Digital monitoring and action planning to reach zero-dose and under-immunised children: Leveraging data for targeted immunisation responses

This implementation research demonstrates that a co-created digital monitoring and action planning tool, piloted in Lahore, Pakistan, effectively strengthens routine immunisation systems by enhancing real-time data visibility, improving decision-making for zero-dose and under-immunised children, and fostering stakeholder ownership to address coverage inequities.

The Gist

Imagine you are trying to find a few specific lost toys in a giant, messy attic. You know they are there, but the lights are dim, the boxes are unlabelled, and you don't have a map. This is exactly the situation facing Pakistan's immunization program. They know there are children who have never received vaccines (the "zero-dose" children) or who stopped halfway through their shots, but finding them in a sea of millions is incredibly difficult.

This paper describes a project that built a digital flashlight and a smart map to help health workers find these lost children and get them vaccinated.

Here is the story of how they did it, broken down into simple parts:

1. The Problem: The "Invisible" Children

In Pakistan, many children miss their vaccines. Some have never had a single shot; others drop out before finishing the series.

• The Analogy: Imagine a school where the teacher has a list of 1,000 students, but 50 of them are missing. The teacher doesn't know who is missing, where they live, or why they aren't there. The list is just a blank spot on the page.
• The Reality: Health managers in Pakistan had similar blind spots. Their data was often slow, incomplete, or stuck on paper, making it impossible to react quickly to help the children who needed it most.

2. The Solution: Building a "Smart Compass"

Instead of just building a new app from scratch in a lab, the researchers decided to build a tool with the people who would use it. They called this "co-creation."

• The Team: They brought together computer experts, health managers, and the actual vaccinators (the people who go door-to-door).
• The Tool: They built two parts that work together like a remote control and a TV screen:
  1. The Mobile App (The Remote): Vaccinators use this on their phones to visit homes. Even if they are in a neighborhood with no internet, they can type in the data. It's like writing a note on a piece of paper that magically saves itself.
  2. The Dashboard (The TV Screen): This is a big, colorful map for the bosses. As soon as the vaccinators upload their data, the map updates. It lights up red in areas where children are missing vaccines.

3. How It Works: The "Traffic Light" System

The most clever part of this tool is how it tells managers what to do next. It uses a simple traffic light system based on how many missing children are found in a small neighborhood (called a Union Council).

• 🟢 Green Light (0–3 missing kids): Everything looks okay. Just keep an eye on things.
• 🟡 Yellow Light (4–8 missing kids): Uh oh, there's a problem. The tool asks, "Why?" (Is it because parents are afraid? Is the clinic too far?). It then suggests specific fixes, like holding a community meeting to talk to parents.
• 🔴 Red Light (9+ missing kids): Emergency! This area is a "hotspot." The tool tells the managers to send a special mobile team to set up a vaccination camp right there, immediately.

4. The Pilot Test: Trying it out in Lahore

They tested this system in a small area of Lahore (a city in Pakistan). They visited 60 homes to see if the system worked.

• The Result: It worked like a charm.
  • Visibility: Suddenly, the "invisible" children were visible on the map.
  • Speed: Managers didn't have to wait weeks for paper reports; they saw the data in real-time.
  • Accountability: The dashboard showed exactly which areas needed help and tracked whether the health workers actually went to fix the problem. It was like having a GPS that not only shows you the traffic jam but also tells you when you've cleared it.

5. The Hurdles: It's Not Just About Tech

Even though the technology was great, the paper admits that technology alone isn't a magic wand. They faced some real-world bumps:

• The "Muscle Memory" Problem: Some older staff were used to paper and found the phones confusing at first.
• The "Key Person" Problem: When a boss changed jobs, the momentum sometimes slowed down because the new person didn't know the system yet.
• The "Battery" Problem: In some places, internet and electricity are unreliable, though the app was designed to work offline.

6. The Big Takeaway

The main lesson from this paper is that technology works best when it's built by the people who use it.

By listening to the vaccinators and managers, the researchers built a tool that didn't just collect data but actually helped them make decisions. It turned a chaotic attic into a well-organized room where every lost toy (child) can be found and put back on the shelf (vaccinated).

In short: They built a digital "search and rescue" team for unvaccinated children. By using a smart app and a clear map, they can now find the children who are falling through the cracks and ensure they get the protection they need.

Technical Summary

1. Problem Statement

Despite Pakistan's Expanded Programme on Immunisation (EPI) achieving high aggregate coverage (approx. 90% in Punjab), significant inequities persist. A critical gap exists in identifying and reaching "zero-dose" children (those who have never received a vaccine) and under-immunised children (dropouts).

• Systemic Issues: Routine reporting systems often render these children invisible due to weak feedback loops, inconsistent data quality, and a lack of real-time monitoring.
• Operational Gaps: Existing digital platforms (e.g., e-Vaccs, NEIR/EMR) focus on attendance and recording but lack the capacity to generate targeted action plans at the micro-level (Union Council/UC). Managers struggle to move from data visualization to actionable interventions for specific missed clusters.
• Context: In urban slums and peri-urban areas of Lahore, disparities in coverage are stark, driven by vaccine hesitancy, migration, and logistical barriers.

