Background: The value of signal-averaged ECG (SAECG) in the risk stratification of Chagas disease (ChD), a potentially lethal illness prevalent in Latin America, remains controversial. The aim of this prospective longitudinal study was to determine the prognostic value of SAECG in ChD, using multivariate models with other established prognostic predictors, and to develop a simple prediction risk score.
Methods: The study enrolled 184 ambulatory ChD patients (107 men; age: 48 +/- 12 years) in sinus rhythm and without other systemic diseases. All patients underwent comprehensive evaluation that included clinical examination, ECG, chest X-ray, 24-hour Holter monitoring, echocardiogram, stress testing, and time domain SAECG. Individual medical therapy was adjusted according to a standardized treatment regimen. The association of potential risk factors obtained by noninvasive evaluation and death was tested by Cox proportional-hazards analysis.
Results: During mean follow-up time of 74 +/- 17 months, 13 patients died. Three independent prognostic factors were identified: left ventricular ejection fraction <50% (HR = 5.2, P = 0.048), ventricular tachycardia at either Holter monitoring or stress testing (HR = 9.9, P = 0.036), and prolonged (>150 ms) filtered QRS complex (HR = 4.3, P = 0.035). A prognostic score developed considering the number of risk factors of each patient had an excellent performance in predicting death (c statistic: 0.92).
Conclusions: Prolonged filtered QRS duration obtained by SAECG is an independent predictor of death in ChD. A prediction score including three risk factors, depressed left ventricular ejection fraction, ventricular tachycardia and prolonged filtered QRS complex, has shown to be useful for stratifying risk categories in ChD.