Re: Changing how we think about healthcare improvement
Jeffrey Braithwaite makes a cogent case for an appreciation of ‘complexity thinking’ in healthcare, drawing on an obvious wealth of personal and published experience.  The challenge for practitioners is to draw on such insights to actually deliver pragmatic approaches to improving existing efforts to deliver care. Such approaches may be informed by experience elsewhere – while Braithwaite asserts that no other industry is as complex as healthcare, many other safety-critical fields have developed a robust approach to managing complex systems which healthcare may benefit from. Several of these such as defence, aerospace, manufacturing, and construction are linked by the academic and practical experience within the Engineering and Design communities. These are in turn closely allied to the fields of Operational Research and Improvement Science – all have well-developed theories around systems and complexity with a range of practical tools to explore, create and evaluate system improvements.
Recent work by the UK’s Royal Academy of Engineering, Academy of Medical Sciences, and Royal College of Physicians to develop a consensus systems approach to improving healthcare has been summarised in the flagship report ‘Engineering Better Care’. This draws on an extensive programme of workshops, discussions and case studies from across the different communities and looks to distil the essence of a systems approach into a series of iterative questions which need to be answered by those seeking to improve care, whether at the local or national level. These questions support Braithwaite’s thesis that both people and systems are at the core of improvement, but goes further to suggest that the areas of design and risk, while less well explored in healthcare, are similarly fundamental. Indeed, risk management and a formal design process are often wholly lacking from medical improvement initiatives.
Many terms used here may need unpacking for the uninitiated. ‘Systems’ involve people, equipment, processes, institutions, and culture and are as applicable to direct clinical care as to large-scale healthcare provision. ‘People’ involve all those – often beyond the obvious – who have a stake in the area of interest. ‘Design’ is not limited to the creation of physical artefacts, but encompasses processes, experiences, or policies. Finally ‘risk’ involves not simply a retrospective attempt to prevent a previous adverse outcome from recurring, but a prospective attempt to both minimise harm and promote excellence. A ‘systems approach’ is then a coordinated effort to address each of these areas, within the resource limitations of a given project, so as to improve care without creating further harm.
One aspect of this work which may be challenging to clinicians is that while clinical insight may be necessary to improvement efforts, it is by no means sufficient. What appear to be obvious solutions, with a clear causal chain, may turn out to be futile or even harmful once scrutinised through a systems lens. Clinical requirements are in fact a subset of the wider suite of stakeholder needs which need to be traded off against each other in order to maximise the chances of effecting positive change while managing risk. This may require facilitation by experts steeped in the necessary methodologies rather than being owned entirely by clinical leaders. This should not be surprising – as noted elsewhere, hospitals would not use clinicians to defend them in lawsuits, or indeed to plan infrastructure developments or manage human resources. In each of these areas a clinical appreciation is essential but needs to be matched by appropriate expertise from elsewhere. As Braithwaite states, improvement efforts are stalling – we should not be afraid to engage with, and learn from, experts in the management of complex systems if we are to tackle this.
Competing interests: PJC was the lead author for the Royal Academy of Engineering report 'Engineering Better Care'. TB, JW, and AK were all contributors to this report.