Chile is past vector control for Chagas. Its cohort shows what comes next.

Chagas disease in Chile has shifted toward what a study published on 14 August calls a post-vector-control, aging-cohort scenario — one in which chronic sequelae increasingly shape outcomes [s1]. The study quantifies that phase using the full national notification record.

The cohort

Researchers assembled a nationwide retrospective cohort of confirmed Chagas cases notified in Chile between 2007 and 2021 — 17,508 individuals — and linked them to mortality and hospital discharge records [s1]. The primary outcome was Chagas-specific death, with death from other causes treated as a competing event rather than as censoring, using cumulative incidence functions, Gray's test, and Fine-Gray subdistribution hazard models, plus complementary cause-specific Cox models [s1].

That methodological choice is the study's backbone. In an ageing cohort, people die of other things, and a conventional survival analysis that treats those deaths as censored will overstate disease-specific risk. Competing-risks methods are the correct tool for exactly this situation.

The headline result is that most of these deaths are not from Chagas

Of the 17,508 individuals, 261 (1.49%) died from Chagas disease and 1,021 (5.83%) from other causes [s1]. The competing risk is nearly four times larger than the outcome of interest.

Ten-year cumulative incidence of Chagas-specific death was higher in men (2.66%) and rose sharply with age, reaching 4.80% at ages 65 to 74 and 8.89% at 75 and over [s1].

Digestive disease, not cardiac disease, carried the highest cumulative risk

The finding most likely to surprise readers familiar with Chagas as a heart disease: cumulative incidence was highest for chronic digestive involvement — ICD-10 code B57.3 — at 6.49%, compared with 2.93% for chronic cardiac involvement (B57.2) and 1.47% for the code covering infection without specified organ involvement (Z22.8), with Gray's test P < .005 [s1].

The adjusted models ran the same way. In Fine-Gray models, Chagas-specific mortality was associated with each 10-year increase in age (subdistribution hazard ratio 2.31), male sex (sHR 1.47), cardiac involvement (sHR 1.92) and digestive involvement (sHR 3.36) [s1].

This does not mean digestive Chagas is more dangerous than cardiac Chagas in general. It means that among people notified in Chile and coded with digestive involvement, Chagas-specific death was more likely — which is heavily shaped by how cases are coded and who receives which diagnosis. A cardiac Chagas patient who dies of heart failure may plausibly be coded to a cardiac cause rather than to Chagas, which would move them out of the numerator entirely. The study cannot rule that out.

The hospitalisation signal

Cause-specific Cox models showed a graded association between recurrent hospitalisations and Chagas-specific death, reaching a hazard ratio of 83.94 for patients with three or more admissions [s1].

An HR of that magnitude is not a risk factor in any useful sense; it is a marker of the state a patient is already in. The authors describe it as consistent with advanced clinical deterioration [s1], which is the right reading. Repeated admission is downstream of the disease process, not upstream of it, and the number is best understood as confirming that the deaths are concentrated among visibly deteriorating patients rather than arriving unheralded.

What the study is for

The conclusion is operational: although Chagas-specific death was uncommon, it was concentrated among identifiable high-risk groups, and risk-stratified chronic care with structured follow-up should prioritise patients with organ involvement or recurrent hospitalisation [s1].

That is a statement about running a chronic disease service rather than about interrupting transmission — the question that follows once vector control has largely done its work. It is a question other countries in the region will face as their own control programmes mature, which is what gives a Chilean cohort study relevance beyond Chile.

Limits

This is a retrospective cohort built from notification, mortality and discharge records, so it describes people who were diagnosed and notified. Anyone infected and never tested is absent from it, and the study offers no estimate of how many such people there are.

Cause of death is taken from death certification, with all the classification variability that entails — and in a disease whose fatal manifestations are heart failure, arrhythmia and gastrointestinal complications, the boundary between "died of Chagas" and "died of a condition Chagas caused" is exactly where certification is least reliable.

The study reports associations, not treatment effects. It does not report whether patients received antitrypanosomal therapy, when, or with what result, so nothing here speaks to the effect of treatment on the outcomes measured.

What to watch

Whether other South American countries with controlled vector transmission publish comparable competing-risk analyses. Chile's dataset is unusually complete; the value of the finding depends on whether the pattern — deaths concentrated in older men with organ involvement and repeat admissions [s1] — reproduces elsewhere.