TABLE II. Key features of the original protocol and differences in the final protocol, which were agreed by most members of the BENznidazole Evaluation for Interrupting Trypanosomiasis (BENEFIT) steering committee.

	Original protocol	Final BENEFIT protocol
Eligibility criteria
Age	18-50 years	18-75 years
Evidence of chronic Chagas cardiomyopathy	Positive serological tests for T. cruzi, and electrocardiographic or echocardiographic alterations, or both, that were characteristic of chronic Chagas cardiomyopathy	Same criteria
NYHA functional class	I or II, exclude those with congestive heart failure (NYHA III or IV)	I, II or III
Living in conditions that predispose to Trypanosoma cruzi infection	Exclude	Include
Previous resuscitation following cardiac arrest	Exclude	Include
Previous sustained ventricular tachycardia	Exclude	Include
Previous insertion of a pacemaker or cardiac defibrillator	Exclude	Include
Previous admission to hospital for heart failure	Exclude	Include
Previous thromboembolic event	Exclude	Include
Endpoints
Primary	Composite of time to cardiovascular death, resuscitated cardiac arrest, sustained ventricular tachycardia, insertion of a pacemaker or cardiac defibrillator, admission to hospital for heart failure, and development of thromboembolic events.	Composite of time to death, resuscitated cardiac arrest, sustained ventricular tachycardia, insertion of a pacemaker or cardiac defibrillator, cardiac transplantation, and development of new heart failure, stroke, or systemic or pulmonary thromboembolic events.
Secondary	Composite of electrocardiographic and echocardiographic changes, as markers of disease progression (surrogate endpoints), throughout the study period. Eventual differences in outcomes between individual countries	Secondary outcomes also included the response to treatment on the basis of results on PCR assay.
Statistical analysis
Sample size	3000 patients (1500 per group) needed to detect a 20% reduction in the relative risk of the primary endpoint in the benznidazole group with 90% power, assuming a 5-year event rate of 30% in the placebo group (at a two-sided α of 0.05). Expect to lose 20% of patients from non-compliance or during follow-up	3000 patients (1500 per group) needed to detect a 26% reduction in the relative risk of the primary endpoint in the benznidazole group with 90% power, assuming a yearly event rate of 8% in the placebo group and 4-6 years of follow-up (at two-sided α of 0.05). Expect to lose 17% of patients from non-compliance and 3% during follow-up.

NYHA: New York Heart Association.