Gene Kim: Welcome back to the Ideal Cast. I'm your host, Gene Kim. I want to let you know that all the amazing talks we showcased at DevOps Enterprise Summit Virtual Europe this year are now available on demand in the video library. Visit itrevolution.com/videos, to watch them. They are fantastic.
You're listening to the Ideal Cast with Gene Kim, brought to you by IT Revolution. [00:00:30] Welcome to another episode of the Ideal Cast. I am recording this at the end of May, 2021 when we are currently in a race to deliver vaccines into everyone's arms in the shortest possible time. Quite literally, the success of this endeavor affects the health and economies of the entire planet. Many months ago, my mentor, Dr. Steven Spear observed that the ability for vaccine [inaudible 00:00:58] to adapt in order to win this race, [00:01:00] has been decidedly mixed. Some vaccine clinics are able to get 100% of the vaccines allocated to them successfully into people's arms. Others struggled to achieve 30%.
The result is not just wasted precious doses, which could have gone to other vaccine clinics where they would have been successfully delivered, but also causes other serious consequences. Every day that people are not vaccinated, they risk getting sick and in turn getting other sick. [00:01:30] Yet, despite it all, against all our expectations, the finish line is potentially, if not within sight, it is now certainly within our ability to contemplate. Just yesterday, the U.S. White House announced that half the U.S. Population has been vaccinated up from 0%, six months ago.
Over the last couple of months, I've gotten a glimpse into the incredible unleashing of human creativity and problem-solving that has made this possible. In conversations with various people involved [00:02:00] in the vaccination process, I heard amazing stories about how vaccination clinics were able to increase the number of vaccinations delivered by four X, six X, by doing things like changing the onsite registration flows, or even reversing the path that people took through a building to put the elevator after the vaccine administration instead of before.
But those operations are happening on a much smaller scale than what is currently happening at Oregon's largest vaccination site at the Oregon convention center. At [00:02:30] the end of March, I had the privilege of spending three hours with Trent Green, senior vice president and chief operating officer at Legacy Health. We were at the Portland Convention Center, where they are currently delivering 8,000 vaccines per day to people who need them. This is up from 2,000 per day in January. It was such an honor to see firsthand how human creativity has enabled them to ramp up vaccination capacity so dramatically. It was literally one [00:03:00] of the most uplifting things I've gotten to see in years, and I am so grateful for the work that Mr. Green and team have been doing. On April 28th, the Willamette Week published an article called Oregon's Largest Vaccination Site is a Logistical Masterpiece. We Take You Inside.
They described the site as a medical Disneyland, like the best airport in the world, comparing it to the operations of a Swiss watch where over 226,000 Oregonians have been vaccinated. [00:03:30] Contrast that to other mass vaccination sites where people described it as having a war zone or apocalypse vibe, requiring people to wait hours in line, sometimes lines to get in another line. Ever since my visit to the mass vaccination site with Mr. Green, I found myself talking about it over and over to so many people, and I had more questions I wanted to ask him. And for so many reasons, I am so glad that he agreed to be interviewed. So today, I have not only Mr. Trent Green, but also [00:04:00] my mentor, Dr. Steven Spear from MIT Sloan as well. I've actually known Mr. Green for years. I've always loved talking with him as he is so full of insights.
In fact, so many things he's mentioned to me often in passing, become areas of intense study for me. In fact, when we were discussing the phenomenon of the square in the Unicorn Project, which describes need for vast escalations to get anything done that requires integrated problem-solving from different silos, he told me a couple of things that then became the topic [00:04:30] of repeated conversations between Dr. Spear and me. I'm so delighted that I have on today, both Mr. Trent Green and Dr. Steven Spear. Mr. Green has been at Legacy Health for 12 years, which is a $2 billion integrated delivery health system based here in Portland, Oregon, which is comprised of eight hospitals, a 600 provider multi-specialty medical group, a regional laboratory and a research Institute. And he has been in the healthcare industry [00:05:00] for over two decades. So, Mr. Green, I've introduced you in my words. Can you introduce yourself in your own words and describe what you've been working on these days?
Trent Green: Well, it's an honor to be here, Gene, with you and Steven. I'm feeling in rarefied air here. I'm a chief operating officer at Legacy Health. As you mentioned, we're an eight hospital system. We own 50% of a health plan. We have well over 800 employed physicians as part of our health system. We are the leading provider of [00:05:30] healthcare in the state of Oregon.
What I am specifically charged with is improvement. So, every day, I wake up trying to figure out how do we reduce costs? How do we improve quality? Ultimately, how do we enhance value for the customer? The way that is expressed actually right now is we're on a lean journey. We are really in the early stages of implementing what we're referring to as a Legacy Health operating system. We're trying to implement a system of improvement that is organized, follows the scientific [00:06:00] method, has elements of real-time problem solving help chain, a topic that you and I have talked a lot about. And hopefully, we have an opportunity to explore a little bit today, tiered huddles leader, standard work, et cetera. We are literally just in the early stages of embarking on that journey.
Gene Kim: Between you and Steve, who will introduce himself in just a couple of moments, I'm probably the most ignorant of how healthcare systems work. Can you describe the role of the chief operating officer versus the chief medical officer, and what are the key leadership positions in [00:06:30] a healthcare organization?
Trent Green: Sure. So, I report to our chief executive officer. And our executive team consists of myself, chief operating officer. We have a chief financial officer. We have a chief medical officer. We have an [IGO 00:06:42], we refer to them as our chief integration officer. He handles a lot of our newer type ventures. And then we obviously have information technology and human resources, all part of the executive team. For my role, chief operating officer, I have responsibility for all hospital operations. I have responsibility for all pharmacy services, [00:07:00] all laboratory services. So, I handle what probably represents 70% of the total business on a day-to-day basis that runs through our facility.
