A controlled trial published in The Lancet Infectious Diseases shows that involving communities and providers, including informal vendors, in an intervention observed a roughly two-thirds adjusted decline in the use of Watch antibiotics in Burkina Faso and the DRC. This result rests on a premise long overlooked: community use of critical antibiotics can be redirected without requiring complete medical training of local vendors.
The essentials
- In 44 villages and neighborhoods in Burkina Faso and the DRC, the weighted prevalence of Watch antibiotic use fell from 26.8% to 17.1% after nine months of intervention (Lancet Infectious Diseases, 2026).
- In control areas not covered, this share instead increased, from 13.4% to 21.2% over the same period.
- The evaluated lever is a combined community and provider package: the intervention directly involved local vendors and residents, with clinics and health centers observing the clearest effects, without requiring medical retraining.
- Regional and national policies are officially based on a One Health approach, but effective coverage of the local circuit varies by country.
- In the absence of intervention, Watch antibiotic use increased in control groups; the trial does not allow deduction of a progression rate for resistance.
Over-the-counter medicine: the first gateway to resistance
At the study sites, antibiotics were obtained from pharmacies, shops, and formal or informal markets, where the qualifications of dispensers varied. This local circuit is an important source of antibiotic supply in rural and peri-urban areas of sub-Saharan Africa.
Watch antibiotics occupy a particular place in this flow. The World Health Organization classifies them among antibiotics with high potential for selecting resistance and requiring monitoring: they remain effective against bacteria resistant to common drugs, but their excessive use accelerates the emergence of additional resistance. Reserve antibiotics are those reserved for very specific indications. Reserving them for severe infections, where no other option works, is a universal recommendation. In informal sales channels, it is nearly impossible to enforce.
The CABU-EICO trial starts from this observation. Its designers selected 44 areas in Burkina Faso and the DRC, countries where the infectious disease burden is heavy and formal health circuits are poorly accessible outside urban centers. Their hypothesis: if medicine vendors understand why certain molecules must be preserved, and if residents know what question to ask before buying, purchasing behavior changes. Not radically. But enough to influence the trajectory of resistance.
Results after nine months of intervention
The protocol did not seek to transform market vendors into licensed pharmacists. It aimed for something more modest and more realistic: giving them a clear rule on critical molecules, and giving buyers a minimal capacity to question their choices.
The result is clear. In the intervention areas, the weighted prevalence of Watch antibiotic use fell from 26.8% to 17.1%. That is a reduction of nearly ten percentage points in nine months. In control areas, the movement is reversed: from 13.4%, this share rose to 21.2%. The gap between the two trajectories documents a real effect of the intervention, and not merely a generally favorable trend.
This figure is notable not only for its magnitude, but because it was achieved without degrading treatment of common infections. Access group antibiotics, first choice when an antibiotic is indicated, continued to circulate in intervention areas; many mild infections require no antibiotic. The trial reduced Watch antibiotic use with no measured negative effect on patient care.
The limits of a hospital-centered strategy
African national plans are officially structured around a One Health approach, with unequal implementation across sectors and countries. Training doctors to prescribe fewer critical antibiotics, controlling prescriptions in public facilities, monitoring resistance in university hospital laboratories. This approach makes sense where the majority of care passes through a health professional.
It leaves out an entire continent of distribution. In many low-resource communities, difficult or delayed access to care fuels widespread self-medication through formal or informal medicine vendors. The antibiotic is often bought based on symptoms, without diagnosis, sometimes in insufficient quantity to complete treatment. These conditions are precisely those that favor the selection of resistant bacteria.
Targeting only hospitals amounts to regulating one tap while leaving a parallel pipe open. The trial in Burkina Faso and the DRC shows what this asymmetry produces: in control areas, the share of critical molecules progressed by nearly eight points while intervention areas reduced their use. The two dynamics coexist on the same territory, separated only by the presence or absence of community intervention.
The question for public health policy then becomes why this asymmetry persists. Part of the answer lies in regulatory frameworks: regulating a public hospital is simpler than regulating thousands of scattered vendors. Another part lies in representations: antibiotic resistance has long been thought of as a problem of overconsumption in wealthy countries, not as a structural issue in low-income health systems.
Training without retraining: the gamble of the community protocol
The CABU-EICO intervention rests on an architecture that deserves to be described precisely, because it is reproducible. Its designers avoided two symmetric pitfalls: leaving medicine vendors with no training at all, or requiring them to achieve a level of competence they cannot quickly acquire.
The protocol trained providers to distinguish Access and Watch antibiotics and apply treatment recommendations according to infections. It combined training, case discussions, therapeutic recommendations, audit feedback, and community campaigns. It was a package of behavioral and educational intervention, not retraining.
On the residents’ side, the intervention provided minimal benchmarks: understanding that a powerful antibiotic is not necessarily the best choice for a simple infection, and being able to ask the vendor a question. This community mobilization links the trial to a broader logic, where local health depends as much on patient behaviors as on care practices. In the field of artificial intelligence applied to health, the same logic of local appropriation determines the effectiveness of deployed tools.
