After over 16 years as a GP, I can tell you what keeps me up at night. It is not the long hours or the difficult diagnoses. It is wondering how much of a difference the care I provide makes for my patients. 

I know my patients well. I know their histories, their families, their struggles. But I cannot tell you with confidence how many patients I have prevented from attending the emergency department or being admitted to hospital. I am unsure how my referral patterns compare to those of my colleagues.  

This is not just my problem. It is New Zealand’s problem. 

Other countries have figured this out. When doctors can see patterns in their own work – whether they prescribe safely or manage chronic conditions well – they improve. When policymakers see what works, they stop guessing. 

We already have the Integrated Data Infrastructure, which links health, education, income and housing data. The missing piece? General practice. Where most health care happens. 

Adding this would give us a complete picture of how health care works for people in their communities and how intersectoral collaboration can work to improve people’s lives. For example, we can see the impact of improved housing standards on conditions such as asthma. 

Canada shows what is possible. Since 2008, over 1,500 GPs have contributed data covering two million patients. The system is run by clinicians, not bureaucrats. Data are stripped of identifying information, analysed only for approved questions, all access is logged, and researchers can only access the part of the database which is relevant to their area of research. 

Here is what matters: doctors get useful feedback. They can see their prescribing patterns, compare themselves with peers, identify patients who need follow-up. When GPs saw their antibiotic prescribing rates, many changed their behaviour. Care improved as unnecessary antibiotic prescribing dropped. 

We could do this. Put GPs in charge of governance. Strip identifying information at source. Provide dashboards that help us better care for our patients. Start small by allowing practices to opt in, then actively expand to rural and high-need areas where the data matter most. 

Privacy concerns are real, but solvable. Patients can opt out. Access is controlled and logged. The Canadians won an international privacy award for their approach. 

Done right, we would finally know whether our care is truly working. We would be guided to improve our prescribing practices. We would know how many acute hospital admissions and emergency department visits our care prevented.  We would know whether different communities are getting equitable treatment. 

I spent nine years building a general practice from scratch. I know what works in theory does not always work in reality. But this does work. Other countries prove it. 

Better data cannot treat patients alone, but better data will help every doctor who does. 

To find out more, read Prabani's research note "Better health through better data" here.​