aprons as a proxy for salmonella

Problem statement

Most fresh food in low- and middle-income countries is bought in traditional (informal) markets, where contamination is high and behavior-change programs try to get vendors and shoppers to handle food more safely. The five-year (2019–2024), USAID-funded EatSafe program in Nigeria and Ethiopia — one of the largest recent efforts of its kind — deliberately did not measure whether its interventions changed contamination in the food. It measured "visible behaviors" instead, because microbial testing at one point in time "may not adequately represent trends," is slow and expensive to procure in-country, and behavior change is too slow to move contamination within a program cycle. The unsolved problem is that no one has a cheap, frequent, validated way to measure market-level food-safety outcomes, so the link between the behaviors programs promote and the pathogens people ingest remains assumed rather than demonstrated.

Why this matters

EatSafe's own formative testing shows how much is at stake: in Nigerian markets "Salmonella was detected in 37% of tomato samples"; in Ethiopia, of 328 kale, tomato and lettuce samples, "7% and 35% were found positive for Salmonella and/or generic E. coli, respectively," and total coliforms "were detected in 89% of samples at high levels," with the report noting the "risk of illness corresponding to observed microbial levels is high, much higher than in the U.S." Programs then spend years on training, branding, food-safety stands and market-management interventions and report outcomes such as apron use, elevated food display and stated intent — behaviors whose relationship to contamination is plausible but unquantified. Without an outcome measure, funders cannot tell effective interventions from theater, and the report itself notes "there is a clear gap between knowledge and action" and that vendors "self-report far greater levels of garbage disposal and stall cleaning than observed."

What’s been tried and why it hasn’t worked

EatSafe reviewed field-deployable assays and found that "portable and relatively simple assays exist, including some that do not need full laboratory settings to be processed," but that "costs and benefits need to be weighed carefully," and it used testing only at baseline — "to assess baseline levels and identify priority actions," build local lab relationships and "spur action by local stakeholders (not in consumer messaging, to avoid scares that could impact market livelihoods)." For evaluation, "the program did not use food testing to assess interventions" because "food testing at one point in time may not adequately represent trends," because behavior change over one year (Nigeria) or eight months (Ethiopia) "without enabling environment improvements" was judged unlikely to shift contamination, and because it is "time consuming and resource-intensive to procure in-country laboratory reagents and equipment." The fallback — assessing "visible behaviors (e.g., wearing an apron, displaying foods elevated from the ground) and evidence of behaviors (e.g., clean hands, clean water)" — is unvalidated as a proxy, and the report concedes that "food safety concepts may be difficult to distinguish from related food attributes such as freshness or overall quality." Contamination pathways are also tangled: food "can become (more) contaminated in the market" but some arrives contaminated and some practices reduce it, so a single test cannot attribute cause. Cold storage, the intervention with the clearest mechanism, "would require substantial investments and maintenance" beyond most vendors and markets.

What would unlock progress

Two things would unlock progress: a validated behavior-to-hazard model that tells evaluators which cheap observable practices actually predict lower contamination for a given commodity, and a low-cost, repeatable proxy measurement (surface or produce-rinse indicator tests, ATP or coliform cards, environmental sampling of shared water and surfaces) that can be run monthly by market staff rather than annually by a lab. The adjacent precedent is water, sanitation and hygiene (WASH), where cheap indicator tests and validated observation checklists replaced expensive pathogen assays for routine monitoring, and hospital infection control, which links audited hand-hygiene compliance to measured infection rates.

Entry points for student teams

A student team could design and pilot a market "safety dashboard" for one commodity in one market: pair a structured observation checklist with a low-cost indicator test (e.g., generic E. coli or coliform on produce rinsates and vendor surfaces) sampled repeatedly over a semester, and estimate which observed practices statistically predict indicator levels. A second team could evaluate candidate field assays against the constraints EatSafe names — in-country reagent supply, cost per test, turnaround — and produce a decision guide for program evaluators. A behavioral-science team could reanalyze EatSafe's published indicators to propose a minimum outcome-measurement protocol that programs could afford. Relevant skills: food microbiology, program evaluation, statistics, human-centered design, public health.