And I partner a lot with our chief medical officer on the clinical aspects, quality aspects, our quality agenda. So, there's a nice diad between myself and our chief medical officer. And then we have another individual who's titled in our system, chief integration officer, who has responsibility for our medical group and a lot of our, what I would characterize as newer ventures. So, we've [00:07:30] gotten into urgent care as a for instance, outpatient surgery. We have, as I said earlier, an investment in a health plan. And so, he has responsibility for managing those newer aspects of where we're trying to take our delivery system all in pursuit of value for customers.
Gene Kim: An important business for people who need to benefit from the healthcare system. That's fantastic. So, Steve, I'm so delighted that you're on today. Could you briefly introduce yourself and maybe take a moment to describe the work that you've done in the healthcare industry? [00:08:00] And if I remember correctly, that was actually done with someone that Trent and I have talked so much about, the honorable Paul O'Neill. Famously, the CEO of Alcoa.
Dr. Steven Spea...: Yeah, Gene. Thank you. And Trent, really great to connect in this fashion. My work in general was around the question, why is it when some organizations try to pull together many specialists towards some collaborative effort towards common purpose, why are some few of them so much better at what they do than everybody else? And you start thinking about the benefit so much better, it means much more value [00:08:30] generated to meet society's needs. It means much greater return on the investment. Some have made the effort, and it means that those doing the work have much richer and more certain realization that what they did was important to somebody else. So, that's the general theme. How do you pull the pieces together so the whole is much, much greater than the sum of the parts?
In terms of work we did in healthcare, healthcare is just simply ripe for introducing this kind of thinking and doing, because it's a sector which originally started [00:09:00] with just individual doctors providing care, and everything was about that magic moment of touch where the term of art is to lay on hands, right? And what happened is, as science and technology advanced in healthcare, the number of specialists involved in bringing magic to that moment increased exponentially. What didn't advance along at the same pace was an understanding of how to do that integration of the many into one.
And so, we got invited in to do some really inspiring work. [00:09:30] Inspiring because of the people with whom we work, not because of what we actually did, but inspiring work in the Pittsburgh healthcare community. Back in the early two thousands, this was picking up off of a couple of pilots we did in the local healthcare community here in Boston with Beth Israel Deaconess hospital. And long and short, it proved out the idea that if you have a systematic way of creating systems, of integrating systems, the ability for people to express their fullest potential, it just goes up geometrically.
Gene Kim: Trent, I trust that resonates with your own experiences, [00:10:00] at least at the high level.
Trent Green: Oh, completely. We talk about this a lot. Our caregivers, clinicians in general, they diagnose problems and they implement solutions. They try a number of things, and not just our positions, but our frontline staff as well. We talk frequently about the amount of real-time problem-solving that happens in our facilities every day. The challenge is that oftentimes, we observed that our people, they solve problems, but they're often work-arounds. They don't get to really fundamentally [00:10:30] solving to root.
And so, what we're trying to now unpack as a system is how do we liberate people to actually solve problems, not just work-around so you can get through to the next thing or the next patient because that happens all the time. We stack up all these work-arounds, they're not all the same. Same problems encountered in one facility is solved this way, it's solved this way in another facility, and neither of which ultimately get to what the root cause of the problem is. And so, what we're trying to step back and do is, how do we create a mechanism [00:11:00] by which we can consistently solve problems ultimately to root and eliminate whatever harm or process breakdown it creates? So what Dr. Spear mentions, it totally resonates with our experience.
Gene Kim: I'd love to put that in a box for a moment. And so, before we talk about integrated problem-solving in the healthcare system in the large, Trent, I'd love to rewind the clock to January or whenever that was when you realized that you might be responsible for standing up a big part of the mass vaccination [00:11:30] clinic. Can you talk about what those first days of operations were like. If I remember correctly, you said you were struggling to get 2,000 doses out a day and eventually heading to 8,000. So, can you describe what that felt like in the early days?
Trent Green: It was a long story that we don't need to rehash on how we got to ultimately standing up a mass vacs site. But when we concluded that we needed to do that, you're right, we ran an operation scheduled to operate 10 hours a day, very frequently operated 12 to 13 hours a day. And [00:12:00] we were lucky if we could crank out 200 doses an hour and ultimately get 2,000 doses in people's arms during a day. And it was extraordinarily high stress. We were trying to be as efficient as possible, but we had never done something like this. What we actually put together was a collaboration of four competing health systems. So, you had that dynamic to manage, but also just the complexity of doing it. We immunize patients every day. We don't immunize [00:12:30] patients with a two-dose regimen that has a lot of care and handling on the actual pharmaceutical side.
So, that added the complexity. So, just a lot of early stress. What we found is, back to these systems and improving systems, we had all these microsystems. We had a registration microsystem. We had a vaccination microsystem. We had a pharmacy microsystem. We had a observation area microsystem. And what we didn't understand in the beginning at all [00:13:00] was how, as you made a change in one of those areas, we forgot that we needed to evaluate the implications in the other microsystems. And so, we created a lot of our own problems and it took us a while to get to where we are now. We're administering close to 8,000 doses a day. We do that in seven hours. So, over 1100 an hour is what we're up to. Now, we can do it in less than 30 minutes and that includes a 15 minutes of observation that's required.
We could never do that in the early [00:13:30] stages. So, we've made a ton of improvement. I think one principle that I learned early was, you can't improve a process until you stabilize a process. And so, one of the things that we struggled with early on is, people would have bright ideas, oh, we should do this. We should do that. We should do this. We attempted to stack way too many improvements on, and it would just make things, frankly, it made things worse either in that microsystem or in another microsystem. So, we learned, and then we started to slow walk our improvements a little bit and do a better job of testing [00:14:00] things before we actually ultimately implement.
Gene Kim: Before we talk about the mechanics of that, could you talk about maybe some of the major milestones as you went from 2,000 a day to 8,000 a day? What strikes as the breakthroughs that you would point to in that journey?
