What can computer simulation offer?

The last half century has witnessed remarkable advances in computing power, machine intelligence, and emergence of novel means of measurement that have led to new and cheaper technologies, advances in science and industry, and more powerful tools to support the efficient functioning of societies. Over many decades, private industry and public sectors alike have harnessed these advances in numerous ways, not least of which is the use of computer simulation, artificial intelligence and supporting technologies to better understand complex systems, solve challenging problems, optimize the allocation of resources, and improve system efficiency, performance, and public safety. For example, complex technological exploits such as space and planetary exploration may not have been attempted without the use of computer simulation. It is difficult to estimate how many lives have been saved globally by our ability to model, simulate and forecast the path, severity and duration of significant weather events before they hit (NASA hurricane simulation system).

Computer simulation is used extensively in disciplines including nuclear physics, astronomy, climatology and meteorology, evolutionary biology, ecology, earth science, materials science and engineering. However, its use to support scientific advances and inform policy and planning decisions in the health and social sectors has lagged behind. For the most part, these sectors rely on comparatively rudimentary decision analytic approaches which, for complex problems, fail to meet decision support needs of policy makers and regional planners. This can lead to investment in interventions that ‘seem like a good idea at the time’ or ‘comprehensive’ approaches where investments are made in a wide a range of ‘possibly effective’ solutions. Such comprehensive strategies often lack focus, result in service systems that are crowded and difficult to navigate, lack sufficient actual investment in time, resources and capacity to implement programs and services effectively, and act to spread finite resources too broadly; diluting the potential impact of investments. What often results are strategies and reforms that produce little or no impact despite significant investments (Box 1), or adverse unintended consequences (Box 2). Further investments and reforms often follow, and the ad hoc trial and error pattern repeats over.

 

Box 1

Mental health and suicide prevention in Australia

In response to a significant rise in young male suicide between the 1970s and late 1990s, the Australian Government implemented the first National Youth Suicide Prevention Strategy which was later expanded to all age groups. Over the following decades, subsequent national frameworks were released advocating a comprehensive population-level approach to suicide prevention, states and territories adopted strategic plans for suicide prevention, over 150 community projects and 27 national initiatives were implemented, there was a Senate Inquiry, a number of Senate Committee reports, a National Action Plan, the establishment of an Advisory Council, community mobilization, and hundreds of millions of tax payer dollars invested, and additional investments made by businesses and philanthropic funding sources.

Despite this significant momentum and investment of resources, there was limited evidence of any impact on overall suicide attempts or suicides in the Australian population, with the exception of declines in suicide by carbon monoxide (CO) poisoning in males. This decline was attributed to changes in the lethality of CO poisoning, the result of a policy requiring new cars to be fitted with catalytic converters, to clean up the noxious exhaust gases produced by petrol engines, and unrelated to any suicide prevention activity or policy.

Despite this lack of impact, the iterative trial and error approach continues, with reform, restructure, additional funding, development of new strategic frameworks and National Action Plan, and ongoing use of traditional program implementation, evaluation and analytic tools and methods.

Box 2

Bed net program in Tanzania, a case study of unintended consequences

The Tanzania National Voucher Scheme (TNVS) was a public-private partnership (supported by donor funding) that provided pregnant women and infants with highly subsidized insecticide-treated nets for malaria prevention between 2004 and 2014. Vouchers were distributed to pregnant women at their first antenatal care visit, and again once the child was fully immunized at 9 months. Women could use the voucher as partial payment toward the purchase of a bed net. The Ministry of Health worked with the producers and distributors of nets in Tanzania to establish the Scheme whereby retailers could accept the voucher in return for a share of the price, remit the vouchers to a wholesaler in return for additional nets to sell. The wholesalers were then reimbursed by the TNVS for each voucher. The TNVS significantly increased household ownership and use of the nets and ensured continuous protection of the vulnerable before, during, and after the universal mass distribution campaigns implemented between 2009 and 2011. The TNVS stimulated and maintained a large national retail network which managed the bed net supply chain.

This scheme was a long running success with the health system integrated vouchers being used to sustain high bed net coverage (approx. 80%) in Tanzania for over a decade. However, the key supporting donor agency believed 80% coverage was too low and implemented a pay-for-performance strategy to incentivize health workers to further scale bed net coverage. This led to the gaming of statistics through collusion between health workers and retailers, which quickly became evident when coverage increased to over 130%. The donor then accused the system of fraud and disinvested due to their policy of zero corruption tolerance and shut down the program. Consequently, Tanzania went from having one of the most successful bed net distribution systems (providing the most vulnerable in the population protection from malaria) to no system. The pay-for-performance strategy has many predictable perversities that could have been mediated had there been the tools to virtually test the strategy before implementing it in the real world.

Developing computer simulation models that represent the complexity of health and social systems, and using them to simulate policy and intervention scenarios (before they are implemented in the real world), can be undertaken in ways that offer significant value to decision makers and researchers including their capacity to:

  • Leverage the tremendous growth of available evidence and data
  • Better understand causal pathways that lead to our most persistent health and social problems
  • Avoid ‘flying blind’ when determining how best to address complex problems with limited available resources
  • Support and nurture learning ecosystems that leverage data, expert and local knowledge and modelling insights to learn more quickly, reliably and deeply how best to improve population health and well-being.

Funding for capacity building in health sector applications of computer simulation is being provided sporadically in Canada, UK, Europe, and the US, including by the Centers for Disease Control and Prevention (CDC) and National Institutes of Health (NIH). But there is a long way to go. Nothing but a sustained, coordinated, global approach to capacity building will accelerate the development of the workforce and infrastructure needed to replicate the evolution seen in other sectors within our current generation. CSART is an alliance of existing centres of excellence in the application of computer simulation and advanced research technologies in health, with deep knowledge, long experience, a history of successful collaboration, and have the building blocks in place that have served as the hallmarks of success in other sectors. CSART is perfectly placed to coordinate global capacity building efforts to deliver more sophisticated decision tools to support health and social policy and catalyze systemic and sustainable improvements in population health and wellbeing. In turn, CSART is in a position to leverage its innovation networks to use generated improvements and insights to shape the design, implementation and refinement of future cutting edge technology frameworks.