BOED-AMR surveillance planner

Which country × pathogen × antibiotic cells should Africa test next to learn the most per dollar?

Code on GitHub ↗
$1,000,000
as submitted
Click countries on the map to include or exclude them. Pathogens:
Information gained · BOED plan
–
Information gained · ad-hoc expansion
–
BOED advantage, same budget
–
total expected information gain, BOED ÷ ad-hoc
Laboratory capacity used
–

Where the plan sends new testing

New cells funded per country. Click a country to list its cells. Purple = this choice; hover a country to compare with BOED. Click a country to include or exclude it.

025cells funded

What each extra dollar buys

Total information gained as the budget grows. Beyond ~$1.5M, lab capacity, not money, is the limit.

Cells in the BOED plan

#CountryPathogenAntibioticEstimated resistance (95% CrI)Info gainCountry has data

Does the uncertainty point to the right places?

Retrospective test: refit on data up to 2018 only, then compare with 2019–2023.

AUC 0.78
Pre-2019 uncertainty separated cells that later shifted from stable ones (1,938 cells, 331 shifted).
Box plot: pre-2019 predictive SD is higher for cells that later shifted than for stable cells; AUC 0.78