Trent Green: Yeah. So our first date of operation at the Oregon Convention Center, which is where we run this mass vac site, was January 25th. We started, first administrations would go in people's arms at nine o'clock in the morning. We would have people there at six in the morning doing training, [00:14:30] doing walkthroughs, making sure everybody was in the right chairs. Theoretically, the last appointment was at seven o'clock at night. But as I mentioned, we routinely went over in those early days. There've been so many improvements. I'm not even sure where to start. I think some key milestones for us have been, we've never wasted a dose, never wasted a single dose. Now, oftentimes, what that meant is, at the end of the day, if we overproduced in pharmacy... Again, one of the microsystems, if we overproduced in pharmacy, there'd be a lot of people [00:15:00] on the phone trying to find people that were eligible to receive the vaccine that could come to the convention center quickly.
We now have a process down where we are extremely efficient at the end of day operation, so that oftentimes, we'll have to open up another vile to finish the day, but we still don't waste and we don't have to find 50 people at the end of the day to administer the vaccine too. So, I think key milestones were a stabilization of processes. We've actually reduced hours now. So, as I said, we were doing [00:15:30] 2,000 in 10 hours, which often ended up being 12 hours. We've now moved it to a seven-hour operation. What we found is, the burnout was real. It's intense when you're in there, and recruitment of volunteers or paid staff and so on and trying to run multiple shifts just became too complex for us. So, we actually got faster when we condensed that time. So, we went from 10 hours... For a period of time, we ran 7:00 AM to 7:00 PM.
That was really taxing for people. So, now, we've moved the operation to five days a week, seven [00:16:00] hours a day when we're, operating. And for instance, my daughter's 16 years old, she became eligible two weeks ago. Last night, I took her to the Oregon Convention Center. I was such a proud father walking her through. She didn't really care about all the things that Dad had done, but we were in and out in 34 minutes. Now, that included 10 minutes of waiting because we were one of the last appointments of the day. And at the end of the day, we do do what we refer to as a pharmacy draw down where we slow down the process a little bit so we don't overproduce. And that included a 10-minute delay, [00:16:30] and that was car door to car door, 34 minutes.
Gene Kim: Awesome. It was such a heady day. And one of the things that really caught my attention was how you moved the area of operations from one of the banquet halls to the huge 116,000 foot area. Can you talk about like what led to that?
Trent Green: Yeah, so we started out actually in, as Gene mentions, in a ballroom and we thought that would be sufficient space for us. Frankly, there's nothing going on at the Oregon Convention Center, so we can have as much space as we wanted. We thought we would have sufficient space in the ballroom, [00:17:00] but this is even another learning. We didn't design that space very well. So, we had, I would call it a waterfall concept. So, you would enter the ballroom, that would be where registration was. Just behind registration is where the vaccination center was. And then the waterfall concept was, you either went left or right and that's where the observation would occur. Well, the problem with having two observation centers is, we had to double the number of staff. We had to have double the number of people that are sitting in there. We had to have two medical tents for people.
And so, we built in some inefficiencies in the beginning. And then, we also [00:17:30] realized we just didn't have enough space to process. We could only end up doing about 500 per hour was our max in that space. It just wasn't enough. So, we moved to the exhibit halls where we had ample space. And then we modified our approach where we created six pods, and the pod strategy was really key because it just enabled people to move that much more rapidly and us to process that many more people efficiently.
When we first moved to the space, we actually had a consult from Starbucks. So, Starbucks had been very [00:18:00] involved in the state of Washington and their mass vacs operations. And so, I met this gentlemen, Josh, at the convention center in February, the first day we had moved into the new space and he said, "Well, do you mind if I take some time studies and I'd be happy to give you some suggestions for improvement?" I said, "No, that'd be great." He spent, I don't know, half a day there, observed all of our systems. Came back, sat down with me and said, "I got to be honest, I don't really have any suggestions for you. You guys are doing such a good job of processing. I want to go [00:18:30] back and study this a little bit more." He said, "You might be able to open one more pod, but beyond that, keep doing what you're doing."
Gene Kim: That's great. In fact, I'll put a link to the FEMA video that was put together that featured the Starbucks person and the back of your head too, Trent.
Trent Green: Yeah, exactly, exactly.
Gene Kim: Before we get to Steve, when you look back at this, what are you most proud of getting to the vaccination clinic this far?
Trent Green: Well, to be a healthcare, any role and healthcare the last 14 months, it's been hard. February 28th is when our first case [00:19:00] was identified in the state of Oregon, and I still remember that day. It was a Friday night. I was supposed to be at a high school girl's basketball game, and I spent the entire time out in the parking lot on the telephone trying to figure out what to do. It wasn't even a case in our hospital. We'd been monitoring COVID, and the last year has been hard. Emotionally, it's been hard for people management of protective equipment. I have to say, though, this vaccination effort has been a bright spot for me personally. I get so much joy. One of the things we lose, I think we've lost is recognition of how important [00:19:30] smiles are with masks. It's really hard to tell when somebody's smiling. But I can tell you, when you go to the Oregon Convention Center and you walk around, you see everybody in masks, you can tell they're smiling. You can see it in their eyes.
And that's just been really gratifying for me. We worked with the governor to actually move teachers ahead of seniors. That was a controversial move, but Oregon had been very slow in restarting schools, and we worked with superintendents of schools who were phenomenal partners to help us test our systems. [00:20:00] Again, we didn't know how to do this and they were organized. We were organized. We figured out a way to get them scheduled efficiently. And so, Gene, I'm going to remember this for the rest of my life, the role that I played in this vaccination effort. And it's just been really personally gratifying for me and for our people. I can tell, people's spirits at work have even been lifted. I'm now vaccinated and they're vaccinated. And there's still protection, but it's been really hard in healthcare and this has been a bright spot in the last 14 months.
Gene Kim: I just read that amazing [00:20:30] article in the Willamette Week describing the vaccination site as a magical Disneyland, like the best airport in the world, like an f-ing Swiss watch. That's got to feel pretty good.
Trent Green: Yeah. It's very gratifying.
