Paul Holmström researches system dynamics in healthcare: Testing ideas before they affect patients
“Healthcare is too important to fritter away in short-term financial measures or political opportunism. We need a common perception of problems and an understanding of the system,” says Paul Holmström.

Testing ideas before they affect patients

When researcher Paul Holmström fell ill, he found himself enmeshed in the very healthcare system that was the focus of his work, and saw even more clearly what improvements are needed. Now he is presenting his licentiate degree, which shows how system dynamics can be applied to develop the healthcare system. The basic idea is to test ideas for improvement in a model first, to avoid putting the day-to-day activities, or patients, at risk.

​Paul Holmström, an industrial doctoral student at Chalmers who is linked to the Centre for Healthcare Improvement, sees many opportunities to improve Swedish healthcare. He is passionate about introducing system dynamics as a method in development efforts, because it allows testing of strategies in theory, before they are put into concrete use.
“It’s better to disrupt operations in a computer scenario than in real life,” Holmström explains. “I was captivated by system dynamics after I saw its potential for testing out policies and strategies at an early stage to eliminate things that don’t work and allow us to focus on things that give results. System dynamics gives us the tools to create an image of the system and all its internal links so that we can simulate changes and examine their consequences over time, pinpoint problems and try out solutions.”

New insights as a patient

He has spent the last few years gathering extensive empirical data in the area. His licentiate work describes 13 cases, and he has contributed to 15 articles and conferences so far. In the midst of his studies he fell ill himself, and as a cancer patient he suddenly got a look at the system he was studying from a completely different perspective.
“It was surreal suddenly being surrounded by the processes I was studying, and experiencing first-hand the paradox between the encounter with skilled professionals and the bizarre system in which they were trying to do their best work. The healthcare system is one of our most important societal functions. It is too important to fritter away in short-term financial measures or political opportunism. Because the healthcare system is complex and complicated, we need tools to allow us to overview the entire system.”
Holmström points out that the healthcare system is characterised by huge variation in patient flows, giant numbers of possible diagnoses and also multiple diagnoses. He sees many advantages to applying system dynamics as a method to deal with this complexity – and to work preventively.
"Between one-third and half of the costs in the healthcare system are related to what we eat, drink and smoke and how much exercise we get. Preventive measures are the solution to the healthcare system’s cost problems. System dynamic modelling allows us to examine the effects of measures over a long period of time in order to choose those measures that have the best outcomes for individuals and society", he says.

"We can simulate the peaks to initiate a relevant discussion on how to manage them without extended waiting times or overworked staff.”

The trimmed-down healthcare organisation we have today presents huge problems because it makes peaks difficult to manage. Here, too, system dynamics can be used as a key to improvement, Holmström says.
“The healthcare system is under extreme pressure. After taking a look at obstetrics and A&E departments, I’m not surprised that staff are looking for other jobs. I get the impression that resource allocation is based on averages with no consideration of how to manage peaks. Resources have been cut back to such a degree that staff perceive the peaks as taking up more time and volume than before. We can simulate the peaks to initiate a relevant discussion on how to manage them without extended waiting times or overworked staff.”
Holmström also pursues action research, which he wants to link to system dynamics. Action research means that researchers are not standing on the side nor analysing events after the fact, but are directly involved in the fields, processes or phenomena to be studied – and they create knowledge in collaboration with those affected.

Action research an important foundation

Combining system dynamics and action research increases the likelihood that the results will lead to genuine change in the organisations, Holmström says.
“Many technically brilliant simulation models have been developed using system dynamics, yet have never led to practical use. One meta-study showed that only 5.3% of published simulations led to implementation. My hypothesis is that they developed their models ‘based on’ the operations, rather than ‘in collaboration with’ them.
“When I work with action research, I get together various stakeholders and combine their perspectives to give them a common understanding of how the system works. After that, we can move on to testing improvement measures in a simulation model, at no risk to the day-to-day operations or to patients.”
The United States, the UK and the Netherlands all apply system dynamics to the healthcare system to a much greater degree than Sweden, where Paul Holmström is realtively alone in his research field. He hopes that his work will contribute to changing that, pointing out that the Swedish healthcare system has a lot to gain from applying system dynamics to develop the healthcare system.
“I show that system dynamics simulation is a practical method for preserving the high standard of Swedish healthcare in the long term at a reasonable cost,” he concludes. “I’ve carried out many smaller projects, but the true benefit is at the regional and national levels.”

Text: Ulrika Ernström


Paul Holmström is an industrial doctoral student at Chalmers’ Department of Technology Management and Economics, and is linked to the research centre CHI – Centre for Healthcare Improvement.
Read Paul Holmströms´ licentiate thesis Bringing System Dynamics into Action Research

Published: Wed 10 May 2017. Modified: Thu 11 May 2017