Data use practices and challenges for maternal and child health decision-making in tribal primary health centres in Andhra Pradesh, India

This qualitative study of tribal primary health centres in Andhra Pradesh reveals that while formal health information systems fail to provide the granular, prospective data needed for maternal and child health decision-making due to structural constraints, frontline workers have developed innovative, complementary information practices to bridge this gap, suggesting that digital health reforms should strengthen rather than replace these local ingenuities.

The Gist

Imagine a health system in the remote tribal hills of India as a massive library.

In this library, thousands of books (data about mothers and babies) are written every single day by the librarians (nurses and health workers). These books are carefully cataloged and sent up a long, winding staircase to the "Head Office" in the city. The Head Office loves these books; they use them to make big charts, check boxes, and report to the government about how well the library is doing.

The Problem:
The librarians at the bottom of the hill (the local doctors and nurses) are the ones who actually need to find a specific book right now to help a sick child or a pregnant woman. But when they ask for the books, the Head Office says, "We can't give you the books. We only send them up. Also, to get a book, you need a special key (a password) that only one person in the middle of the hill has, and that person's phone doesn't get a signal."

So, the librarians are stuck. They have the data, but they can't use it to make life-or-death decisions.

The Three Big Hurdles (The "Why")

The paper identifies three main reasons why this library system is broken for the people on the ground:

1. The "One-Way Street" Design: The digital system was built like a one-way street. It's designed to send information up to the bosses, but it doesn't send anything back down to the doctors. It's like a mailman who only delivers letters to the city but never brings any mail back to the village.
2. The "No-Signal" Lock: To open the digital system, you need a code sent to your phone via text message (an OTP). But in these remote, hilly areas, cell phone signals are as rare as rain in a desert. If you don't have a signal, you can't get the code. If you can't get the code, you can't see the data. It's like trying to open a door with a key that only works when you are standing in a specific spot with a clear view of the sky.
3. The "Single Keyholder" Bottleneck: Because of the signal issues, the system was set up so that one specific data entry clerk holds the master key (the phone with the login credentials) for the entire region. If that one person is sick, on leave, or just busy, nobody in the whole district can access the data. It's like having a library where only one person has the key to the front door, and if they lose it, the whole building is locked.

The "MacGyver" Solution (The "How")

Faced with a locked door and no keys, the local health workers didn't give up. They became like MacGyver, using whatever they had lying around to build their own tools. They created a "shadow system" to do the job the official system failed to do:

• The WhatsApp Group Chat: Instead of waiting for the official report, nurses and doctors started snapping photos of patients and sending them to a group chat. It's like a neighborhood watch group where everyone shares updates instantly. "Hey, Mrs. Reddy is due for her shot today!"
• The "DIY" Spreadsheet: One doctor, frustrated by the broken system, built his own tracking list on a free app called Google Sheets. He asked his team to fill it out manually. It's like a librarian who, tired of the broken computer, starts writing the catalog on a whiteboard because it actually works.
• The Thursday Meeting: They started holding weekly face-to-face meetings where everyone gathers to compare notes and fix errors. It's a "town hall" where they triangulate information from different sources to get the truth.
• The Human Bridge: They rely heavily on that one data clerk (the keyholder) to print out summaries and send them via WhatsApp. The clerk has become an unofficial "data hero," manually compiling the info everyone needs.

The Big Takeaway

The paper argues that we shouldn't look at these "shadow systems" (WhatsApp, spreadsheets, meetings) as failures or cheating. Instead, we should see them as brilliant local inventions.

The local health workers are not lazy or bad at their jobs. They are highly skilled problem-solvers who have built a parallel, working system because the official one is broken for their specific environment.

The Lesson for the Future:
If the government wants to fix maternal health in these areas, they shouldn't just throw new, fancy computers at the problem. They need to:

1. Listen to the locals: Understand that they need specific names and addresses, not just big percentages.
2. Fix the locks: Change the password rules so they don't rely on a single person's phone or a perfect cell signal.
3. Build on what works: Instead of replacing the WhatsApp groups and spreadsheets, the new official system should integrate them. Make the official system work with the local ingenuity, not against it.

In short: Don't try to replace the local "shadow library" with a new one; fix the door to the main library so everyone can finally get in.

Technical Summary

1. Problem Statement

The study addresses a critical paradox in India's tribal primary health centres (PHCs): while these facilities generate vast amounts of Maternal and Child Health (MCH) data through routine health information systems (RHIS), this data rarely informs local clinical or programmatic decision-making.

• The Gap: There is a disconnect between data collection (upward reporting for state/national monitoring) and data utilization (local action).
• Context: Tribal areas in India face unique challenges including geographic isolation, poor connectivity, and complex governance structures (specifically the Integrated Tribal Development Agency or ITDA model).
• Specific Issues:
  • Formal systems provide only retrospective, aggregate statistics, whereas local decision-makers need prospective, individual-level, geographically disaggregated data.
  • Recent security measures, such as One-Time Password (OTP) authentication, have inadvertently created access barriers in low-connectivity settings.
  • A critical governance vulnerability exists where a single data entry operator controls access to the entire sub-district's data portal.

