The Data Gap in Māori Health

When I was managing information systems at RAPHS, I saw this gap up close. Our systems were good at capturing what happened in a clinical encounter: diagnosis codes, prescriptions, lab results, referral pathways. Standard primary care data.

But the Māori health programmes we supported operated on a different model. Whānau ora, community outreach, home visits, cultural assessments. The system wasn't built for that. So we improvised. Free-text notes. Workaround fields. Spreadsheets that lived outside the formal system.

That meant the official data told one story about our Māori health services, and the reality was something else entirely. The clinical data said we were delivering a certain volume of care. It couldn't say whether that care was culturally appropriate. It couldn't capture whether the whānau felt supported. It couldn't reflect the community health worker's assessment that the real barrier was housing, not medication adherence.

This isn't a technology limitation. It's a design choice made decades ago when health IT systems were built around a specific model of care, and that model didn't include te ao Māori.

Tania and I have talked about this at length. Her perspective, grounded in research and kaupapa Māori frameworks, complements what I've seen operationally.

She's right. And the operational consequence is significant. Most health IT systems treat data governance as an access control problem: who can see what, role-based permissions, audit trails. That's necessary but insufficient.

Māori data sovereignty, as articulated by Te Mana Raraunga, asks different questions. Who benefits from this data? Does the community it represents have input into how it's used? Are insights flowing back to the people the data came from?

In my experience, most health organisations answer "no" to at least two of those questions. Not because they're acting in bad faith, but because their systems weren't designed to support that kind of governance.

The standard primary care dataset in New Zealand captures a narrow view of health. It's optimised for funding models, reporting requirements, and clinical audit. It works well for those purposes.

But when you try to use that same data to understand Māori health outcomes, the gaps become obvious.

Determinants of health are invisible. Housing quality, food security, employment stability, cultural connection. These factors drive health outcomes for Māori communities, and none of them appear in standard clinical data.

Community-level patterns disappear. Health data is collected and reported at the individual level. But Māori health operates in a whānau context. A grandmother's diabetes management is connected to whether her mokopuna have stable housing. That connection doesn't exist in the data.

Intervention effectiveness is unmeasurable. If you can't capture what a kaupapa Māori health programme actually does, you can't measure whether it works. And if you can't measure it, you can't fund it. This creates a vicious cycle where the most culturally appropriate interventions are the hardest to justify in a data-driven funding environment.

The fix isn't a new platform or a better data model. It starts with who's in the room when these systems are designed.

Every health IT implementation I've been part of has had a design phase. Requirements gathering, workflow analysis, data modelling. The people in the room are usually clinicians, IT staff, and managers. Occasionally a patient representative.

Rarely is there genuine Māori participation in the design of the data architecture itself. Not just "what screens do you need?" but "what does this system need to capture about your community's health, and who should govern that data?"

That participation changes the outcome. When Māori health workers helped us redesign parts of our reporting at RAPHS, we ended up with a hybrid approach. The standard clinical data stayed in the formal system. But we built structured fields for community health assessments, cultural indicators, and whānau engagement that previously lived in free text or spreadsheets. It wasn't perfect, but it meant the data started to reflect what was actually happening.

Three things, none of them easy.

Health IT vendors need to design for multiple models of care. The Western biomedical model is one lens. It's not the only one, and systems that only support that lens will continue to produce incomplete data about Māori health.

Funding models need to accommodate what the data actually shows. If structured data about community-based, kaupapa Māori health interventions exists, funders need to recognise it. The current system rewards what the current data captures, which creates a bias toward clinical interventions over community ones.

Governance needs to be shared. Health data about Māori communities should include Māori governance. Not advisory groups that meet quarterly, but genuine shared decision-making about how data is collected, stored, analysed, and used. The operational complexity is real. But the alternative is continuing to make decisions about Māori health using data that doesn't represent Māori reality.

The conversation about data in health is getting louder, and new technologies will intensify it. The question is whether we use this moment to redesign systems that work for all communities, or simply digitise the gaps we already have.