I forget what hospital these charts are even from, but it doesn’t really matter. It’s become more common, in my gemba visits, to see hospital departments charting metrics. They are often part of “huddle boards” that teams gather around on a daily basis (or, ideally, each shift).
It’s good to see daily metrics, as opposed to monthly averages, since it’s better to have something closer to “real time” data — if you react and respond appropriately.
It’s also good to see simple, handwritten charts… that’s often part of what makes daily cycles possible as opposed to waiting for computer reports and fancy charts being printed out by somebody.
Below is an example of a chart with a daily average being plotted. 100% is ideal and a higher number is better.
In the chart, days with 100% are rewarded with a smiley face.
The low points in the chart don’t have accompanying frowny faces.
That’s probably well and good, because this looks like a pretty stable system to me. To my eye, without drawing a formal SPC chart, the average performance of the system is about half way between the top and bottom points in the chart. About half of the points are above average and half are below average, which is no surprise.
Below is the graph with my estimated average/mean in the blue dashed line:
In a stable system, we’d expect daily performance to fall between 100% and a calculated “lower control limit,” which I’ve estimated and drawn with a darker blue line.
SPC tells us that any random day would produce a result between 100% and the LCL. It’s all the same system. A smiley face on the good day probably doesn’t really reinforce anything positive, since it seems that people are basically doing their jobs the same every day. There’s likely not any better effort on good days than there is on bad days.
Factors that prevents a 100% performance level are systemic. It’s not as simple as asking “why did we have a good day today?” or “why did we have a bad day today?” It’s all the same system. We can improve that system, and should, but that’s different than troubleshooting or looking for a root cause on any particular single day with “bad” performance.
This organization has, like many, set a goal for what they consider to be “good” or “acceptable” performance. Here, they have already color coded green as a good day and red as a bad day (which doesn’t help if you’re color blind).
Note that the average appears to be in the red zone, so they are going to, ON AVERAGE, have a “bad day.” Asking “what went wrong?” on a bad day or having to explain each of those individual days is the wrong approach.
Asking “what went wrong that day?” is looking for a “special cause” when the system’s variation appears to be driven by “common cause” variation. A better question would be, “how can we improve the system?”
This same department had a second chart hanging right next to the first, see below. In this chart, down is better and the ideal performance would be zero.
This chart has many of the same dynamics. It appears to be a stable system that has an average and control limits as I’ve estimated and drawn below:
The higher points on the chart had explanations written down on the chart.
Those points don’t appear to be “special causes,” as they don’t appear to fall outside of control limits. Therefore, there’s no “reason why” we had a bad day… it just happens. It’s chance. It’s common cause variation.
It’s like a participant in the Red Bead game drawing 16 red beads instead of 9. It’s noise in the system.
We often don’t want to admit or recognize that there’s randomness in system performance and measures. But, it’s as real as the gravity that would make that chart fall to the floor if the magnet were removed.
This chart has the same problem of the red and green. Again, the typical, expected, average performance is red. This system is not capable of consistently meeting the goal set by management. The system is set up to “fail,” as are the people working in it.
That can’t feel good… every single day huddling around a chart that’s just as likely to be red as it is green.
I’m not suggesting changing the target. That’s just putting a happy face on things. I’m suggesting that the team should focus on improving the system and continuously improving (I hear Dr. Deming suggesting that).
I asked about the YELLOW zone. Why is there a yellow zone?
I was told that there had been too many data points in the red zone. Yeah, if that yellow zone were also red, almost every single day would be in the red.
The person said, “We’re not supposed to have yellow zones, but it was demoralizing to always be red… so we added the yellow for that chart.”
Instead of spending time on all of that, what if the time were spent actually improving the system? It’s puzzling that leaders will spend so much time on things like that… what’s the goal? What’s the color coding? But, those same leaders might say, “We don’t have time for improvement.”
Yeah, you do. Use your time differently. Lead differently.
I highly recommend the book Understanding Variation by Donald Wheeler, which explains all of this better than I do. The book is life changing. Check it out.
Thanks for reading! I’d love to hear your thoughts. Please scroll down to post a comment. Click here to receive posts via email.
Now Available – The updated, expanded, and revised 3rd Edition of Mark Graban’s Shingo Research Award-Winning Book Lean Hospitals: Improving Quality, Patient Safety, and Employee Engagement. You can buy the book today, including signed copies from the author.