2. Methodology

The study employed an Embedded Implementation Research (IR) design under the MAINSTREAM Initiative, utilizing a co-creation approach to ensure the solution was contextually relevant and operationally feasible.

• Study Design: A mixed-methods approach combining quantitative pilot data and qualitative feedback (Key Informant Interviews, Focus Group Discussions).
• Study Site: A Union Council (UC) in District Lahore, Punjab, Pakistan.
• Sampling Strategy:
  • Technique: Lot Quality Assurance Sampling (LQAS) using Probability Proportional to Population Size (PPS).
  • Sample: 6 clusters identified within the UC, with 10 households per cluster, totaling 60 households.
  • Target: Children aged 0–23 months and their caregivers.
• Development Process:
  1. Co-Design: Iterative consultations with EPI managers, vaccinators, and IT specialists to define requirements (offline capability, GPS tagging, simplified UI).
  2. Pilot Implementation: Deployment of a custom mobile app for data collection and a Power BI dashboard for analysis.
  3. Fidelity Tracking: Field observations, real-time troubleshooting, and stakeholder feedback loops to assess acceptability, adoption, and adaptability.

3. Key Contributions & Technical Architecture

The primary contribution is a Digital Monitoring and Action Planning Tool that bridges the gap between data collection and operational decision-making.

A. Technical Stack

• Backend: MySQL database for structured, scalable data storage.
• Frontend (Field): Custom Android mobile application for vaccinators (supports offline data entry, GPS tagging, and real-time syncing).
• Frontend (Management): Microsoft Power BI dashboard for visualization, analytics, and action tracking.

B. Core Functionalities

1. Real-Time Data Visualization:
   • Filters data by Province, District, UC, and gender.
   • Interactive Maps: Geo-spatial visualization of zero-dose clusters at UC, district, and provincial levels.
   • KPI Tiles: Displays coverage rates, zero-dose counts, and maternal immunization status.
2. Threshold-Based Decision Support (Action Planning):
   The system uses LQAS thresholds (based on a sample of 60) to automatically recommend interventions:
   • 0–3 Zero-Dose Children: Low risk; continue monitoring.
   • 4–8 Zero-Dose Children: Medium risk; triggers analysis of root causes (e.g., unawareness, access) and suggests targeted community sessions or follow-up visits.
   • 9+ Zero-Dose Children: High risk; recommends immediate targeted vaccination campaigns or mobile outreach.
3. Hierarchical Drill-Down (Information Tree):
   Allows users to explore why a child was missed (e.g., linking sources of awareness to specific reasons for non-vaccination).
4. Accountability & Tracking Module:
   • Enables supervisors to assign tasks, set deadlines, and track completion.
   • Includes an "Edit" function for field staff to update action plans based on local context.
   • Features a printable checklist for field teams to ensure transparency.

4. Results

• Feasibility & Acceptability: The tool demonstrated high acceptability among vaccinators, supervisors, and managers. It was perceived as a tool that "makes work easier and more transparent," reducing paperwork and improving data accuracy.
• Visibility of Gaps: The pilot successfully identified zero-dose clusters and mapped coverage gaps at the UC level, which were previously invisible in routine paper-based systems.
• Fidelity: The intervention was delivered with high fidelity. Field workers accurately recorded data, and supervisors actively used the dashboard to assign and monitor follow-up tasks.
• Refinements: User feedback led to critical improvements, including:
  • Optimization of UC selection logic.
  • Addition of contextual tooltips.
  • Mobile-responsive dashboard reconfiguration (originally desktop-focused).
  • Implementation of a 30-minute auto-refresh feature for data syncing.
• Challenges Identified:
  • Barriers: Leadership turnover, system fragmentation, limited digital literacy among some staff, and workload pressures.
  • Infrastructure: Occasional syncing delays and connectivity issues in the field.

5. Significance and Implications

• Shift from Monitoring to Action: Unlike previous platforms that focused on recording, this tool integrates data visualization with actionable planning, enabling managers to not only see gaps but execute specific, tiered interventions.
• Equity and Accountability: By making zero-dose children visible and assigning specific tasks to supervisors, the tool strengthens accountability and promotes health equity in underserved urban and peri-urban areas.
• Scalability Pathway: The study outlines a clear pathway for scale-up, emphasizing:
  • Interoperability: Integration with existing national systems (DHIS2, NEIR, NADRA).
  • Capacity Building: Continuous training for data literacy and tool usage.
  • Policy Support: Institutionalizing the tool within routine workflows through policy endorsement.
• Model for LMICs: The co-creation model offers a replicable framework for Low- and Middle-Income Countries (LMICs) to design digital health solutions that are technically robust, operationally relevant, and sustainably adopted by local health systems.

Conclusion: The study confirms that a co-created, digital monitoring tool can effectively strengthen routine immunization systems by improving data utility, fostering accountability, and enabling targeted responses to reach the most marginalized children.