Validation of malaria-attributed deaths using verbal autopsy studies: a systematic review

• Published in: Malaria Journal
• Main Authors: Ronald Carshon-Marsh, Susan Bondy, Theodore Witek, Prabhat Jha
• Format: Article in Journal/Newspaper
• Language: English
• Published: BMC 2024
• Subjects: Verbal autopsy; Minimally invasive tissue sampling; Complete diagnostic autopsy; CHAMPS; Malaria; Cause of death; Arctic medicine. Tropical medicine; RC955-962; Infectious and parasitic diseases; RC109-216; Arctic
• Online Access: https://doi.org/10.1186/s12936-024-05035-5; https://doaj.org/article/4f2f8eb1d5444627bc1daa1e459e8f36

Abstract Background Malaria contributes substantially to the persistent burden of child deaths in sub-Saharan Africa. Accurate and comprehensive malaria mortality data are crucial to monitor the progress in reducing malaria incidence and mortality. Verbal Autopsy (VA) ascertains the cause of death despite its limitations leading to misclassification errors. Minimally Invasive Tissue Sampling (MITS) is being conducted in some settings as an alternative to Complete Diagnostic Autopsy (CDA). The present study examines the validity of malaria-related deaths comparing VA diagnoses with those obtained through MITS and/or CDA. Methods A comprehensive literature search for original studies in English language using Ovid MEDLINE, Ovid Embase, CINAHL via EBSCO, Scopus, The Cochrane Library via Wiley, Google Scholar and searching the MITS Surveillance Alliance papers was carried out. The reference period was January 1, 1990–March 31, 2024. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines were adopted. Results Among 71 articles identified in the databases, 21 matched the eligibility criteria. Qualitative syntheses showed that malaria Cause Specific Mortality Fractions (CSMFs) across various studies ranged from 2 to 31%. Plasmodium falciparum was mostly responsible for these deaths and the most common complications were anaemia and cerebral malaria. The sensitivity and specificity of the VA validation studies ranged from 18.4% to 33% and from 86.6% to 97%, respectively, and there was a high level of misclassification for both InSilico and Expert Algorithm VA for malaria compared to MITS. The overall concordance rates between MITS and CDA diagnoses ranged from 68 to 90%, with the highest concordance seen in deaths due to infectious diseases and malignant tumours. Clinical data increased diagnostic coincidence between MITS blind to clinical data and the gold standard CDA by 11%. Conclusions The comprehensive review finds that MITS demonstrated better accuracy compared to VA in diagnosing ...