Video Transcript: Running More than One PDSA

David M. Williams, PhD, Improvement Advisor, IHI

There’s no real science in terms of saying, “Here’s the right answer.” There are a couple things that happen. One is that you want to do PDSAs that are building on each other. So, when I have a particular change idea, I want to keep testing that idea and building based on the last test, and the last test, to get more knowledge — and go until I feel like that change is something I learned about, and it’s working.

At the same time, no single change is ever “the thing” that’s going to likely result in your improvement. So, you’re likely to be doing other things that you want to test concurrently — and that’s okay, as well. That helps you build knowledge about this sort of system of change ideas that will result in the outcome.

The one caution that I always throw out is I want people to be careful to not intermingle them too much — so that you can’t tell which one made a difference or not. There are some more advanced strategies — like planned experimentation — that’ll enable us to turn on change one and two but leave off change three and four and do combinations to try to recognize relationships of what works — but, in general, I try to help people focus in on saying, “I want to look at this particular change or these parallel changes, and do them in a way that’ll help me learn, and, as I discover which change works in what way, then I can start to think about how they interact with each other.”