The financing of such programs remains the major point of uncertainty. Training vendors, producing educational materials, ensuring follow-up on sales: these costs are modest compared to those of hospital programs, but they require a political decision and dedicated budget that most national action plans on antibiotic resistance do not yet provide for.
Two trajectories for the next twenty years
The trial documents a moment. The question that opens concerns the long-term trajectory, in a context where antibiotic resistance already kills over one million people per year worldwide, with a disproportionate burden on sub-Saharan Africa.
Two scenarios emerge from the trial data. In the first, policies to combat antibiotic resistance integrate local vendors and residents as regulated actors. Current national action plans update their targets to include the informal circuit. Short trainings are deployed at scale, with a simple system for tracking sales of critical molecules. The reduction documented in Burkina Faso and the DRC can then be reproduced in other countries with similar profiles, before resistance crosses thresholds difficult to reverse.
In the second scenario, the hospital approach remains dominant. Local vendors remain outside national programs. Access to critical antibiotics continues to grow in informal circuits, at the pace observed in the trial’s control areas. Resistance progresses in rural and peri-urban communities that depend most on these circuits, without the usual surveillance systems, centered on hospital laboratories, detecting it sufficiently early.
These two trajectories depend on concrete decisions that can be made or avoided. The number of low-income countries that will integrate local medicine vendors into their national action plan against antibiotic resistance constitutes a readable signal. A few African countries have already begun to broaden the definition of front-line health actors to include community agents; extending this to antibiotic dispensing represents an additional step, not an impossible qualitative leap.
Economic stakes reinforce the health argument. Antibiotics remain the reference treatment for the deadliest bacterial infections. Losing their effectiveness on critical molecules forces health systems to more complex, more costly, less available treatments in low-income contexts. Comparison with other long-term public health challenges is instructive: expanding vaccination programs to poorly covered territories followed a similar arc, moving from pilot projects to nationally funded policies over several decades. The trajectory of antibiotic resistance could follow the same logic, provided that community actors are recognized as levers, and not merely as problems.
Points the trial does not resolve
The CABU-EICO trial answers a specific question: the intervention works in the short term in rural and peri-urban contexts in Central and West Africa. The answer is yes. Several questions remain open.
The first is that of sustainability. Nine months of intervention document an effect. Nothing in the trial allows us to know whether this effect persists after the program ends, or whether the behaviors of vendors and buyers are maintained without external follow-up. Community health trials regularly show effects that fade after research teams withdraw.
The second is that of transfer. Burkina Faso and the DRC share characteristics: high density of informal vendors, low hospital coverage, significant infectious disease burden. Other sub-Saharan African countries present different profiles. The efficacy of the protocol in dense urban contexts, or in countries where formal health systems are more developed, remains to be established.
The third is that of spillover effect. The trial measures the proportion of patient visits resulting in dispensing or purchase of a Watch antibiotic. The trial also plans to directly measure community carriage of resistant bacteria; the main article primarily reports Watch antibiotic use and care. The link between reduced sales of critical molecules and slowing of resistance is biologically plausible and supported by literature, but the trial does not prove it directly.
These limitations do not reduce the scope of the result. They indicate where subsequent research must focus, and what decision-makers must monitor before generalizing the protocol. In global health, controlled trials like this one remain the most reliable instruments for distinguishing what works from what seems reasonable. Antibiotic resistance has long lacked this type of evidence for the informal circuit.
The actors who can shift the lines
The WHO publishes an annual list of priorities for fighting antibiotic resistance. The national action plans that member states commit to submitting form the main governance framework. So far, these plans have mainly targeted hospital prescriptions and agricultural uses, two sectors easier to regulate.
The CABU-EICO trial now provides a basis for expanding this framework. Its designers, and the institutions that funded it, can bring this result into negotiations on national action plans in participating countries. Burkina Faso and the DRC have local proof, produced in their own context. This is a stronger position than generic recommendations built on data from other continents.
Organizations working on access to medicines in sub-Saharan Africa also have a role to play. The trial reduced Watch antibiotic use with no measured negative effect on patient care; it alone does not allow concluding that it improved access to care or to Access antibiotics as a distinct outcome. This combination responds to a frequent objection, that regulating antibiotic use risks depriving already under-medicated populations of access to treatment.
The question that remains open for policy actors is that of financing a scale-up. Training medicine vendors across thousands of villages requires resources that current national programs do not provide for. Global funds dedicated to health, some of which now include antibiotic resistance among their priorities, constitute a possible lever. The issue is to make this financing a decision taken before resistance in informal circuits is documented as a crisis, rather than after.
Sources
- CABU-EICO controlled trial, The Lancet Infectious Diseases, 2026 (via PubMed)
- World Health Organization, AWaRe Classification of Antibiotics (Access, Watch, Reserve), WHO website
- WHO, National Action Plans on Antibiotic Resistance, Global Monitoring Framework
