Rb. Bestetti et al., NONINVASIVE PREDICTORS OF MORTALITY FOR PATIENTS WITH CHAGAS HEART-DISEASE - A MULTIVARIATE STEPWISE LOGISTIC-REGRESSION STUDY, Cardiology, 84(4-5), 1994, pp. 261-267
This prospective investigation was conducted in an attempt to identify
noninvasive predictors of mortality for patients with Chagas' heart d
isease through a multivariate stepwise logistic regression study. Fift
y-six patients with a positive complement fixation test for Chagas' di
sease were followed up at the Cardiomyopathy Clinic of our institution
from April 1990 to April 1992. Patient age was 59 +/- 17 years; 28 (5
0%) were male. Upon admission, 19 patients (33%) were in the New York
Heart Association (NYHA) class III and 8 (14%) in the NYHA class IV, S
ystolic blood pressure was 125 +/- 23 mm Hg, diastolic blood pressure
76 +/- 11 mm Hg and resting heart rate 77 +/- 11 beats/min. Forty pati
ents (71%) were given digitalis and 39 (69%) angiotensin-converting en
zyme inhibitors. Plasma Na+ was 140 +/- 4 mEq/l, K+ was 4.34 +/- 0.73
mEq/l and creatinine level 1.34 +/- 0.31 mg/100 ml. Cardiomegaly was o
bserved in the chest X-ray of 41 of 51 (79%) available patients. Atria
l fibrillation was observed in the resting ECG of 24 of 54 (44%) avail
able patients, premature ventricular contractions in 23 (41%), right b
undle branch block in 26 (46%) and left anterior hemiblock in 26 (46%)
patients. Echocardiography revealed a left ventricular ejection fract
ion of 0.45 +/- 0.16, left ventricular systolic dimension of 51.23 +/-
13.53 mm and left ventricular diastolic dimension of 62.94 +/- 19 mm.
Sixteen (28%) patients died during the 2-year study, 11 of them sudde
nly. By univariate analysis, left ventricular ejection fraction (p = 0
.03), left ventricular diastolic dimension (p = 0.03), NYHA class IV (
p = 0.0004) and digitalis use (p = 0.04) were found to be associated w
ith mortality. In the multivariate model, however, only left ventricul
ar ejection fraction was retained as an independent predictor of morta
lity. Actuarial survival was 75% for patients with left ventricular ej
ection fraction >0.30, and 40% for patients with left ventricular ejec
tion fraction <0.30 (p = 0.03). We conclude that patients with Chagas'
heart disease having a left ventricular ejection fraction <0.30 deter
mined echocardiographically are at very high risk of dying.