Gene Kim: Steve, you had told me something some months ago that, literally my jaw dropped and that also became the focus of lots of effort and other studies. You made the observation that some hospitals, as they're delivering vaccines, they can get a hundred percent of vaccines in people arms, where some are struggling [00:21:00] to get 30%. And then you also made the observation that some school systems, once the decision is made to return to in-person learning, some can reopen on a dime, maybe weeks, others take months or quarters. And you made the claim that both of those are probably a very good proxy for an organization's ability to adapt, learn, act upon learnings, re-engineer all the relevant processes and really unleashing the creativity of the entire workforce, especially on the front. So, to what extent [00:21:30] does the story that Trent told, does that affirm your hypothesis on a scale of one to 10? One is not at all, 10 is exactly in line with what you expected.
Dr. Steven Spea...: Yeah. I'm going with a spinal tap, this is an 11. No, no. I just want to, first of all, say thank you to Trent and his colleagues for all the hard work they've done both in the gloomy times and now to get us to the end of the tunnel. Just to put things in perspective, people were talking about having to be a masked society well into 2022 and even 2023. And the fact [00:22:00] that by mid 2021, we'll have some return to normal is fantastic. And again, to Trent and his colleagues that during the dark periods, they took on real personal risk to help usher the rest of us through this terrible event. What I'd like to encourage anyone listening to this is to listen to Trent's account with great appreciation. It's well-deserved, but also to recognize that in his account, there's some general lessons which are applicable to them too.
And I'll just pick up on a couple. One was this idea of, the problem was [00:22:30] we didn't run a system, we were running a series of microsystems. That is common, and Gene and I have spent a lot of time talking about why we end up with these very complex organizations which looked like these kind of loose confederations, loose collections of microsystems that don't really come together. The reality is, that's true and where you get systems of systems to parlay one of our favorite books, team of teams, is by actually having someone draw the line through the systems to figure out the sequence, the dependencies by which value gets created. [00:23:00] And when you start seeing the system of systems, then you can start doing all the system level [inaudible 00:23:06] and the system level improvement to make things better. So, that was one point. And again, why I encourage people to listen beyond the well-deserved appreciation is that anyone listening to this podcast, they are probably working within a microsystem, which is part, not part of a microsystem and a larger system, it is a microsystem next to a lot of other microsystems.
So that's one. The second thing that Trent said, of the many things he said, but the second thing that comes to [00:23:30] mind is this idea that first we need a standard before we start to improve. And that's so key, because until we actually make a sound declaration of what we think is going to work, we can't find out we're wrong. Until we create a standard, create a declaration, set up the hypothesis for an experiment, it's just air confetti, it's just nothing.
And even if the first standard you create, and inevitably, something as complex and novel as a mass vaccination, the first standard you create is going to be deficient in some way, at [00:24:00] least you could see the deficiencies in the standard and then quickly close the loop and find out what's wrong with the standard and modify, modify, modify so that you get to a better standard, whether it's the realization that waiting to the left and to the right complicates things out. Fine, one waiting place. From one location to another location, fine. Again, that wasn't in the first standard, but it was in version three and four and 12 and 24 and et cetera, et cetera. Similarly, like the slowdown in the pharmacy, which is, if we don't stop the pharmacy filling syringes with some lead time... And then you probably, once you [00:24:30] had that realization, I'm willing to bet you had six iterations on just how much time you needed, right?
Trent Green: Well, probably 15.
Gene Kim: Here at IT Revolution, we achieved something pretty amazing just before the DevOps Enterprise Summit Virtual Europe Conference. We added every one of our past conference talks dating all the way back to 2014 into our video library. It is now the best place to [00:25:00] watch these videos and it is getting better and better all the time.
PART 1 OF 4 ENDS [00:25:04]
Gene Kim: ... watch these videos, and it is getting better and better all the time. It's super easy and fast to search for talks, either by descriptions, speakers, companies, or topics. Slides are available for all the talks, and for all of the plenary talks this time around, we have posted transcripts for them too. In other exciting news, we have launched the DevOps Enterprise Journal, Spring 2021, which features white papers from some of the most popular 2020 DevOps Enterprise Summit presentations. Download it at itrevolution. [00:25:30] com/journal.
Dr. Steven Spea...: One thing, and then I'll stop gushing praise on this is, like I said, there were three. One was, we live inside microsystems rather than a system of systems and this... you need a standard to make an improvement. But the other is, where Trent ended, which was, you can see the smiles behind the masks. February 28th, 2020, no one was smiling. [00:26:00] But when this kicked off with your mass vaccination, probably, January 2021, still, no one was smiling, even though you were giving a vaccinations, because it was just so overburdensome. It was so frustrating, it was so difficult to actually do the right thing.
And now here we are in mid April and people are doing quote unquote, the same work, but they're doing it with gracefulness. They're doing it with appreciation. And the thing I just want to offer, why is anyone ever asked to do work, where they can't smile-
Gene Kim: Yeah.
Dr. Steven Spea...: ... with the sense that they've done something that [00:26:30] someone else will appreciate? I almost have to get to the point that, it's a choice that we make and that people responsible or other people make, that we make a choice as to whether we're going to be energetic and open-minded, and inquisitive like Trent and his colleagues. And try to get systems to evolve and adapt and adjust so that the people in this system want to smile because they know that they're doing something appreciated. And it's an alternative that's chosen, if we don't do that. Anyway, this has been a fantastic... I'm going to stop talking because I want to listen.
Gene Kim: [00:27:00] Well, in fact, Trent, I think, as Steve was talking, it kind of revealed why I was so excited to get the three of us together. And I think part of it is just the notion of theory building versus theory testing.
Trent Green: Yeah.
