ABC | Volume 110, Nº5, May 2018

Original Article Almeida et al MTWA and malignant arrhythmias in Chagas disease Arq Bras Cardiol. 2018; 110(5):412-417 Table 1 – Characteristics of the sample. All (96) Case group (45) Control group (51) p Number of patients 96 45 51 - Male sex * 48 19 29 0.220 Mean age (years) 55 62 49 - Age > 60 years * 27 26 1 < 0.001 Mean ejection fraction (%) 48,8 39 58 - Reduced ejection fraction (< 45%) * 37 31 6 < 0.001 Beta-blocker use * 47 37 10 < 0.001 *Number of patients Table 2 – Factors related to the presence of ventricular arrhythmias in the multivariate logistic regression model. p OR 95%CI LL 95%CI UL MWTA 0.044 5.17 1.05 25.51 Beta-blocker 0.139 3.73 0.65 21.40 Sex 0.118 0.27 0.05 1.39 LVEF 0.011 0.91 0.85 0.98 Age 0.005 1.13 1.04 1.22 LL: lower limit; UL: upper limit; MWTA: microvolt T-wave alternans; LVEF: left ventricular ejection fraction. Statistical analysis Initially, the case and control groups were compared regarding their clinical characteristics by use of Fisher exact test. The variables tested were sex, age (older or younger than 60 years), reduced or preserved LVEF, and beta-blocker use. There was a significant disparity between the groups, and to assess the association between MTWA and the occurrence of malignant ventricular arrhythmias, multiple logistic regression models were adjusted, including the potential confounding covariables. The covariables age and LVEF were entered into the model continuously. The model calibration was assessed by use of Hosmer-Lemeshow test. The results were expressed as odds ratio (OR) with its respective confidence interval. The significance level adopted was α = 0.05. All analyses were performed with the R statistical software, 3.3.2 version. 27 Result This study recruited 96 patients with CCC as follows: 45 patients (46.8%) with an ICD, constituting the case group; and 51 (53.1%) without an ICD and no known history of ventricular arrhythmia, constituting the control group. Table 1 describes the sample. Of the total sample, 48 patients (50%) were of the male sex, 42.2% of the case group and 53.1% of the control group, p = 0.220. Of the patients with an ICD, 57.8% were older than 60 years, while of those with no ICD, only 1.96% were older than 60 years, p < 0.001. Of the total sample, 37 patients had reduced LVEF (38.5%), 31 patients (68.9%) in the case group, and 6 patients (19.6%) in the control group, p < 0.001. In addition, the distribution of beta-blocker users was as follows: 37 patients in the case group (82.2%), and 10 patients in the control group (19.6%), p < 0.001. The MTWA test had a non-negative result in 51 patients (53.1%) as follows: 36/45 patients (80%) in the case group and 15/51 patients (29.4%) in the control group, OR = 9.60 (95%CI: 3.41 – 27.93). Because of the difference in characteristics between the groups, a logistic regression model was created to correct the disparities between them, including age, sex, LVEF and beta-blocker use. Table 2 shows the results of data analysis. The model showed that the difference is statistically significant between the case and control groups regarding the result of the MTWA test [OR = 5.17 (95%CI: 1.05 – 25.51)]. The Hosmer-Lemeshow test showed good calibration of the model (p = 0.872). Discussion In this case-control study with adjustments for other significant variables, we observed the association between the non-negative result of the MTWA test and the occurrence of ventricular tachyarrhythmias in patients with CCC, with OR = 5.17 (95%CI: 1.05 – 25.51), suggesting that MTWA may play a role in the assessment of the risk for sudden death of patients with Chagas heart disease. The occurrence of ventricular tachyarrhythmias seems more common in Chagas disease than in heart diseases of other etiologies. 4 However, there is neither a method nor a score to properly identify patients at risk for sudden death due to those arrhythmias. 414