Engaging communities through participatory learning action for the control and prevention of diabetes: a protocol for the Process Evaluation of the EMPOWER-D trial in Pakistan and Afghanistan

This paper outlines the protocol for a mixed-methods process evaluation of the EMPOWER-D trial, which aims to assess the implementation, mechanisms of impact, and contextual factors of a participatory learning and action intervention for preventing type 2 diabetes in community settings across Pakistan and Afghanistan.

The Gist

Imagine you are trying to fix a leaky roof in a house, but instead of just handing the homeowner a bucket, you gather the whole neighborhood to figure out why the roof is leaking, how to fix it together, and what tools they need to keep it dry forever.

That is essentially what this paper is about. It's not the final report on whether the roof is fixed; it's the instruction manual and the diary of how the team plans to check if their neighborhood project is actually working.

Here is the breakdown of the EMPOWER-D project and its "Process Evaluation" (the diary part), explained simply:

1. The Big Problem: The "Silent Storm"

In Pakistan and Afghanistan, a massive health storm is brewing: Type 2 Diabetes.

• The Scale: It's like a flood. In Pakistan alone, millions of people have it, and many don't even know they are sick. By 2050, the number of people with diabetes is expected to double.
• The Challenge: The local health systems are like small rowboats trying to fight a tsunami. They don't have enough doctors or medicine to treat everyone, so they need a different strategy.

2. The Solution: The "Community Garden" Approach

Instead of doctors telling people what to do, this project uses a method called Participatory Learning and Action (PLA).

• The Analogy: Think of it as a Community Garden.
  • In the past, experts might have just dropped off seeds and told people, "Plant these."
  • In this project, the community members meet in groups (like a garden club). They look at their own land, decide what's growing wrong (maybe they are eating too much sugar or not moving enough), and they design their own plan to fix it.
  • They meet regularly, share ideas, try solutions, and then sit back and say, "Did that work? What should we try next?"
• The Goal: To empower people to become the "gardeners" of their own health, rather than just passive patients.

3. The Experiment: Testing the Garden in Different Soils

The researchers are testing this "Community Garden" idea in three very different places:

1. Rural Pakistan: A village setting.
2. Urban Pakistan: A busy city setting (Karachi).
3. Rural Afghanistan: A difficult, conflict-affected setting where life is very hard.

They know that a garden that grows in a sunny city park might not grow in a rocky mountain village. So, they are adapting the "seeds" (the activities) to fit the local soil.

4. The "Process Evaluation": The Detective's Notebook

This specific paper is a protocol (a plan) for a "Process Evaluation."

• What is a Process Evaluation? Imagine you are baking a cake.
  • The Main Trial is tasting the cake at the end to see if it's delicious (Did diabetes go down?).
  • The Process Evaluation is the detective work while you are baking. It asks:
    • Did we follow the recipe?
    • Did the oven temperature change?
    • Did the baker get tired?
    • Did the neighbors smell the cake and want to join in?
• Why do we need it? Sometimes a cake tastes bad not because the recipe is wrong, but because the baker didn't have enough sugar, or the oven was broken. This study wants to know exactly what happened during the "baking" so they can fix the recipe for the future.

5. How They Will Investigate (The Tools)

The team will use a mix of tools to get the full picture, like a detective using a magnifying glass, a camera, and a notebook:

• Attendance Sheets: Did people actually show up to the garden meetings? (The "Dose").
• Interviews & Focus Groups: Talking to the "gardeners" (participants) and the "garden leaders" (community mobilizers) to hear their stories. "Did you feel empowered? Was it too hard?"
• Photovoice: In rural Pakistan, participants will take photos of their daily lives and the changes they see. They will bring these photos to meetings to tell stories that words alone can't capture.
• Supervisor Reports: Checking if the leaders are following the rules and guiding the groups correctly.

6. The Three Big Questions They Are Asking

The study is looking at three main things (like checking the foundation, the walls, and the roof of a house):

1. Implementation: Did we build the garden as we planned? Did we have to change the blueprints because of the weather (context)?
2. Mechanisms of Impact: How did it work? Did it work because people felt more confident? Did it work because the whole neighborhood started supporting each other?
3. Context: What outside forces helped or hurt? For example, in Afghanistan, women's movement is restricted. How did that change the way the garden was built?

7. Why This Matters

This is the first time this specific type of community approach is being tested in Afghanistan and Pakistan together.

• The Payoff: If this works, it's a blueprint for how to fight diabetes in places where money and doctors are scarce. It proves that you don't need a high-tech hospital to save lives; sometimes, you just need a community that cares and a plan they made together.
• The Future: If they find out what works and what doesn't, they can teach other countries how to build their own "Community Gardens" to stop the diabetes flood.

In short: This paper is the roadmap for a team of detectives who are going to watch a community-led health experiment in real-time, taking notes on every success and stumble, so that one day, this method can be used to save millions of lives across the world.

Technical Summary

Based on the provided preprint, here is a detailed technical summary of the EMPOWER-D Process Evaluation Protocol.

1. Problem Statement

Type 2 Diabetes Mellitus (T2DM) represents a critical public health crisis in low- and middle-income countries (LMICs), particularly in South Asia.

