CODA Democra MITS

Durada

01/07/2025 - 31/07/2027

Coordinador

ISGlobal

Finançadors

The Gates Foundation

Minimally Invasive Tissue Sampling (MITS) has been validated as a reliable alternative to the complete diagnostic autopsy (CDA) for cause-of-death (CoD) determination in low-resource settings. However, conventional MITS (c-MITS) remains resource-intensive — with per-case costs of USD 600–1,000, multiple sample types, and extensive microbiological, histopathological, and molecular workup — limiting sustainable integration into routine mortality surveillance in LMICs.

Democra-MITS aims to develop and validate simplified, age- and CoD-specific MITS protocols (s-MITS) that preserve diagnostic accuracy while substantially reducing procedural complexity, time, and cost.

The project is structured around two objectives:

1. Retrospective development. AI-assisted modelling applied to existing datasets from CHAMPS, CaDMIA/CaDMIA Plus, and São Paulo (~10,000 MITS cases) will quantify the diagnostic contribution of each sample–test unit across neonates, children, non-pregnant adults, and maternal deaths. Expert-adjudicated DeCoDe scores on a gold-standard subset (~250 cases) will train ordinal regression, LASSO, Random Forest, gradient boosting, and Bayesian network models. Decision-tree optimisation and Pareto frontier analyses will identify minimal diagnostic sets retaining ≥90% of the diagnostic information of c-MITS. Candidate protocols will then be retrospectively validated against paired CDA data.
2. Prospective validation. Refined s-MITS protocols will be tested in a multicentre, within-case comparative study of 200 deaths across three geographically and epidemiologically distinct sites (Maputo and Quelimane, Mozambique; Pokhara, Nepal). Sequential c-MITS and CDA will be performed on each case, enabling direct assessment of diagnostic agreement, sensitivity and specificity by CoD category, procedure time, and cost.

Expected outputs. A set of validated, age- and cause-specific s-MITS protocols and modular MITS kits, accompanied by evidence on diagnostic accuracy, feasibility, and cost relative to CDA — supporting scalable postmortem surveillance in LMICs in line with WHO mortality surveillance priorities and SDGs 3 and 17.

Total Funding

1.499.944,00$

Gates Foundation (INV-094224)

Coordination

ISGlobal

Partners

Maputo Central Hospital, CISM (Mozambique); Gandaki Medical College (Nepal).