patients were hospitalized and diagnosed with HF (I50), most of the male sex (55%). The mean length of stay was 14.8 days, and the in-hospital mortality of that population with HF was 24.8%. 6 In the present study, of the patients admitted to our ward in 2014, the in-hospital mortality of those with HF and infection was 26.9% versus 17.0% of those without infection (p = 0.05). The increase in mortality due to infection has been also reported in the OPTIMIZE-HF Registry. 4 Comparing the characteristics of the patients with and without infection based on the variables analyzed, those with infection decompensated with a milder ventricular impairment than that of those without infection, suggesting that decompensation resulted from the overload and systemic changes that infection causes and not only from the severity of cardiac impairment. Patients with infection 365