Gene Kim: I think to be able to... I love the principle of parsimoniousness in science, [inaudible 00:27:15] the goal is to create the fewest number of principles, explain the most amount of observable phenomena, come up with some hypotheses, and be able to test them. Right? And then gain some surprising insights. So just even in this conversation so far, I'm hearing a lot of that. So I have a theory for you, or like a hypothesis, and one of [00:27:30] them... Sorry, I was reviewing the pictures I took at the vaccination clinic at the convention center. And clearly, it was a lot of good ideas that led to this massive expansion in capacity. I'm assuming that a lot of those good ideas didn't come from you. So, can you talk about the process of which good ideas were surfaced and maybe in two paths, one where they all pile up and make a big mess. Right? Where you can't isolate whether it was actually better or worse and why-
Trent Green: Yeah, yeah.
Gene Kim: ... versus the more systematic sequencing. Yeah, your role in that.
Trent Green: Yeah, sure. We [00:28:00] tried to adhere to this principle that the people closest to the work, they're the ones that know the work and can improve the work. And so, we did a lot of listening to them. What barriers were they encountering? What suggestions did they have? So we did a lot of listening.
I want to comment on the standards. One thing we realized, very early on, is that we actually needed to document and adjust, on nearly a daily basis, standard work for people. What would happen is, we would have a new vaccination lead for the day. Well, [00:28:30] the vaccination lead who was here yesterday, they wanted to do it this way. And we said, "No. Here's the standard that you have to apply for every vaccinator so... And we're not having a lead make an individual adjustment." They came from different healthcare systems.
One thing that we did pretty early on, it was really smart, is, every morning when we would meet for our daily safety huddle, we had a plan for the day. This is how many appointments we have, these are our hours of operation, this is, et cetera, et cetera, et cetera. And we went through it for every area. And then on the back, we had a QR code, and the QR code was for [00:29:00] suggestions. What we wanted to do is actually have people capture in real time, in whatever area, what suggestions they had for improvement. And we made a principal early on that you could not make, you could not implement any improvements at that point in time. You had to document what the issues were. So you had to do a little investigation, capture, what was the experience, what happened, et cetera.
And so, we'd really tried to push it to the frontline. And we spent a lot of time [00:29:30] on... This is another thing I'm learning is, and I'm guilty of this, it's often hard for me, as a senior executive in a healthcare system, I think I fairly often place the burden on myself, that I have to know everything and have a solution for everything. And sometimes it's hard for me to even step back and have a beginner's mind on things.
So what we really tried to push people on is, don't give us a solution, help us really understand current state. So, do a better job for us documenting what current state is, and then let's collaboratively [00:30:00] figure out what would... let's go back to what we think the ideal future state is. And then what are the gaps and how do we solve? So don't come with solutions, come with a good understanding of current state.
Gene Kim: Even in the idea box, right, there's actually a little bit of a structure to be able to say... not just the idea, but, "Describe what you think will this improvement will lead to." Right?
Trent Green: Yeah, exactly. And we did do some testing. We actually set up... Well, one of the big challenges that we had, we first wanted to target waiting. We wanted the process to run as efficiently as possible. We threw a lot [00:30:30] of people at the problem early, as opposed to trying to solve with improvements in process. But one thing, Gene, we've not been able to solve, and you can attest to this since you've been, we have a lot of motion. There's a lot of movement. There's a lot of walking, and we had a hypothesis early on. Why don't we create a pod model and let's move the vaccinator and the registrar around the patients, as opposed to have the patients move?
So the concept was, that would have, call it, let's... for the sake of discussion, we would have [00:31:00] eight chairs in a kind of circle or semicircle. And the registrar and vaccinator would move from patient to patient. We had a model where you'd have two registrars to one vaccinator, registration takes longer than vaccination. And they would move around and the patient would sit there. They would get registration, vaccination, and their 15 minutes of observation in that chair. We tried to test that, and that was too much for the team. That became very stressful for the vaccinator [00:31:30] actually, because they were trying to catch up the entire time.
So anyway, just an example of, that was a suggestion that was made, that maybe this would be a way that we can attack this excess movement problem, waste, that we have. And that came up through discussion. We kind of batted around. And one afternoon, actually, for three or four days, we ran a model where, on the side of our vaccination center, we actually tested that. And we tested how quick it was and was that going to actually... And we ultimately determined that was [00:32:00] too much too fast, if you will. And we ultimately, didn't think that we were going to be able to process as many people in using that approach.
Gene Kim: And so, can you talk a little bit about, what was your specific role?
Trent Green: My biggest role and the role of my colleagues, I don't want to leave out the chief operating officers of the other health systems who were partners with me in this, we set out, kind of, the overarching principles. We're going to use a single electronic health record. Everything is going to be scheduled. [00:32:30] These are the hours of operation. We set those out, and then, again, we let the people do the work. Occasionally, there were things where we needed to break ties, or we need a larger decision that needed to be made. But the actual true guts of kind of the operation, we let the people that were closest to the work do it.
I will also say, I need to mention, that we had two performance excellence engineers, I would call them, who worked with us. And they helped study a lot of the ideas. They would observe, back to Steve's [00:33:00] point about the kind of systems is, they saw the whole thing. They were all... All they cared about was overall cycle time. They didn't care about what the cycle time was for registration, they wanted the overall process of vaccinating individuals. And they helped us see the whole. And they were really, really instrumental in advising me on improvements and kind of standards that I needed to enforce or principles that I needed to enforce.
Gene Kim: So, Randy was one of these?
Trent Green: Randy and Eileen. Yep.
Gene Kim: Fantastic. And I also have to imagine... So I'm thinking [00:33:30] about, so the way I remember it, whether it was pharmacy or vaccination or registration, they all had a department lead that would rotate in and out.
Trent Green: Correct.
Gene Kim: So I would imagine that those leads would be responsible for their area, but certain things, they are also embedded in a larger system. And so, things that would affect the groups around them would require some person in a higher level authority that could see and sequence or prioritize or something, to what extent is that true? And that really became in your domain.
