Written by Martin Marshall, Lead for Improvement Science London and Professor of Healthcare Improvement at UCL.
First posted on Improvement Science London on July 18, 2012.
The English language is a fascinating thing; powerful in the hands of Shakespeare and Yeats, inadequate in the hands of, well, me. An uncomfortable light has been shone on my communication skills in recent months as I have tried to explain and promote the science of improvement. The people that I want to influence – decision makers in the health service and academics – bring a healthy mix of enthusiasm and scepticism to the table. At the start of one of our first ‘engagement’ events I asked the participants whether they were familiar with and understood the term ‘improvement science’. To my surprise, the vast majority of people put up their hands. There then followed a forensic hour-long examination of both words, on their own and in combination, and a lively, sometimes heated, discussion. I had planned to repeat my straw poll but I wasn’t sure that I wanted to reveal what I suspect would have been a complete reversal of the earlier poll. The science of improvement may be an intuitively appealing term but it is a problematic one too.
Since that first meeting people have become more interested in the principles of improvement science and less inclined to search for a tight definition, but the question refuses to go away. ‘I really like the features of the science of improvement’, people say, ‘the ways in which it brings together the expertise of service based decision makers and academics, but is it really a science? Surely improvement is a goal, an ambition, perhaps even an imperative, but not a science?’
I want to rise to this challenge. I have no doubt that describing improvement in scientific terms is both reasonable and useful. I’ll justify this view by going back to the basics. Science is a way of knowing. It is not the only legitimate way of knowing (contrary to what some of my tutors told me at medical school), nor should it necessarily be described as the ‘best’ way of knowing, but it is the most rigorous approach to acquiring knowledge. It is characterised by systematic ways of thinking, using observation and/or experimentation to build theories and evidence that aim to produce knowledge that can be generalised or transferred beyond the location in which it was created. This contrasts with other common ways of acquiring knowledge such as those based on superstition, authority, intuition, rationalism or experience. Science has developed as a way of reducing the intrinsic biases associated with these popular and every-day ways of thinking.
So, let’s look at efforts to improve services at any level in the health sector, from large scale policy design to small scale practices in the front line. You are likely to see some activities that are planned in a systematic way, that are informed by the best possible research evidence and social theory, that are based on rigorous observational, and sometimes experimental, data, and that aim to produce learning that others can reflect upon, and replicate or adapt. It’s difficult to argue that this isn’t science, unless you are one of those people who think science only happens in test tubes. But you will also see shoddy improvement activities, poorly planned, ignorant of the evidence, detached from consensus views about how change can be achieved, using misleading data, unwilling to consider the wider implications of their work and inclined to make exaggerated claims of impact. This, in the words of Ben Goldacre, is Bad Science.
The application of Good Science to improvement work benefits patients. Bad Science is at best misleading, at worst wasteful and damaging. So, let’s be proud to call Improvement Science.