Infectious

Diarrheal disease

The condition of having at least three loose or liquid bowel movements each day.

Balanced: 5th of 142Biggest burden: 8th of 142Most neglected: 2nd of 142

Living starter kit

Not yet expert-checkedUpdated Nov 13, 2026

Start building on diarrheal disease

Diarrheal disease is a leading cause of death in young children, and large study datasets exist: tens of thousands of children with clinical and stool data, and cohorts that followed children for years. The hard problems are telling which germ actually caused an illness, predicting which children will fall behind in growth afterward, and building triage rules that hold up in new places.

Know diarrheal disease well? No expert has checked this kit yet.Check our work →

Open datasets4

GEMS childhood diarrhea study (ClinEpiDB) ↗
Clinical and stool test data from GEMS, a large case-control study of childhood diarrhea in several countries; downloads need study-team approval.
22,567 participants in the portal release
ClinicalApply for accessLicense not stated
MAL-ED birth-to-five cohort (ClinEpiDB) ↗
Data from MAL-ED, which followed children from birth to age five across multiple sites, tracking infections, diet and development; downloads need study-team approval.
2,145 participants in the portal release
ClinicalApply for accessLicense not stated
Bangladesh and Mali diarrhea cause data ↗
De-identified data on children with diarrhea in Bangladesh and Mali, with local weather, for predicting whether a virus caused the illness.
302 children enrolled in the associated validation study
ClinicalOpen downloadLicense not stated
UNICEF diarrhea treatment data ↗
Country tables from UNICEF on how many children with diarrhea receive recommended treatment.
EpidemiologyOpen downloadLicense not stated

Benchmarks and tasks3

Viral-only diarrhea prediction, Bangladesh and Mali ↗Predict whether a child's diarrhea is caused by a virus alone, from symptoms plus local season and setting, against the published Bangladesh and Mali validation study. Clinical
Growth faltering after diarrhea: GEMS to MAL-ED ↗Predict which children will fall behind in height after diarrhea, from basic clinical details, against models built on GEMS and tested on MAL-ED. Clinical
NIRUDAK dehydration severity models ↗Predict how dehydrated a patient older than five is (none, some or severe) from clinical signs, against the published NIRUDAK models. Clinical

Open questions4

01
Can a model tell viral from other diarrhea in a new place, where different germs are common?
Knowing an illness is viral lets doctors skip unneeded antibiotics, but the best model differed between Bangladesh and Mali.
ClinicalEpidemiologySource: External validation of diarrhea etiology prediction ↗
02
Can simple measurements identify which children will fall behind in growth after diarrhea, in any setting?
Falling behind in growth has lifelong effects, and the GEMS-to-MAL-ED study shows how hard it is to carry a rule from one population to another.
ClinicalSource: Growth-faltering clinical prediction study ↗
03
Can dehydration be judged reliably in older children and adults, across clinics?
Most dehydration guidance was built for young children; NIRUDAK is one of the few studies of older patients.
ClinicalSource: NIRUDAK derivation study ↗
04
Which mix of infections, diet and home environment best predicts later problems in a child's development?
MAL-ED was designed to study how these factors interact.
ClinicalEpidemiologySource: MAL-ED programme ↗

Who's working on this4

University of Utah and Brown University
The collaboration tested a phone-based model for predicting viral-only childhood diarrhea in Bangladesh and Mali.
Its public data and clearly specified predictors make a reproducible modelling task.
No public invitation foundEvidence, Feb 9, 2022 ↗
GEMS team, University of Maryland
The team coordinated a multi-country case-control study of childhood diarrheal disease.
Its linked clinical and stool data support work on what causes diarrhea and what follows it.
No public invitation foundEvidence, date unknown ↗
MAL-ED network (FNIH and Fogarty International Center)
The network studies infection, nutrition and child development across multiple sites.
Its long-term data supports models of repeated illness and later growth, not just one-off diagnosis.
No public invitation foundEvidence, date unknown ↗
NIRUDAK team, Brown University and icddr,b
The collaboration built clinical models of dehydration severity in older children and adults with acute diarrhea.
Published predictors and public workbooks show how a triage model was built.
No public invitation foundEvidence, Mar 10, 2021 ↗

Not in the kit yet

The two biggest datasets, GEMS and MAL-ED, need approval from the study teams to download. None of the datasets states a license, so check terms before commercial use.
All three benchmarks are published models to compare against, not live leaderboards.
GEMS and MAL-ED shift dates for privacy, which limits fine-grained outbreak modelling.
Most data covers young children; the NIRUDAK benchmark is for people older than five.

This kit keeps growing as new data, benchmarks and teams appear. Know something we missed? Tell us →

Curated by the Unsolved project team. Every link was checked on the date shown; licenses are as stated by each source.