Trent Green: [00:34:00] Yeah, that is absolutely true. And that helped create our tiered huddling structure, a morning huddle with all the leads for the day. And we would talk about what was happening. If there was a change to a standard, we'd communicate the change to the standard. About two hours into operation, on a daily basis, we have a safety huddle. Has anything come up that we need to know about, that we're concerned about? And then in the afternoon, about two hours before we close, we did a plan for the day, a look ahead. What are we learning today? Is there anything that we're learning today that we need to think about and adjust for tomorrow? And then [00:34:30] finally, at the end, end of the day, we would bring a much larger group of people who also incorporated scheduling.
So, what I just referred to, Gene, was kind of the daily operation on the ground. But then there are other microsystems that are involved here, or other players, technology, scheduling, et cetera, that we needed to bring in and have a larger conversation about any improvements or challenges that we might have experienced. So at the end of the day, we had what we referred to as, the [00:35:00] highest level tier. We referred to it as our CXO tier. So you'd have the chief operating officers and the chief medical officers of all the systems who would hear what happened that day and be responsible for, if there was any major process improvement or change or resource that was needed that was elevated to our attention to help solve.
Gene Kim: Got it. And just to confirm my understanding, so is that you have the sort of the departments, you have the leads-
Trent Green: Yep.
Gene Kim: ... you have the, I'm kind of guessing, the incident commander-
Trent Green: Yep.
Gene Kim: ... and then the CXOs above that.
Trent Green: Exactly.
Gene Kim: I'm guessing that [00:35:30] it was really that incident commander who played an instrumental role in changing areas of system outside and for anything that involved more than one component, one area.
Trent Green: Yeah. They needed to be able to see the whole and understand the impact on the whole. And when I sat in that role, I often would consult with these performance excellence engineers, just get their insight on suggestions that were made. I want to stress, we tried not to make, on a daily basis, changes. I think we realized early on, we made way too many changes. [00:36:00] We did not have a stable system, so we couldn't improve the stable system. It was like the confetti that was mentioned earlier. Anything might improve, might help, or it might... it may not help, or it might help in this area, but not in another area. And so, we really tried to take good care in overall improvements.
Gene Kim: This is so great. In fact, I mean maybe as a segue to the other, sort of like, big, aha moment. And then, Steve, I'd love any of your reflections. One of the most startling and amazing things that you had showed, both me and Dr. Chris [Drear 00:36:27] , was just sort of the planning that went [00:36:30] around this. I mean, you had talked about if you kind of did the hard costs or soft costs. Right? One, the expense to run this operation, right? If you include the FEMA staff and National Guard, right? You're looking at, I think, it was three quarters of a million dollars a day-
Trent Green: Yep, exactly.
Gene Kim: ... just for [inaudible 00:36:44] purposes. And the thing that really, I just was not prepared to hear, but it was probably one of the most hopeful things I've ever heard was... Sorry for if I mangle this, but you said, "With the advent of Moderna and the potentially, and hopefully, the Johnson and Johnson vaccines, there were scenarios where the center could achieve [00:37:00] its desired goals in, say, 11 weeks, that with the vaccines that were easier to store and administer, the burden of this extremely expensive mass vaccination site, might be coming to an end." So whether it's 11 weeks or 17 weeks, I guess what blew me away was that strategic level thinking of like, "How long does this need to be around?" So if I understand correctly, that was kind of the thinking that was obviously, in your domain. Right? [inaudible 00:37:23] was like, "How long does need to be around? When can we release FEMA, National Guard staff?
Trent Green: Oh, a hundred percent. Back to [00:37:30] the last 14 months being incredibly difficult, we didn't have in our legacy health strategic plan that we were going to create a mass vaccine site five years ago. We didn't even have this in our plans, frankly, six months ago. As a result, we're having to beg, borrow, and steal. We're paying overtime to nurses to work in the site. And so, we're adding burden to our organizations. And yet, the goal is, let's get as many people vaccinated as quickly as possible.
And we aligned on that [00:38:00] really quickly. And I met with Pat Allen, the director of the Oregon Health Authority, very early on and was just doing math for him. "Look, this is a two dose regimen. This is how many people we have in the state of Oregon. If you want to assume, for planning purposes, that herd immunity is, conferred at 70%, this is how many doses, in the state of Oregon, we have to administer a day. In the metro area, we have to do this many." And so we used that as an organizing principle for, how do we get this going? And how do we turn this off by [00:38:30] no later than the fall?
So very early on, we outlined kind of those high level objectives or vision for the center. It seems like we are going to realize that much earlier than we had anticipated, which is great. I actually, as gratifying as this has been, it will be equally gratifying for me to turn this off. While it has been gratifying, it has also been grueling. And it's been hard on our people, just in terms of additional work to do. We're having a [00:39:00] hard enough time managing patients in our hospital, let alone, now we've got to add resources to support vaccination efforts. So, it has been great. I am also looking forward to turning it off. Having said that, we're trying to catalog all of our learnings here because we may have to turn this back on, if these vaccines required boosters, we need to be prepared to quickly mobilize again, if we have to.
Gene Kim: Gene here. Holy cow, I didn't want to break in and interrupt Mr. Green's amazing story, but I think now it would be a pretty [00:39:30] good time to share some reactions and additional context. So, number one, I mentioned what an honor it was to get a glimpse behind the operations at the vaccination center with Mr. Green. And when he described how difficult things have been for healthcare systems during the COVID pandemic, I can't even really imagine what it must be like.
For a hint, I'd refer you to two interviews that NPR Planet Money Indicator did with Dr. Patrick Cawley, who is CEO over the Medical University [00:40:00] of South Carolina Health System in Charleston, which has over 10,000 employees, including 1,200 doctors. The first interview was in April 2020, and later in December, 2020. He described what life has been like for hospital administrators. He talked about the preparation for potential surges due to COVID, and the declines and surgeries, procedures, ambulatory visits, and hospitalizations, which resulted in $30 million in losses that month.
