Effect of a community-based behavioural intervention bundle to improve antibiotic use and patient management in Burkina Faso and DR Congo: a cluster randomised controlled trial

This cluster randomised controlled trial in Burkina Faso and the Democratic Republic of Congo demonstrated that a community-based behavioural intervention bundle significantly reduced Watch-group antibiotic use without negatively impacting patient management, offering a promising strategy to mitigate antimicrobial resistance in sub-Saharan Africa.

The Gist

The Big Picture: A "Community Detox" for Antibiotics

Imagine a neighborhood where everyone has a "magic potion" (antibiotics) they use for almost everything: a cold, a stomach ache, a scratch, or a fever. The problem is, this magic potion is losing its power. Because people use it too much and for the wrong things, the bad germs are learning how to fight back. This is called Antimicrobial Resistance (AMR).

In two places in Africa (Burkina Faso and the Democratic Republic of Congo), the germs were getting very strong, and people were using the "Watch-group" antibiotics (the heavy-duty, broad-spectrum ones) way too often, even when they didn't need them.

The Goal: The researchers wanted to see if they could teach the whole community—both the doctors/sellers and the patients—how to use these "magic potions" more wisely, like a team learning to play a game better together.

---

The Experiment: The "Bundle" Approach

Instead of just giving a lecture to doctors or just telling patients "don't do it," the researchers created a Community Intervention Bundle. Think of this as a three-part toolkit designed to change habits from both sides of the counter.

1. The "Coach" for the Sellers (Supply Side)
The researchers gathered all the people who sell or prescribe medicine: doctors in clinics, pharmacists, and even the informal street vendors.

• The Analogy: Imagine a sports coach gathering the whole team (not just the star players) for a series of training camps.
• What they did: They held three rounds of workshops. They didn't just read a manual; they used role-plays, quizzes, and feedback. They taught them: "For a cough, try waiting. For a fever, check if it's malaria first. Save the heavy-duty antibiotics for the really dangerous infections." They introduced a new "Rulebook" (the WHO AWaRe Book) that acts like a GPS for doctors, telling them exactly which medicine to pick for which problem.

2. The "Hype Team" for the Community (Demand Side)
While the sellers were learning, the community was getting a message too.

• The Analogy: Think of a town hall meeting mixed with a street festival.
• What they did: They used theater plays, songs, soap distribution, and door-to-door visits. They explained that taking antibiotics for a virus (like a cold) is like using a sledgehammer to crack a nut—it doesn't work and it breaks the nut (your body's defenses). They encouraged people to wash hands and drink clean water so they wouldn't get sick in the first place.

3. The "Secret Shoppers" (The Test)
To see if the training actually worked, the researchers used Simulated Patients.

• The Analogy: Imagine a mystery shopper going into a store, but instead of buying a shirt, they pretend to be sick with a specific symptom (like a bad cough or a fever) to see how the seller reacts.
• The Process: Trained actors went to clinics and shops pretending to be sick. They recorded whether the seller asked good questions, examined them, or just immediately handed over the heavy-duty antibiotics.

---

The Results: Did It Work?

The study compared villages that got this "Bundle" treatment against villages that just continued with their normal routine (the control group).

The Good News:

• The "Heavy Hitters" Dropped: In the villages with the intervention, the use of the powerful "Watch-group" antibiotics dropped significantly.
  • The Metaphor: It's like a town that used to use a fire hose to put out a candle. After the training, they switched to using a cup of water. The "fire hose" usage dropped by about 33% compared to the towns that didn't get the training (where usage actually went up).
• No Harm Done: Crucially, the "mystery shoppers" found that the quality of care didn't get worse. Doctors still asked questions and examined patients; they just prescribed fewer unnecessary drugs.
• The Shift: People started realizing that for minor illnesses, sometimes "no medicine" is the best medicine.

The Challenge:

• The "Street Vendors" were Harder to Change: While the clinics and pharmacies improved, the informal street vendors were harder to sway.
  • The Metaphor: The clinics were like a school where the teachers listened to the new rules. The street vendors were like a busy marketplace where customers were shouting, "I want my medicine now!" and the vendors felt pressured to sell it just to keep their customers happy. Without strict laws or police enforcement, it was hard for them to say "no" to a customer demanding antibiotics.

---

Why This Matters

Think of antibiotics as a shared battery for the whole community. If everyone uses it recklessly, the battery dies, and when someone really gets sick (like with a severe infection), there's no power left to save them.

This study showed that you don't need to build new hospitals or spend millions of dollars to fix this. You just need to:

1. Teach the sellers the right rules.
2. Educate the buyers so they stop demanding the wrong things.
3. Work together as a community.

By doing this, the "battery" (our antibiotics) will last longer, and the "bad germs" won't learn how to defeat our medicine as quickly. It's a small step that could save lives in the future.

Technical Summary

1. Problem Statement

Antimicrobial resistance (AMR) is a critical global health threat, with sub-Saharan Africa bearing the highest estimated mortality burden. A primary driver of AMR is the inappropriate use of antibiotics, particularly the overuse of "Watch-group" antibiotics (broad-spectrum agents with higher resistance potential) for conditions where they are not indicated (e.g., viral infections or minor ailments). In low- and middle-income countries (LMICs), antibiotic use is often driven by self-medication, limited diagnostic capacity, and a lack of adherence to treatment guidelines. While the World Health Organization (WHO) introduced the AWaRe Antibiotic Book (categorizing antibiotics into Access, Watch, and Reserve groups) in late 2022, there is a scarcity of experimental evidence on how to effectively implement these guidelines in community settings across diverse healthcare providers (formal and informal) in sub-Saharan Africa.

