Can applying a complexity lens to health organisations be more than an academic exercise with practical applications for service delivery? This question was central to a two-day meeting last week in London at the Health Foundation between researchers, policy makers and health practitioners. The event was organised by The Knowledge to Action Group (KTAG), an international team of senior academics and coordinated by two leading members of Fuse’s Complex Systems research programme, Prof David Hunter and Dr Emily Henderson. The group brought together the crème de la crème of health research and knowledge mobilisation, with about 40 handpicked participants from across the UK to debate a green paper on the topic (which is a tentative government report and consultation document of policy proposals for debate and discussion).
The essence of this meeting was about getting to grips with the consequences of doing research and developing interventions in a world that is inherently complex (which is not to be confused with just being complicated!). The green paper (Knowledge-to-Action: Addressing Complex Problems in Health Systems, May 28/29 2015) described complex systems as non-linear; dynamic; and having no single point of control. Because of these characteristics, outcomes in a complex system are unpredictable and resistant to interventions. So are health systems really complex? Anyone trying to get research evidence into practice will quickly discover that knowledge mobilisation is prone to all the characteristics outlined above for complex systems.
In one of the breakout sessions during the event the difference between complicated and complex was illustrated with the example of building a rocket versus raising a child. Building a rocket is complicated but will lead to a predictable outcome (successful launch) if all the different pieces are correctly assembled according to the guidebook. However, raising a child is complex (as many parents will testify) and often leads to chaos in many academic households. This is bad news for academic parents and health service managers: you cannot steer the thing (but you are still accountable for it) and even if you find the solution to a problem in the system it is likely not to work when you try it again. The dynamic nature of systems means that what works in one context or at one point in time may not work the same as the system changes.
How do you work within such a system? Fortunately, the event provided four case studies of research projects where the academics had managed to work within a complex health system to produce change across different areas, often in co-production with policy makers and health practitioners. The authors of the green paper distilled four themes from these case studies to inform future research and interventions:
- the balance of central and distributed authority in organisations and systems;
- the importance of emergence (continuous learning and adaptation);
- the need for co-produced knowledge; and
- a range of leadership positions and styles.
These four themes were discussed in more detail in various breakout sessions over the two days.
Participants embraced the spirit of the event and discussed a wide range of topics (from the need for a critical discourse department to finding positive deviants), whose relationships with the four themes were sometimes unclear and mostly unpredictable. Nevertheless, some key messages emerged from the event that will soothe the nerves of academics and service managers. By visualising complex organisations as “patterns of conversations between interdependent individuals” (as quoted in a presentation from Allan Best on the first day), working in complex systems becomes conversation management and engagement. According to the researchers involved in the successful case studies, this works all the better outside the normal context of everyday work by taking commissioners to the pub or baking a cake for a meeting with service managers to get the conversation really going. Isn’t it ironic that to work inside complex systems, one needs to go back to basics in a more simple system?