[00:40:30] In that April interview, he described how they temporarily had to lay off 900 care team members. And actually, they had performed that reduction in force the day before the interview, which was just heartbreaking to hear. In that interview, Miss [inaudible 00:40:46] asked about how strange it was that they're having to do layoffs, despite that so many people are showing up in the emergency room. Dr. Cawley responded about how he's had that same conversation a dozen times a day with people outside of the hospitals and health [00:41:00] system, because most of the revenue comes from surgeries or procedures, and those dropped 60 to 70% in the prior three weeks. And then he goes on to talk about the, almost wartime-like preparations they're having to go through, preparing for the expected COVID surge in four to six weeks, where they had set up 500 more beds in campus gyms and from reconfiguring internal space inside their hospitals.
But in my mind, the more remarkable [00:41:30] interview was in December where he described the first vaccinations is being administered to frontline healthcare workers. He said, "I've seen multiple people getting vaccinated here." He described it as a surreal experience. He said it was the first time he's ever seen people cry while being vaccinated.
I'll quote from the interview, "I've never seen that before. For many people, it was a surreal experience and there was an emotion by many folks who received it. There's just been so much on people during the last nine [00:42:00] months. Think about what we've all been through, how we've completely upended our lives and our children's lives, and there's just so much weighing on people. So, you know, it didn't surprise me at all to see that happen. There's just this full on elation as well, it's great to see all those different sorts of emotions mixing."
And by the way, I have goosebumps as I'm reading this. He continues, "And finally, we're at the end of one journey and the beginning of another. So here we are giving vaccines [00:42:30] for this thing that has wreaked so much havoc. And this is what will give us our lives back, this is what will allow us to be free and move around and get back in schools. And what scientists have done here, in such a short period of time, is nothing less than putting a man on the moon or some of those other great accomplishments."
Holy cow, these were two great interviews. I will put links to the transcripts for those two interviews in the show notes. So that gets us to number two. Mr. Green mentioned how, [00:43:00] quote, "I'm going to remember my role in this vaccination effort for the rest of my life. Well, I don't have much more to add than that, I just want to underscore how critical of a mission this is to vaccinate as many people as we can, as quickly as we can, so that we can return to societal normalcy."
What Dr. Cawley described was almost passing the baton to the next stage in the vaccination rollout. And my feelings during this interview and editing it is one steeped with a feeling of gratitude [00:43:30] and awe, and even elation learning about how Mr. Green and team went from delivering 2000 vaccines per day to 8,000 vaccines per day. It's also mixed with a sense of grief because vaccines aren't yet available around the world. The most notable right now is India.
Last week, I had numerous conversations with friends who described how they had family and friends die in the last several weeks. As I said in the beginning, this mission requires that we get everyone on the planet vaccinated as soon as [00:44:00] possible. And hopefully all the miracles we've talked about today will not only make vaccines more widely available, but the lessons learned in these vaccination role outs will help us all achieve this goal faster.
Okay. Number three, Dr. Cawley described the availability of COVID vaccines as being the equivalent to landing a man on the moon. That's a super, interesting observation. So Dr. Spear and I talked about how amazing it is that we have not just one vaccine approved for emergency use, but five vaccines approved [00:44:30] for emergency use. And so at first glance, this seems nothing short of miraculous. After all, the typical vaccine takes five to 10 years to develop. In early 2019, most estimates were suggesting that it would take two to three years for a vaccine to be delivered.
But apparently, thanks to Project Warp Speed and other incredible heroics, somehow it seems to have unlocked a massive level of focus and investment that enabled vaccines to be developed [00:45:00] in a record time, but also mass manufactured and distributed at a scale we may never have seen before. But I will suggest that this is actually doubly miraculous because of something Dr. Spear showed me last month. In the pharmaceutical space, it's called Eroom's law. So, you'll note that's the reverse of Moore's law.
So of course, Moore's law is observing that CPU performance is increasing exponentially. Now for nearly, say, five decades, where as CPU [00:45:30] performance has doubled every three to five years. Eroom's law is showing that, in pharmaceuticals, for every $1 billion in R and D spend, the number of therapeutics making it to market is dropping exponentially. So in 1950, $1 billion of R and D would yield a hundred therapeutics that made it to market. By 2020 that has dropped to less than one. So the curve is a mirror image of Moore's law, dropping logarithmically, [00:46:00] as time progresses.
I am reading from the Wikipedia entry for Eroom's law. It is the observation that drug discovery is becoming slower and more expensive over time, despite improvements in technology, a trend first observed in the 1980s. The cost of developing a new drug roughly doubles every nine years inflation adjusted. This is in contrast to the exponential advancements of other forms of technology over time and was deliberately spelled as Moore's [00:46:30] law spelled backwards.
So Steve and I have discussed that maybe Eroom's law is what naturally happens when you increase the number of functional specialties that need to work together. Maybe it is a reason why team of teams had so much trouble in the before state dismantling the terrorist networks. Maybe it describes why pharmaceutical development is getting more and more expensive and taking more and more time. Maybe it describes what happened in software development and delivery. [00:47:00] But somehow, when survival depends upon it, like in team of teams or what birthed DevOps, or we suggest might have happened that enabled five COVID vaccines to be created, that have been approved for emergency use. Or in the mass vaccination site that Trent oversaw, somehow organizations can break away from this dominant architecture that suffers from Eroom's law and create genuinely new ways of working. Steve and I gave a presentation on this at DevOps Enterprise [00:47:30] Summit last week, and I'll put a link to that in the show notes.
Number four, I want to describe how challenging it was in this interview to cover all the ground I wanted to cover, even in the two hours that we had. In just a little bit, in this interview, I'm going to switch focus to how the lessons learned during the COVID vaccination rollout process could inform how to improve the overall health care system. But I wanted to share some of my firsthand observations from those amazing three hours [00:48:00] at the Portland Convention Center in March.
I was there with Dr. Chris Drear. He is an emergency physician who I met several years ago. And in June he will be starting as the Chief Medical Officer at Columbia Memorial. He has a long background of using theory of constraints to improve flow in a healthcare setting and being a medical practitioner, it was so fun and so useful to be able to compare notes with him and ask him questions about what we saw together at the convention center.
