Verbal autopsy is a tool that originated in low-resource settings in Africa and Asia in the 1950s1, serving to attempt to identify cause of death when medical autopsy was difficult or impossible. It consists of a structured interview of a close relative or caregiver of the deceased, and includes circumstances, symptoms, and other pertinent narrative information regarding the death.2 While reliability and validity of verbal autopsy are debated, it has extraordinary potential as a pragmatic method of assessing cause of death to contribute to vital statistics that could not otherwise be obtained.2
In a recently released article in Pediatrics, Martina Knappett and colleagues from Uganda, Kenya, South Africa, Canada, and the US collaborated in a study that reviewed verbal autopsy data to characterize the cause of post-discharge death among children under age 5 years admitted to 1 of 6 hospitals across Uganda with a diagnosis of suspected sepsis (10.1542/peds.2023-062011). For context, childhood mortality in low- and middle-income countries (LMICs) is a major contributor to global under-five childhood mortality, with the majority of deaths occurring in sub-Saharan Africa, where 1 in 14 children dies before reaching their fifth birthday – a risk 15 times higher than high-income countries.3 Understanding and prevention of deaths is hampered by a lack of vital statistics, due both to resource constraints leading to infrequent health surveys and the community (vs hospital or facility) location of many infant and child deaths.
This study is a secondary analysis of a multi-site prospective study that enrolled children less than 5 years of age with suspected sepsis, conducted July 2017-March 2020 among 6 Ugandan hospitals representative of both rural and urban areas. Children were a median of 5.4 months old at the initial admission, lived a median of 26.2 (range 11.4-41.9) kilometers from a hospital, and died either at home (45.0%) or in-transit seeking care (18.3%) at a median of 27.5 (range 9.0-74.0) days after admission; the majority of their mothers had no schooling (62.3%). These stark demographic barebones helped me understand the critical need for additional information, without which preventive interventions cannot be implemented.
- Of 6,545 children enrolled in the parent study, 6191 (94.6%) survived to hospital discharge,
- Of whom 366 (5.9%) died within 6 months post-discharge and were thus eligible for the study, and
- Of these study-eligible children, 361 (98.6%) had a verbal autopsy completed by trained field officers using a structured interview.
Two physicians with relevant experience reviewed each verbal autopsy report to determine primary cause of death, possible secondary causes of death, and significant contributing factors; discordant results were reconciled by consensus discussion and reviewers assessed the confidence of their conclusions with a 3-point Likert scale (very, somewhat, or not confident).
To share an overview of results:
- The leading cause of neonatal (<28 days of age) post-discharge death was sepsis and other infections (78.9%)
- Among older children (1-59 months) the leading causes were pneumonia (27.7%), sepsis (23.3%) and malaria (8.9%)
- The most common contributors to death were malnutrition (50.5%) and anemia (25.7%)
- Key issues of discussion included additional results, differences over time, study limitations, comparison to other studies and barriers to care
This well-conducted study gives us a window into global pediatrics – the world we share – and gives insights that can contribute to strategic interventions, for example incentivizing follow up of high-risk children with malnutrition and anemia, that may aid in preventing post-discharge mortality in this vulnerable population.
References
- Méthode Pour L’Enregistrement Par Des Non-médecins Des Causes Élémentaires de Décès Dans Des Zones Sous-Développées. WHO; Geneva, Switzerland: 1956
- Bailo P, Gibelli F, Ricci G, Sirignano A. Verbal Autopsy as a Tool for Defining Causes of Death in Specific Healthcare Contexts: Study of Applicability through a Traditional Literature Review. Int J Environ Res Public Health. 2022 Sep 17;19(18):11749. doi: 10.3390/ijerph191811749. PMID: 36142022; PMCID: PMC9517079
- Unicef Data. Under-five Mortality. Updated January 2023. https://data.unicef.org/topic/child-survival/under-five-mortality/# Accessed 9/25/23