Lean retreat, day 2

Previous post here

Continuing our exploration of how to reduce how many Medicare patients are readmitted after leaving the hospital, this morning we started getting our hands dirty, thinking out all the activities that take place - and can go wrong - when a patient is sent home.

Now, I have some experience with this, having been sent home seven times from my seven admissions in 2007, but I never had any idea how many things were going on behind the scenes. In particular I had no clue how often "final" paperwork (in the computer) had to get "unfinalized" and modified - and much how of a pain the computer systems make it. Plus, there are many many places where information can fall through the cracks, because this is all far less automated than I would have, naively, thought.

As always, the useful question quickly becomes, what can we do to improve this? It's no good ranting, because even a rant leaves you with this: "Okay, got it. Now: you're at where you're at. What can we do that will make any difference?"

And here's where it gets inspiring, because Lean has produced remarkable results without requiring vast new IT systems, and especially since its methods are well within the reach of ordinary people. For one thing, you don't have to figure it all out first - you think, and try things, and see what works. For another, today's eye-opening exercise was done with Post-Its, paper, and Sharpies.

More formally, "think, try, see what works" is called PDCA: Plan, Do, Check, Act [or Adjust]. It's really the scientific method applied to practical managerial problem solving. And it works. I've personally witnessed the transformation of the customer experience in this hospital's orthopedics department; when I first went there the wait was horrid (three hour visits were not unusual), and today people are in and out in 52-56 minutes - consistently.

This stuff works.

In the exercise pictured here, we mapped out four different stages of a patient being discharged, going home, and eventually returning to the clinic (doctor's office) for a follow-up visit. Apologies for the poor picture, but you can vaguely see "swim lanes" for the different people involved: the top row, with green Post-Its, is the patient; the row below is the case manager; other rows are the inpatient MD, the nurse, pharmacist, etc. All these people have specific responsibilities in discharging a patient. Each activity was written on a stickie. There were hundreds.

Then we went back and, as a group, looked at every single activity and marked it in the Lean way:

• Value Added. (Is this specific activity something the customer came here for?) Getting instructions or therapy: yes. Waiting: no.
• Not Value Added. Everything else: filling out paperwork, re-checking medication lists, paging the attending physician, you name it. Some might be necessary, but none of them is what the customer came here for. Lean strives to eliminate these.
• Not Value Added, but Required. A subset of Not Value Added.

In principle all "Not Value Added / Not Required" items should be eliminated. But a vital part of the Lean method is that you start where you're at, try things, and improve incrementally. And even if you did eliminate all such things, you'd still continue making everything leaner and leaner.

Yellow and orange stickies, known as "stormclouds," were added, showing areas where things can go awry. Clusters indicate key areas to engineering reliability into the process.


Does it work? Heck yes: more data on that in tomorrow's installment.