Feel the fear and do it anyway: Improving health services and systems carries clear risk of failure

In high-income countries such as Aotearoa New Zealand, the public expects a great deal of our health system.

We expect our health services to be effective at producing positive clinical outcomes – a very reasonable expectation, built into all health professional training, that health services are delivered safely. We expect health services to be accessible to people who need them.

And we also need health services to be delivered efficiently, because any dollar wasted is a dollar that cannot be spent where it will be more beneficial.

Health services certainly need to respond to patients’ needs and expectations. Thinking on a bigger scale, services need to be resilient and adaptable in the event of major shocks (earthquakes, pandemics, and so on).

Our health workforce also needs to be valued and not subject to unmanageable stresses.

All these health system values are important, but they are also often in tension with one another. A resilient health service is unlikely to be a lean, efficient one. There are well-known trade-offs between geographical equity of access to surgery, and surgical quality and safety.

And it is certainly possible for a health service, or a whole health system, to do a reasonable job of managing these multiple imperatives, while still exhibiting poor outcomes at the population level, and producing major inequities of health outcomes between population groups.

Tradeoffs occur every day, at every level of our health system

When thinking about what a high-performing health system looks like, population health and more equitable distribution of health outcomes now sit at the very core.

However, the sorts of actions and decisions required to enhance population health and to reduce inequities have the potential to – in the short term at least – adversely affect other health system values. For example, funding a new drug that may clearly have beneficial clinical outcomes for groups that are already comparatively well off, but have little benefit for Māori and Pacific populations, would increase inequities, and be less justifiable if equity is prioritised.

Addressing population health and inequity, therefore, increases the complexity of our health system challenges.

Health managers, clinical leaders, policy-makers and the system as a whole face the short-term and long-term challenges of how to manage potentially conflicting imperatives. These trade-offs occur every day, at every level of our health system.

The idea of mapping and managing trade-offs has always been important in the training of policy analysts. Most often, formal approaches to training “policy designers” has focused on techniques for calculating the costs and benefits, advantages and disadvantages of specific policy proposals against multiple criteria, and then finding the optimal trade-off. The results depend largely on how different criteria are weighted against one another.

We see this rationalist approach attempted in how Pharmac makes decisions about funding new drugs (although we don’t get to see how the criteria are weighted), and in how waiting lists are managed for elective surgery (via clinical prioritisation access criteria).

But there are major informational limitations to these rationalist approaches. They require a central “brain” with access to credible, quantifiable data about benefits, other consequences, and costs, when such information is not readily available for most health services.

Health systems are hardly unique in requiring the management of multiple, potentially conflicting imperatives. Many other systems do exactly that and examples can be found across an enormously diverse range of phenomena, including biological evolution, the growth of cities, software programming and the evolution of languages. All are examples of complex systems in which multiple parameters are juggled, and none are governed by a central “brain”.

A particular image is useful for understanding the implications of complexity. Imagine you are tramping in an area in which the landscape consists of one large mountain rising from a plain. You want to reach the highest point above sea level in this landscape, and clearly this can be done by climbing to the top of the peak. This mountain is smooth and rounded such that it is possible to climb it from any side.

Now imagine that you are dropped by helicopter somewhere in the Southern Alps without a map. You are near the top of one mountain, but you think there may be higher mountains around you, but you can’t be sure because the mountain range is covered in cloud. To reach a higher point than where you are, you will need to tramp downhill before you can ascend the neighbouring mountain, which might not be possible to climb. There is a considerable chance that you will end up at a lower altitude if you decide to leave your starting point.

The first landscape is “smooth”, the second is “rugged”. Complex systems are analogous to rugged landscapes – and sophisticated spatial models of systems involving multiple parameters actually produce this pattern.

I find the image of a rugged landscape most useful as a metaphor for how to think about improving health services and health systems. It suggests that this process is more one of experimentation, with the clear risk of failure, than one of omniscient calculation by a central brain.

Most of the talk about improving health systems – in legislation, policy documents and local plans – draws from an image of improvement that evokes a rather smooth landscape in which everyone knows, and agrees, which way is up.

But health system improvement in a complex system involves considerable risk, such that in attempts to change things for the better, at whatever level of the system, it should be expected that some important aspects of the health service deteriorate.

But the risk of failure should be weighed against the downsides of staying put.

Even if this is a more realistic and pragmatic image to guide health sector leaders, the high political visibility of health policy will, unfortunately, preclude our political leaders from portraying our health system as a complex system, even if they understand from experience the ruggedness of health policy.

Tim Tenbensel is professor, health systems, in the Faculty of Medical and Health Sciences at the University of Auckland