Billy Boland: Fail fast, learn fast

billy_bolandI have been talking about failure a lot at the moment. How it should be encouraged and celebrated. We all do it often enough, so we might as well try to enjoy it. Being open to failure and learning from it is a key skill in quality improvement (QI). The more you do it, the more chance you have of succeeding.

When I was working in academic psychiatry, I struggled with the upfront planning that is necessary for research design. Big trials, randomised controlled trials, and the like can’t afford to fail. Upfront investment (of time and money) is sizable and a lot rides on getting it right. If you realise that you haven’t accounted for significant co-founders or that your samples are wonky and unrepresentative once the data is in, it can already be too late. Thousands of pounds have already been spent, and hours of yours and other people’s time have been wasted. Failure is not an option. I found it stressful.

But the rapid, small scale experimentation that can be used in QI appeals to me much more, and is suited to the kind of figuring out that we want to do when we seek to change things for the better in healthcare. The Plan-Do-Study-Act (PDSA) cycle can often be done small scale, in the clinic, and quickly.

I saw a great example of this at a QI conference at the Royal College of Psychiatrists recently by Dr Japna Satara, a foundation doctor working in psychiatry. She wanted to introduce physical health checks for patients in the mental health crisis team. Her first intervention, staff training, had a modest impact but not as much as she had hoped for. She wasn’t satisfied. She went on to produce a standardised template to aid in completing and recording physical health checks and measured its impact again. Sure enough, performance improved once more. In developing her interventions she sought to understand potential levers for change (for example, staff wanted a tool to support the implementation of their learning), then tackled them. The whole process was done in her four month placement. Dr Satara won the oral presentation prize at the conference for her work.

It’s the practical application of experiments in the workplace that makes it so attractive. By adopting an experimental attitude, it opens the door to trying things out. It can minimise the personal risks associated with change and make it easier for clinicians to have a go. It can be easier to try out something new if you don’t see it not working as a kind of personal inadequacy.

Failure in this way is a source of intelligence. If a change didn’t work, what does that tell you? With any change you need to measure what happens as a result. If things get worse rather than better, you might want to think again or stop your change altogether. Or if they are getting better but not as much as you would like, what else can you do?

So I’m less afraid of failure now. I have a different attitude to it. Of course, I still have a weakness for getting things right the first time, and can enjoy unwarranted pride in success straight off. But generally now I can see things not working out as an opportunity for learning. On a good day, I can even enjoy it. Give it a try.

Billy Boland is a consultant psychiatrist and associate medical director for quality and safety at Hertfordshire Partnership University NHS Foundation Trust. You can follow him on Twitter @originalbboland.

Competing interests: None declared.

  • evidmed

    I think generally concentrating and feeding back which parts in the system work successfully (in the “study” part of the QI cycle) not only increases staff morale, but also is a vital solid block to build on.

    I always found trials too narrow in selection, publication of trial results too bound by statistical projections, then critical appraisal methods too academic. It is always the implementation which is interesting as it is real. That is where QI projects excel. It is real life and if it works, it works in the area/system/staff group for which it was intended …and so the reward is so much more rewarding..