• Prevalence: Pakistan has the highest global age-standardized prevalence of T2DM (31.4% of adults), with projections indicating a 103% increase by 2050. Afghanistan also faces rising trends (11.7% prevalence) with a high rate of undiagnosed cases (71.4%).
• Systemic Gaps: Health systems in these regions face significant capacity gaps, making pharmacological interventions alone insufficient.
• Intervention Gap: While Participatory Learning and Action (PLA) has proven effective for diabetes prevention in Bangladesh (DMAGIC project), there is a lack of evidence regarding its adaptation, implementation, and effectiveness in other LMIC contexts, specifically in fragile and conflict-affected settings like Afghanistan and diverse urban/rural settings in Pakistan.
• Evaluation Need: There is a scarcity of rigorous process evaluations (PE) for complex community-based interventions in LMICs, particularly those examining how and why interventions work (or fail) across different cultural and political contexts.

2. Methodology

This paper outlines the protocol for an embedded mixed-methods Process Evaluation (PE) of the EMPOWER-D trial. The study follows the UK Medical Research Council (MRC) framework for evaluating complex interventions.

• Study Sites: Three distinct contexts:
  1. Rural Pakistan (Khyber Pakhtunkhwa): Cluster RCT (Peshawar and Swabi).
  2. Urban Pakistan (Karachi): Feasibility Cluster RCT.
  3. Rural Afghanistan (Greater Kabul region): Feasibility trial in a conflict-ridden setting.
• Intervention: The EMPOWER-D intervention adapts the PLA approach, utilizing facilitated men's and women's groups. Community Mobilisers (CMs) guide a cycle of identifying health problems, planning solutions, implementing actions, and reflecting on outcomes. The cycle runs for 18 months in Pakistan and 12 months in Afghanistan.
• Evaluation Design: A convergent mixed-methods design integrating qualitative and quantitative data.
• Three Core Domains of Evaluation:
  1. Implementation: Assessing fidelity, dose (intensity), reach, adaptations made, acceptability, and sustainability.
  2. Mechanisms of Impact (MoI): Exploring changes at individual (agency, self-confidence), group (collective efficacy), and community (norms, social cohesion) levels.
  3. Context: Analyzing external factors (cultural, political, economic, gender dynamics) influencing delivery and outcomes.
• Data Collection Methods:
  • Quantitative: Attendance logs, supervisor checklists, fidelity reports, and trial survey data (analyzed descriptively).
  • Qualitative: In-depth Interviews (IDIs) and Focus Group Discussions (FGDs) with participants, CMs, supervisors, and stakeholders.
  • Participatory Visual Methods: Photovoice will be used in rural Pakistan, where participants document experiences via photography to stimulate discussion during FGDs.
• Data Analysis:
  • Quantitative data will be analyzed using Stata for descriptive statistics.
  • Qualitative data will undergo thematic analysis using a Theory of Change (ToC) framework, supported by software (ATLAS.ti/NVIVO).
  • Data will be triangulated to identify commonalities and context-specific influences across the three sites.

3. Key Contributions

• First Multi-Country PE: This is the first process evaluation of a PLA intervention for diabetes prevention conducted simultaneously across multiple LMICs, including a fragile state (Afghanistan).
• Methodological Innovation: It integrates Photovoice into a diabetes prevention trial to capture visual narratives of community change and challenges.
• Contextual Adaptation: The study explicitly documents how a standardized intervention is adapted to vastly different contexts (rural vs. urban Pakistan; stable vs. conflict-affected Afghanistan), providing a blueprint for scaling NCD interventions in diverse settings.
• Theory of Change Validation: It tests the assumptions of the PLA approach in new cultural settings, specifically examining how gender dynamics (e.g., restrictions on women's movement in Afghanistan) impact the intervention's mechanisms.

4. Results

• Status: As this is a protocol paper, no empirical results (outcomes or findings) are presented yet.
• Planned Outputs: The study aims to produce:
  • Three reports on the adaptation process.
  • Three reports on implementation and reception per site.
  • Comparative manuscripts analyzing cross-site differences.
  • Specific publications on the Photovoice methodology and Mechanisms of Impact.
• Timeline: Data collection is scheduled to conclude in early 2027, with analysis and dissemination expected by mid-2027.

5. Significance

• Policy and Practice: The findings will inform the adaptation and scale-up of participatory approaches for Non-Communicable Disease (NCD) prevention in resource-constrained health systems.
• Fragile Settings: By including Afghanistan, the study addresses a critical gap in literature regarding health interventions in conflict zones, offering lessons on feasibility and safety.
• Sustainability: The evaluation assesses the potential for long-term sustainability and integration into local civil society and health structures, moving beyond the trial period.
• Transparency: Publishing this protocol enhances transparency in clinical trials and sets a standard for reporting process evaluations in LMICs, ensuring that future interventions are evidence-based and contextually responsive.

Trial Registration: ClinicalTrials.gov (NCT06561126 and NCT06570057).
Ethical Clearance: Approved by the National Bioethics Committee of Pakistan and the Aga Khan University Ethics Review Committee.