What CHAMPS Learned From 8,500 Child and Stillbirth Death Reviews

What the CHAMPS summary covers

The U.S. Centers for Disease Control and Prevention published a CHAMPS Surveillance Summary on 10 September 2026. It describes mortality surveillance conducted from December 2016 through December 2024 in population-defined catchment areas in Bangladesh and six African countries: Ethiopia, Kenya, Mali, Mozambique, Sierra Leone and South Africa.

CHAMPS stands for Child Health and Mortality Prevention Surveillance. The network was designed to produce more specific cause-of-death evidence in high-mortality settings, where vital registration and verbal autopsy may not identify pathogens, maternal contributors or several conditions in the same causal chain.

This was a descriptive surveillance programme, not a randomized intervention trial. Its results describe enrolled deaths in selected catchments and should not be read as national prevalence estimates.

From 18,784 eligible deaths to 8,500 reviews

CHAMPS identified 18,784 eligible deaths. Families consented to enrolment for 15,612 deaths, or 83.1%. Minimally invasive tissue sampling was completed for 9,415 deaths, and multidisciplinary cause-of-death review was completed for 8,500, representing 90.3% of deaths investigated with tissue sampling.

The reviewed deaths included 3,199 stillbirths, 3,230 neonatal deaths and 2,071 deaths among infants and children aged 28 days to under five years. These groups differ in biology, clinical pathways and preventable opportunities, so the report analysed them separately.

Consent, timing, refrigeration, notification and operational feasibility influenced whether tissue sampling could occur. A complete review therefore represents a deeply investigated subset, not every death in the catchment areas.

How a CHAMPS review is assembled

Trained teams combined minimally invasive tissue sampling with microbiology, molecular testing, histopathology, clinical-record abstraction and verbal autopsy interviews. Multidisciplinary panels then reviewed the evidence using standardized frameworks.

Panels could identify an underlying cause, other conditions in the causal chain, contributing maternal conditions, comorbidities and health-system factors. This multi-cause approach is why several percentages in the report overlap and should not be added together.

CHAMPS also used community engagement and bereavement-sensitive consent. Families could be offered a clinical explanation of findings and referral support. These procedures helped make postmortem investigation feasible, but consent and local acceptability still shaped inclusion.

Preventability was assessed for 7,558 deaths

Of 7,558 deaths with a preventability assessment, 6,103, or 80.7%, were classified as preventable or possibly preventable through improvements in already available maternal, newborn and child-health interventions.

This does not mean one intervention would have prevented every death. It was a panel assessment based on the causal pathway and opportunities such as antenatal care, obstetric management, timely clinical care and infection prevention. The denominator excludes reviewed deaths without a preventability assessment.

The result is relevant to service planning in the studied catchments, but it is not a national rate for Bangladesh, an Africa-wide rate, an India estimate or a global proportion.

Causes differed by age group

Stillbirths were predominantly attributed to perinatal asphyxia or hypoxia, reported in 79.1%, often alongside maternal hypertensive disorders, placental abnormalities, chorioamnionitis and other conditions.

Among neonatal deaths, the report identified preterm complications in 39.7%, asphyxia or hypoxia in 37.7%, and sepsis in 36.5%. Among infants and children, frequently identified causes included lower respiratory infections in 37.4%, sepsis in 36.9%, malnutrition in 27.3%, malaria in 22.1% and diarrhoeal disease in 17.3%.

Multiple conditions were common: 44.3% of neonatal deaths and 67.9% of infant and child deaths had two or more conditions in the causal chain. Excluding stillbirths, infection contributed to 3,146 deaths, or 59.3%. Among 2,749 deaths with one or more pathogens in the causal chain, 44.5% were polymicrobial.

Nutrition and maternal health were part of the same picture

Postmortem measurements indicated substantial undernutrition among infants and children with available data. Non-mutually-exclusive indicators of underweight and wasting were each reported at 61.3%, and stunting at 43.0%.

A main maternal condition, most commonly placental complications or hypertensive disorders of pregnancy, was identified for 71.1% of stillbirths and 58.2% of neonatal deaths. These findings support integrated thinking about maternal care, newborn care, infection control, nutrition and paediatric services. They do not establish that one factor alone caused every death in which it appeared.

Why selected catchments are not countries

CHAMPS sites were selected through a multistage process considering mortality, geography, research infrastructure and government commitment. The sites were population-defined, but they were not random samples of entire countries.

Deaths also had to meet residence rules, be identified by local systems and, for tissue sampling, be reported quickly enough. Families had to consent. These features create selection and feasibility limits. The findings reveal detailed causal pathways within the surveillance system, but cannot establish the frequency of each cause across every community in the seven countries.

What the Bangladesh site means for India

Bangladesh gives the report regional South Asian relevance for mortality surveillance, maternal and newborn services, infection prevention, diagnostic capacity and nutrition. However, no Indian catchment was included.

The India-facing lesson is methodological: stronger death notification, diagnostic evidence, multidisciplinary review and respectful community engagement can produce more actionable information than a single nonspecific cause. The percentages should not be transferred to India without Indian data.

What this surveillance can and cannot show

CHAMPS can show which causes and contributors were identified in investigated deaths, how often multiple conditions appeared, and which service opportunities review panels considered relevant. It can demonstrate that standardized postmortem surveillance is feasible in participating resource-constrained settings.

It cannot prove that adopting one programme would reduce mortality by a particular amount, compare randomized intervention groups, measure national prevalence, eliminate selection bias, make a personal diagnosis or explain an individual family’s loss outside a formal clinical process.

What would require an update

This page should be updated if CDC materially revises the summary, CHAMPS publishes later years, new national sites become mature enough for inclusion, methods change or intervention evaluations test whether identified opportunities improve outcomes.

Medical review status: Pending. MedoPulse has checked the cited CDC source, but this source-led page has not been reviewed by a MedoPulse clinician.

Educational notice: MedoPulse provides general educational information, not diagnosis, treatment or individualized medical advice. Cause-of-death conclusions require formal clinical and surveillance processes.