What can CLABSIs and cucumbers teach you about PDSA cycles?

Don Goldmann, MD, Chief Medical and Scientific Officer, Institute for Healthcare Improvement

Iím going to talk to you today about Plan-Do-Study-Act cycles, commonly known as PDSA cycles. So letís just go over for a second what those terms mean. And Iíll start by telling you that this is the method we use in science, we use it in everyday life, and we certainly use it in improving health care.

So, plan. By plan, we mean we are going to run an experiment. We have a hypothesis. We are making a prediction about what will happen when we run this experiment and we are going to determine whether or not that prediction holds up. Thatís where the ďdoĒ comes in. We run the experiment and then we study. Thatís the ďs.Ē We see whether or not our prediction was verified when we acted on our plan. Based on what we find on the data we get from the ďPDS,Ē we then revise our plan if necessary and act to change our plan and have another hypothesis and run another experiment.

So thatís pretty simple and youíve probably seen many examples of this in health care even if nobody called it a PDSA.

Letís look at an example of trying to prevent central venous catheter-associated blood stream infections, or CLABSIs. Now, we all know that a checklist has been shown to be very effective in getting folks to reliably do the right evidence-based practices when they put in a line and that reduces infection. The question is, how do you actually get that done in practice? How do you execute on that evidence-based plan?

Well, we know that among the things youíre supposed to do when you put in a central venous catheter are to prep the skin with a certain kind of antiseptic, chlorhexidine, and to use a large surgical drape. Those are two of the elements of the checklist. The question is, how do you get them to be at the bedside reliably so that the care team has them at their beck and call when they are ready to put in the catheter?

So, the physician is ready to put in the catheter. You have a theory that if you include the surgical drape and the antiseptic in the packet of supplies the team brings to the bedside, the physician will always be able to do it. He wonít have to be looking all over; she wonít have to be hunting for these supplies. And thatís your theory and you predict this will improve care. So you do that experiment and, lo and behold, not surprisingly probably, the drape is there, the antiseptic is there, the physician uses them when he puts in the catheter and an infection is probably less likely. Thatís a PDSA cycle in health care.

So I know youíre dying to try this in health care and thatís great, but I recommend that first, you try it in your personal life so you get comfortable with it. So that when you go to health care, you feel like, Iíve done this before (and) this isnít that hard.

But Iím going to talk about, of all things, cucumbers. I care about vegetables. I have raised beds. I spend a lot of time and effort. I donít want stuff to go wrong. So this year, I decided Iím going to plant cucumbers. I prepared the soil, put in the fertilizer, dropped in the seeds ó 3 seeds to a hill ó they sprouted, thinned them to one, and things (are) growing like crazy. I run it along my bed and am looking at it with admiration.

Then, I come back from work to home and I see that itís been chopped off by a weed wacker.

Turns out, every Monday thereís a guy that comes and he does our lawn and he has a weed wacker and he chopped off my cucumber. And this results in a pretty stunted cucumber.

So I said, look, I have a theory that if I propped this cucumber vine up on the raised bed more carefully, the weed wacker wonít get to it. Thatís my prediction. And so I put the cucumber vine up on the raised bed, came back and, by golly, the head was chopped off again. My prediction was not correct. When I studied the result, it was a disaster for my poor cucumber. So I had to revise my plan. I had to act in a different way.

My next experiment was to put a large flower pot in front of the cucumber vine, hoping that the guy with the weed wacker would understand that he wasnít supposed to meddle with this thing and, lo and behold, I come back and another PDSA cycle failed. Heíd been mangling my cucumber again.

At this point, I thought about something that you probably use in your life and we certainly sometimes default to in health care ó and thatís called brute force. I wanted to find the guy, the weed wacker guy, and basically grab him by the shoulders and shake him and say, ďYouíre killing my cucumber.Ē But I restrained myself. This is improvement, and we donít want to have blame. We want to do things better, make things better, for the weed wacker guy as well as myself.

So my next PDSA cycle was to put the cucumber up on a trellis. I figured if I had a trellis ó and I have a very pretty one, as you can see itís in the shape of a leaf; itís really quite attractive ó I prop the cucumber up on that so that the weed wacker really couldnít reach it. And lo and behold, look at it, it thrived. My prediction was: Get the thing out of the way so that the weed wacker absolutely canít get to it. This is called a forcing function. The forcing function is that the darn weed wacker just wonít get that high. He canít do it. And my cucumbers are thriving. Iím telling you I had the best marinated cucumbers ó a little onions, salt, and vinegar, just like my mom used to make it. This was a really good PDSA cycle.

So thatís all there is to it. You can do that in your own life. You can do it in health care. Remember, you have to have a precaution. You have to start with a theory. Your job is to see whether or not your theory is a good theory. Most of our theories, we have some doubt. So go test it. Run a PDSA.