In an ideal world, large healthcare providers such as the UK’s National Health Service (NHS) would allocate funds to the places where patients need the most care. But the real world is full of patients who don’t stay in one place, crowded cities packed with health services and less densely populated rural areas with few clinics or doctors.
Healthcare is a complex and deeply personal issue. Some people refuse to seek medical advice unless absolutely critical, and it’s long been debated whether people seek medical services because they need to or simply because they have easy access to them. Looking at patient outcomes hasn’t provided a definite answer, and the very availability of services in some areas may introduce a statistical bias that’s hard to eliminate.
What’s for certain is that frontline medical and healthcare services are critical at both an individual and society level. A positive human future hinges on ensuring major public health spenders such as the NHS are able to track how they spend funds so they can prove value for money and accurately plan for future service provision.
At PA, we work with public bodies, governments and regulators to develop pricing and cost strategies, and run complex modelling based on the latest data science techniques to help find an answer to this challenging question – and ultimately to provide better outcomes to end users.
By using new ways of collecting and analysing data, combined with pioneering research, it’s becoming easier to understand the correlation between patient need and location, and see how that affects spending. In doing so, we believe it’s possible to better meet the needs of patients and ensure healthcare services are run in a way that enables them to do more with less.
When MIT economists Heidi Williams and Amy Finkelstein, and Matthew Gentzkow of Stanford, explored the issue in 2016, they found 50-60 per cent of the variation in healthcare service volumes was due to place-specific factors rather than underlying need. In Miami, the researchers found healthcare providers spent $14,423 per Medicare patient in 2010, the same figure in Minneapolis was just $7,819 per patient.
This isn’t an issue for the US alone. One of the main aspects of our work with the NHS has been to explore how it can allocate its money differently to achieve better outcomes. To gain better insights into policy approaches that could reduce variations in spending allocations, we used complex data collection and analysis techniques to model the amount of patient activity at an English NHS Trust. This enabled us to forecast the financial implications of different scenarios for the next 20 years.
By applying this knowledge within the NHS, we can distinguish between spending driven by patient need and spending driven by levels of accessibility. From that, we can work out what the patient need is, and the NHS can allocate funding in a more cost-effective way.
These data-driven techniques have big implications for future health policy in the UK and beyond. For example, health systems could pilot approaches that factor internal migration effects on health spending. When we ran initial tests on data from the Office for National Statistics, we found enough people move around the UK health systems to understand the impact of place on demand.
We also need to better understand unwanted differences in clinical practice. So far, insights into this have come more from basic benchmarking approaches, such as the NHS Right Care analysis. A more sophisticated picture of what variations are controllable, and what is driving them, can create more standardised practices across the whole health system.
Through better data collection and analysis and the right insights, we can make those connections between patients and place, create efficiencies, shape policy approaches, improve financial forecasting and drive more consistency – ultimately creating a positive human future for patients.