2. Methodology

• Study Design: A qualitative study embedded in the diagnostic phase of an Action Research project, utilizing an interpretivist orientation.
• Setting: Three tribal PHCs under the ITDA Rampachodavaram in the Alluri Sitharama Raju (ASR) District, Andhra Pradesh. These sites were purposively selected to represent variations in terrain, connectivity, and tribal population composition.
• Participants:
  • 8 Key Informant Interviews (KIIs): 5 Medical Officers (MOs), 1 District Programme Officer, and 2 Data Entry Operators (DEOs).
  • Participant Observation: Conducted at weekly cross-sectoral convergence meetings involving MOs, Auxiliary Nurse Midwives (ANMs), ASHAs, and Anganwadi workers.
  • Document Review: Analysis of paper registers, Mother and Child Protection (MCP) cards, HMIS reports, and informal tracking tools.
• Analysis: Transcripts were independently coded by two analysts using Braun and Clarke's reflexive thematic analysis (six-phase approach). The Open Data Institute's data ecosystem framework was applied to distinguish between formal data flows and "soft value exchanges" (informal knowledge flows).
• Ethics: Approved by the Institutional Ethics Committee of Sree Chitra Tirunal Institute for Medical Sciences and Technology; followed the CONSIDER statement for research involving Indigenous peoples.

3. Key Contributions

This study offers several novel contributions to the literature on health information systems in low-resource settings:

1. Documentation of "Soft Value Exchanges": It systematically identifies and categorizes the informal information practices (WhatsApp networks, self-built tools, human intermediaries) that local actors construct to compensate for formal system failures.
2. Identification of a Structural Single Point of Failure: It reveals a previously undocumented governance vulnerability where a single Multipurpose Health Data Entry Operator (MPHDEO) holds the sole mobile device required for OTP authentication, effectively acting as a gatekeeper for all PHCs in the jurisdiction.
3. Reframing Local Practices: It argues that these informal practices are not "system failures" or mere workarounds, but expressions of local ingenuity and latent infrastructure that should be strengthened rather than replaced.
4. Context-Specific Governance Analysis: It highlights how uniform national policies (like OTP authentication) fail in specific tribal contexts due to connectivity and infrastructure constraints.

4. Key Results & Findings

The analysis identified three interconnected domains:

A. Unmet Information Needs (The "What")

Local decision-makers do not lack motivation or data volume; they lack accessibility and granularity.

• Need for "Names over Numbers": MOs require individual-level line lists tied to specific villages (Sachivalayams) to plan outreach, rather than aggregate percentages.
• Prospective vs. Retrospective: Systems currently report what was missed; MOs need systems that flag what should be done next (e.g., upcoming immunization due dates).
• Actionable Intelligence: Data must be available at the point of decision-making, not just at the district/state level.

B. Structural Constraints (The "Why")

Three structural barriers prevent formal systems from meeting these needs:

1. Infrastructure Mismatch: Digital tools designed for connected urban settings fail in tribal areas with inconsistent mobile networks. The 2024 shift to OTP-based authentication exacerbates this, as receiving an SMS can take 20+ minutes or be impossible without signal.
2. Upward-Only Data Flows: Systems are designed for supervisory accountability (data moving up), with no feedback loop returning actionable insights to the facility level.
3. The MPHDEO Bottleneck: Institutional login credentials for the RCH Portal are linked to a single MPHDEO's mobile phone. This creates a single point of access failure; if this individual is unavailable or transferred, all PHCs lose access to the formal system.

C. Complementary Information Practices (The "How")

To bridge the gap, local actors have developed four distinct practices:

1. WhatsApp Networks: Used for real-time case communication (photos, alerts) between ANMs, ASHAs, and MOs.
2. Self-Built Tracking Tools: MOs create personal Google Sheets trackers to replicate "due-list" functionality missing from the portal.
3. Cross-Sectoral Convergence Meetings: Weekly in-person meetings (Thursdays) to resolve data discrepancies and triangulate information across registers, specifically designed for days when connectivity is poor.
4. Intermediary-Consolidated Reports: The MPHDEO synthesizes data across facilities and distributes summaries via WhatsApp, acting as a human data intermediary.

5. Significance and Implications

• Policy Reform: Authentication governance for health portals must be context-sensitive. Centralized OTP systems in low-connectivity areas create access barriers that undermine system utility. Credential registration should move from individual mobile numbers to institutional devices or multi-user systems.
• System Design: New MCH digital tools must be preceded by participatory diagnostics of the local information ecosystem. They should integrate existing "soft value exchanges" (like WhatsApp) rather than attempting to replace them.
• Feature Requirements:
  • Offline Functionality: Not a feature, but a prerequisite.
  • Granularity: Tools must drill down from aggregate indicators to named beneficiary lists organized by geography.
  • Feedback Loops: Systems must return actionable data to the facility level, not just collect it.
• Strategic Approach: Complementary practices (convergence meetings, human intermediaries) represent latent MCH infrastructure. Reform should aim to formalize and strengthen these practices rather than bypassing them with new, disconnected systems.

Conclusion: The gap between MCH data and local decision-making in tribal India is structural, not behavioral. Effective digital health reform requires acknowledging the limitations of current governance models and leveraging the adaptive, local ingenuity already present in these communities.