2. Methodology

Study Design:
The study was a cluster-randomised controlled trial (cRCT) conducted in 44 villages/neighbourhoods across two health districts: Nanoro, Burkina Faso, and Kimpese, Democratic Republic of Congo (DRC).

• Randomisation: Clusters were allocated 1:1 to intervention or control arms, stratified by site and provider type (e.g., presence of a health centre vs. only vendors).
• Intervention: A co-created, behavioural intervention bundle targeting both the supply side (healthcare providers) and demand side (communities).
  • Supply: Three rounds of educational and feedback sessions for health centres, private clinics, pharmacies, and informal vendors. These sessions focused on four priority infections (cough/bronchitis, pneumonia, diarrhoea, acute fever) using AWaRe Book guidance.
  • Demand: Mass health education campaigns (theatre, videos, door-to-door visits) promoting handwashing, sanitation, and rational antibiotic use.
• Duration: The intervention was rolled out over nine months (Jan–Sep 2023 in Nanoro; Apr–Dec 2023 in Kimpese).

Data Collection & Outcomes:

• Primary Outcome: The change in the weighted prevalence of Watch-group antibiotic use (patients receiving at least one Watch-group antibiotic) comparing intervention vs. control clusters, pre- and post-intervention. The analysis was adjusted for healthcare utilisation rates to estimate community-wide impact.
• Secondary Outcomes: Prevalence of any antibiotic use, patient management scores (via simulated patient visits for five infection scenarios), and process evaluation via qualitative interviews.
• Sample Size: 5,532 patients surveyed at baseline and 4,898 post-intervention across 63 health centres, 60 pharmacies, and 41 informal vendors. 1,092 simulated patient visits were conducted.
• Analysis: Negative binomial regression with cluster-robust standard errors was used to calculate adjusted Prevalence Ratios (PR), accounting for overdispersion and non-independence.

3. Key Contributions

• Dual-Target Approach: This is the first study to simultaneously target both antibiotic dispensing (supply) and community demand through a co-created intervention bundle in sub-Saharan Africa.
• AWaRe Book Implementation: It is the first experimental study to evaluate an intervention informed by the WHO AWaRe Antibiotic Book, published only two months before the study began.
• Comprehensive Provider Scope: Unlike previous studies focusing solely on formal clinics, this trial included informal medicine vendors and community pharmacies, acknowledging their significant role in antibiotic distribution in LMICs.
• Methodological Rigor: The study utilized simulated patient visits to assess clinical management quality (history taking, examination, dispensing) without the ethical and logistical challenges of following up on real patients, ensuring safety monitoring.
• Population-Level Estimation: By adjusting antibiotic use prevalence for local healthcare utilisation rates, the study estimated the true population-wide impact rather than just facility-level metrics.

4. Key Results

Antibiotic Use Reduction:

• Watch-Group Antibiotics: The intervention resulted in a statistically significant reduction in Watch-group antibiotic use.
  • Intervention Arm: Decreased from 26.8% (baseline) to 17.1% (post-intervention).
  • Control Arm: Increased from 13.4% to 21.2%.
  • Effect Size: Adjusted Prevalence Ratio (PR) of 0.33 (95% CI 0.14–0.78), indicating a substantial relative reduction.
• Any Antibiotic Use: The prevalence of any systemic antibiotic use also decreased significantly in the intervention arm (56.2% to 37.5%) compared to a slight increase in the control arm (PR 0.48).
• Site Variations: Reductions were most pronounced in health centres and private clinics. In Kimpese, private clinics contributed significantly to the reduction. In Nanoro, health centres drove the change.

Provider Type Differences:

• Formal Providers (Clinics/Health Centres): Showed the greatest reduction in inappropriate prescribing.
• Informal Vendors: Reductions were minimal. Qualitative data revealed that informal vendors faced intense pressure from customers to sell antibiotics and lacked the authority or regulatory backing to refuse sales.

Patient Management:

• Scores: Baseline patient management scores were low, particularly among informal vendors (mean score ~0.1) compared to health centres (mean ~6.6).
• Change: Changes in management scores were limited. There was no evidence of harm (e.g., worsening of care), but improvements in clinical history-taking and physical examinations were modest.
• Simulated Visits: No major increase in Watch-group antibiotic use was observed during simulated visits, except for a potential increase in diarrhoea cases at pharmacies and informal vendors.

Qualitative Insights:

• The intervention successfully raised awareness about the risks of self-medication and "street medicines."
• However, financial barriers and drug availability in formal health centres (e.g., stock-outs, indirect costs like transport) continued to drive patients toward informal vendors.

5. Significance and Implications

• Feasibility of Behavioural Bundles: The study demonstrates that a contextualised, low-cost behavioural intervention can substantially reduce the use of high-risk antibiotics in primary care settings without negatively impacting patient management.
• Stewardship Strategy: Reducing Watch-group antibiotic use at the community level is a critical strategy to slow the emergence of resistance in community-acquired pathogens.
• Limitations of Education Alone: While effective in formal settings, the study highlights that education and feedback alone are insufficient to curb antibiotic sales by informal vendors. Stronger regulatory measures or structural interventions (e.g., ensuring drug availability in public clinics) are likely needed to address the informal market.
• Policy Relevance: The findings support the integration of the AWaRe Book into routine primary care training and suggest that interventions must address both supply (provider training) and demand (community awareness) to be effective.

In conclusion, the CABU-EICO trial provides robust evidence that community-based, multi-component interventions can effectively optimize antibiotic use in sub-Saharan Africa, though addressing the informal market requires more than just educational strategies.