Factors associated with postdischarge death in children aged 5–14 years discharged from Tanzanian tertiary hospitals: a prospective observational cohort study

By
Elton Roman Meleki
Stella Mongella
Francis Fredrick Furia
October 8, 2026

Bmj Paediatrics Open

Overview

In this prospective cohort across three Tanzanian tertiary hospitals, 22 of 744 children with known outcomes (2.9%) died within three months after discharge. Mortality was associated with HIV infection, chronic kidney disease, prolonged symptoms, leukocytosis, and discharge against medical advice; more than half of deaths occurred within one month.

Background

Postdischarge mortality is a recognised paediatric outcome in low- and middle-income settings, but evidence has focused mainly on children younger than five years. Estimates vary across studies, in part because populations and follow-up periods differ, and age-specific evidence for older children remains limited. The study addresses this evidence gap for Tanzanian children aged 5–14 years discharged from three tertiary hospitals. Its findings describe the timing, probable causes, and factors associated with deaths during the three months after discharge.

Data Highlights

Measure	Finding
Study design and follow-up	Prospective cohort; 10-month enrolment period and telephone follow-up at 1, 2, and 3 months after discharge
Participants	786 children aged 5–14 years; median age 7 years (IQR 5–9); 467 (59.4%) male
Known postdischarge outcomes	744 participants
Deaths within 3 months	22/744 (2.9%)
Deaths within 1 month	More than half of deaths
Deaths occurring at home	13/22 (59.1%)
Leading probable causes of death	Pneumonia and malignancies, each 18.2%
Factors associated with postdischarge mortality	HIV infection: HR 9.83 (95% CI 2.3–41.5); chronic kidney disease: HR 18.26 (95% CI 1.3–39.26); symptoms lasting >2 weeks: HR 4.7 (95% CI 1.5–14.73); leukocytosis: HR 7.94 (95% CI 1.97–32.03); discharge against medical advice: HR 55.4 (95% CI 4.46–393.11)

Key Findings

Among 744 children with known outcomes, 22 (2.9%) died during the three-month postdischarge follow-up.
More than half of deaths occurred within one month of discharge, and 13 (59.1%) occurred at home.
Pneumonia and malignancies were the leading probable causes of death, each accounting for 18.2%.
HIV infection, chronic kidney disease, symptoms lasting more than two weeks, and leukocytosis were associated with postdischarge mortality in Cox regression analyses.
Discharge against medical advice was also associated with mortality (HR 55.4; 95% CI 4.46–393.11).
The source background reports that prior studies of older children in Mozambique and Kenya found postdischarge mortality of 2.2% and 3%, respectively, while a Tanzanian cohort of children aged 5–12 years reported 16% over one year; follow-up duration and study populations differed.

Clinical Implications

The study authors state that these data may guide postdischarge follow-up by identifying children at risk and informing appropriate follow-up timing. The observed concentration of deaths in the first month and associations with chronic illness and discharge against medical advice are findings relevant to follow-up planning, but the study does not establish the effectiveness of any specific intervention.

Conclusion

In this cohort of Tanzanian children aged 5–14 years, postdischarge mortality was 2.9% among those with known outcomes, with most deaths occurring within one month. Mortality was associated with several clinical factors and discharge against medical advice.