Can Self-Testing Reduce Epidemic Deaths in Africa?

Self-testing has become one of the most attractive ideas in epidemic preparedness: fast, decentralized, easy to scale, and potentially useful in health systems where laboratory capacity is limited. However, a new modelling study on epidemic-prone diseases in Africa delivers a sharp warning for policymakers: distributing test kits is not the same as controlling transmission.

Published in Diagnostics, the study examines whether self-testing can reduce epidemic peaks and avert deaths when added to standard health-system testing. The researchers focused on African Union priority pathogen archetypes, including Ebola, Influenza A, Cholera, Coronavirus and Mpox, using a deterministic SEIR model to compare standard-of-care testing with additional self-testing.

Untargeted population-wide self-testing produced only modest epidemic mitigation, with a median reduction in peak disease prevalence of 1.9%. Its impact depended most strongly on whether people adhered to post-test isolation, followed by testing intensity and pathogen characteristics.

The test kit is not the intervention

Self-testing is often discussed as a breakthrough in diagnostic access, especially after its expansion in HIV and COVID-19 responses. In principle, it allows individuals to collect samples, perform tests and interpret results at a time and place of their choosing. That makes it one of the most decentralized diagnostic tools available.

However, the study shows why epidemic control is more complex than access alone. For fast-moving infections, the public health value of testing depends on what happens after the result. A person who tests positive must isolate quickly enough, for long enough, and with enough support to interrupt transmission. Without that behavioral step, self-testing may detect infections without meaningfully changing the epidemic curve.

Adherence to isolation was the strongest driver of epidemic mitigation, with a partial rank correlation coefficient of 0.784. Self-testing intensity followed at 0.617, while lower transmissibility and longer infectious periods also improved impact. Test sensitivity mattered, but less than isolation behavior.

The finding should reshape how governments and donors think about diagnostics. The most accurate or widely distributed test will not deliver its full value if people cannot afford to isolate, fear stigma, lack food support, live in crowded housing, distrust the result, or do not understand what action to take. In outbreak settings, the real intervention is not the test kit. It is the chain linking testing to isolation, care, confirmation, surveillance and community support.

Mass self-testing runs into a scale problem

The findings challenge the idea that population-wide self-testing can be a general epidemic-control tool in resource-constrained settings. In the model, even a favorable scenario with low transmission and low background health-system testing showed limited impact at existing operational benchmarks. At the WHO Afro benchmark of 10 tests per 10,000 people per week, the reduction in peak prevalence was only 0.45%. To achieve a 10% reduction in peak prevalence, the model required about 34 self-tests per 10,000 people per day, roughly 24 times higher than the benchmark.

The benchmark sits far below the testing intensity needed for a meaningful reduction in peak prevalence under the modelled conditions. For African health systems already facing limited diagnostic capacity, workforce constraints, fragile supply chains and competing health priorities, sustaining that level of population-wide self-testing through an epidemic would be extremely difficult.

Self-testing means policymakers should avoid treating it as a low-cost substitute for health-system capacity. Epidemic preparedness still requires laboratory networks, confirmatory testing, trained health workers, data systems, surveillance, risk communication and isolation support. Self-testing may widen the front door to diagnosis, but it cannot replace the public health machinery that must act on the result.

The biggest gains may come in the deadliest outbreaks

The study's most interesting nuance is that self-testing may be more valuable for reducing deaths than for flattening epidemic peaks, especially for high-mortality pathogens with moderate transmission. The number needed to self-test to avert one death varied sharply by pathogen archetype: 1,512 tests for Ebola-like pathogens, 22,590 for Coronavirus-like pathogens, 55,453 for Cholera-like pathogens, 117,231 for Influenza A-like pathogens and 355,708 for Mpox-like pathogens.

With a highly lethal disease, even a small reduction in transmission can translate into meaningful survival gains. Ebola-like outbreaks, which often have clustered spatial dynamics, may be more suitable for rapid, targeted self-testing than lower-mortality or faster-spreading pathogens. The authors suggest that self-testing could be explored in communities near outbreaks or after funerals as an enhanced form of contact tracing, although they stress that this remains speculative and would require confirmatory testing, quarantine protocols and biosafety safeguards.

Rather than mass distribution across entire populations, self-testing may be most useful when targeted to places where conventional diagnostics are delayed or absent: villages near an outbreak, recent migrants into affected areas, high-risk contacts, underserved communities, or locations where stigma prevents early presentation to health facilities. Stakeholders consulted in the study also questioned population-wide deployment and emphasized targeted use where diagnostic systems are weakest.

The goal should not be to make self-testing universal for every epidemic-prone disease. The goal should be to identify where it can deliver the greatest marginal benefit, fastest, with the fewest wasted resources and the least social harm.

Preparedness must fund behavior, not just diagnostics

Donors and governments often fund commodities because they are visible, measurable and easier to procure. However, the model suggests that the biggest determinant of impact may be less visible: whether people can act on a positive result. It requires investment beyond test procurement. Community education, risk communication, food and income support during isolation, clear referral pathways, confirmatory testing, safe quarantine arrangements and trusted local leadership may determine whether self-testing reduces transmission or simply generates data. The study also highlights the risk of false-positive burdens, especially where repeated isolation of healthy people could undermine acceptability over time.

The authors also highlight the limits of their work. The pathogen archetypes are not disease-specific forecasts and should not be interpreted as empirical effectiveness estimates. The model uses simplified assumptions and does not include treatment benefits, even though treatment can reduce mortality for diseases such as Ebola and Cholera. It also does not model targeted self-testing through contact tracing or social networks, which could prove more efficient than population-wide deployment.

Still, the study provides a useful warning for the next phase of diagnostic policy. Self-testing is unlikely to function as a general epidemic-mitigation tool under constrained health-system capacity because the volumes needed to meaningfully reduce transmission may exceed operationally plausible thresholds. However, early, decentralized deployment during Ebola-like outbreaks may still offer a feasible path to reduce mortality and support response if future studies can identify practical deployment strategies.