Deaths with preceding hospitalisations within 180 days in eight countries in sub-Saharan Africa and South Asia: A secondary descriptive analysis of the Child Health and Mortality Prevention Surveillance (CHAMPS) network

Objectives To describe (1) the proportion of deaths that were in recently hospitalised children and (2) causes of mortality among deceased children aged 0–59 months with preceding hospitalisations who enrolled in a mortality surveillance programme. Design Descriptive study using prospectively collected data. Setting Eight Child Health and Mortality Prevention Surveillance (CHAMPS) community and healthcare sites in sub-Saharan Africa and South Asia. Participants Deaths among children aged 0–59 months enrolled in CHAMPS 2016–2023. Interventions None. Primary and secondary outcome measures Deaths with antecedent hospitalisations within 180 days of death. Causes of death determined by expert panels who reviewed clinical data and histopathologic and microbiologic results from postmortem minimally invasive tissue sampling. Results CHAMPS enrolled 8548 deaths; we excluded 3688 neonates who died before discharge or ≤24 hours of birth and 482 with unclear information on antecedent hospitalisations. Out of the 4378 remaining deaths, 16.7% (95% CI 15.7% to 17.9%) were deaths that occurred within 180 days of a hospitalisation (n=733/4378). Of these, 55.7% (95% CI 52.0% to 59.3%) occurred outside healthcare facilities. Among included deaths with minimally invasive tissue sampling completed (n=337), lower respiratory tract infections (41.2%, 95% CI 36.0% to 46.7%), sepsis (39.8%, 95% CI 34.5% to 45.2%) and undernutrition (n=92, 27.3%, 95% CI 22.7% to 32.4%) were most common causes of death among cases with antecedent hospitalisations. The greatest proportion of deaths with antecedent hospital admissions occurred among cases aged 1–11 months (48.0%, 95% CI 44.4% to 51.7%), compared with those aged 0–1 months (21.7%, 95% CI 18.8% to 24.9%) and those aged 1–5 years (30.3%, 95% CI 27.0% to 33.8%). Moreover, the greatest proportion of deaths with antecedent hospital admissions occurred among infants/children with weight-for-age Z-score of <−3 (62.5%, 95% CI 56.5% to 68.0%) compared with those with weight-for-age Z-score of ≥−3 (37.5%, 95% CI 32.0% to 43.5%). Conclusions We observed a high proportion of deaths with antecedent hospitalisations within 180 days among young children across eight sites in sub-Saharan Africa and Asia. Among those deaths, children aged 1–11 months and undernourished infants were over-represented, suggesting early follow-up as a potential point to focus targeted support and future research.