Intended for healthcare professionals

CCBYNC Open access
Analysis Quality Improvement

Changing how we think about healthcare improvement

BMJ 2018; 361 doi: (Published 17 May 2018) Cite this as: BMJ 2018;361:k2014

So how do you do and research health improvement if it is so complex?

I welcome Braithwaite's analysis on the elusiveness of systems-wide progress in quality in health and believe that it is a wake up call to policy and decision makers, clinicians and researchers working in the field. He has clearly explained the value and contribution complexity science can bring to this issue. And has thrown down the gauntlet that our current conceptualisations are 'broken' and we need to change "our collective mindsets".

Braithwaite helpfully provides us with some recommendations on how to get change (see the "attractors" and "repellents" in Box 1 and twenty "enablers" in his table). But , to me, questions still remain (for example) about when and for whom should these suggestions be used? What outcome(s) result when they are used? How do these suggestions fit into the concepts of "work-as-imagined" and "work-as-done". To illustrate, in Box 1 "Systems can change when: ... Stimulated by medical progress...", but in what circumstances does this happen, for whom and to what extent, how and why? To progress we need to better understand these recommendations - what causal processes underlie then, when do they operate, for whom and what outcomes do they cause.

Fortunately, help is at hand. Using a realist lens to understand these recommendations would be one way to understand causation behind these rules and also to work out the contexts in which to use them, for whom and what outcomes they cause. If primary research is being used to make sense of these recommendations then a realist evaluation approach could be used (1). When evidence synthesis is needed to learn about the recommendations from across health improvement studies then realist review is more suitable (2).

For those wanting to pick up the more modest gauntlet of researching his recommendations using realist reviews and realist evaluations, so as to make them even more useful, quality and reporting standards and training materials and resources exist -

In addition, a vibrant community of realist researchers may be accessed through a JISCM@il listserv -

(1) Pawson R, Tilley N. Realistic evaluation. London: Sage; 1997.

(2) Pawson R. Evidence-based policy: a realist perspective. London: Sage; 2006.

Competing interests: I teach a module on realist review and realist evaluation: I have been and am currently paid to provide realist methodological support to a number of funded projects (for list of projects please see:

20 May 2018
Geoff R Wong
Clinical Research Fellow and GP
Nuffield Department of Primary Care Health Sciences, University of Oxford
Radcliffe Primary Care Building, Radcliffe Observatory Quarter, Woodstock Road, Oxford, OX2 6GG