[00:48:30] I'll share some pictures I took that day. I would share more but many I can't because I didn't obtain permission from all the people being vaccinated, which are protected under HIPAA regulations. So the day started out at 11:00 AM, well before the noon opening that Mr. Green alluded to. Walking into the convention center, you could see, already, a line of people starting to queue up, well before their appointment time. We met Trent outside of the volunteer area. He was in a reflective vest, as [00:49:00] were most of the staff, but his had the title of, Incident Commander, emblazoned on it. Just by that you may have already gleaned that they were using the FEMA Incident Command System, which so many of us in the DevOps community have some familiarity with.
And if you don't, check out the talk that Brent Chapman gave at the DevOps Enterprise Summit in Las Vegas in 2018, describing his experiences as an SRE leader at Google, as well as at Apple and Netflix. He's not only a technology leader, [00:49:30] but he's also spent time as an air search and rescue pilot, and as an incident commander for major festivals, such as Burning Man.
So one of the first things that Trent showed us was the ballroom that he mentioned in this interview, the first site where they administered vaccines. We retrace the steps that people would take as they parked their cars, go up the elevators, then queue up to get vaccinated. He described the many changes that they made to improve flow until they realized that the amount of space [00:50:00] was going to cap them at around 500 vaccinations per hour.
PART 2 OF 4 ENDS [00:50:04]
Gene Kim: [inaudible 00:50:00] was going to cap them at around 500 vaccinations per hour. And then he took us to this amazing conference room that overlooked the nearly 116,000 square feet of exhibit hall space. It was like being on the catwalk overlooking the plant floor in the Phoenix project. I have a panoramic picture where you could actually see that vast space divided up into the initial queuing area, onsite registration, a special area where people would go [00:50:30] if they had allergies so they could be seen by doctors and other exceptions. They would then queue into I think six vaccination pods. Behind curtains, you could sort of see the pharmacy preparation area where there were scores of people preparing the Pfizer vaccine for administration. I'm looking at the CDC site about the Pfizer vaccine. Apparently there are six doses per vial. You have to thaw them for two hours. If you click into how to thaw, [00:51:00] prepare and administer the Pfizer BioEnTech vaccine, you see a list of 14 steps.
The steps include how you need to mix it in the vials, how you need to invert it 10 times, mix it in the syringe. You have to use it within six hours. And through this cascade of work, the end product from that work center are trays with syringes, each with vaccine doses in them, ready to be jabbed into people's arms. So that gets to the next area of flow, which is where I suppose [00:51:30] the action really happens: the jabbing. So there were about six or so pods in each. There might've been 10 or so jabbers, they seem to be grouped by what healthcare system they came from. One was primarily National Guard, who I understand were often medics. And so once you get jabbed, you then flow into an observation area where they would sit and be observed for any reactions for 15 minutes. And it's during that time, when they would be approached to schedule the [00:52:00] second appointments.
And as Trent said, people could get in and out within 35 minutes. What is so interesting to me is that in like Dr. Spear's description of the [inaudible 00:52:12] production system, the number of improvements are so vast that they're difficult to list off one by one from memory. So what we observed that day was this incredible process that was just flowing, which was the in-state derived from hundreds, if not thousands, of little improvements made day [00:52:30] after day or week after week for three and a half months. Which gets us to number five. I love that Mr. Green mentioned that there was a suggestion box QR code that everyone could use to describe the current condition, their proposal or suggestion, along with what the expected improvement would be.
I love that the structure helped facilitate thinking through experiments. Like a well-formed feature request or a bug report form where the people working on solving [00:53:00] the problem can get the information they need without bogging down the person having to fill out the form. Which gets us to number six. Holy cow, there's so many great things to reflect upon in this interview. But I love how Trent talked about the slower cognitive problem solving that leaders were tasked with, such as setting the system level goals, going through the math of how many people do we need to vaccinate to get to 70% herd immunity. [00:53:30] How many can we possibly deliver per day? How many people do we need to deliver those vaccines? What would be the shift of operations? Where are we going to source those people? At one point, Trent was showing Dr. [inaudible 00:53:43] and I the results board. So around the time we visited, they were doing about 8,000 vaccines per day and all but 10 of those were Pfizer vaccines.
And so apparently there's this huge logistical burden for the 10 Moderna vaccines that were being delivered. [00:54:00] And so there were discussions going on about how to phase out the Moderna delivery and move that to other vaccination sites. Especially since it's apparently an easier vaccine to handle and administer, which would allow them to focus on just specializing on the Pfizer vaccine. All of those things, I think, represent those skills that the best leaders bring to bear when they think about solving big problems. Setting those system level goals, designing the [00:54:30] organization, assigning the roles, responsibilities, and the relationships between them, and of course, maybe most importantly, assessing and helping judge the performance of the system and help enable the next improvement. Okay, let's go back to the interview where I ask Mr. Green about the fast versus slow integrated problem solving styles that he's had to employ early, versus late in the vaccination journey.
Steve, before we ask you for your impressions, Trent, one of the things that Steve and I have [00:55:00] been trying to conceptualize is sort of the different modes that leadership operates in. And just like in the book, Thinking Fast and Slow, we're seeing this really kind of two modes. One is the slow cognitive problem solving that is, I think, often attributed to planning. So when you talk about these goals right, end the vast vaccination center by no later than fall, right, as you think about what's the ideal distribution mixes of Pfizer versus Moderna versus the fast [00:55:30] tempo of operations.
And I think kind of our observation is that if you look at whether it's sort of the before state and team of teams, there's a pattern where leadership overreaches into operations, right? So using kind of the much [inaudible 00:55:44] mechanisms of going up and down the organization is slower. It doesn't have as much high fidelity and often ends up leading to not great outcomes. I guess the question is, did you find yourself operating more in this kind of slower state of the slower activities of improvement and planning [00:56:00] and less so in this kind of overreaching into operations since does that resonate with you?
Trent Green: Yeah, it completely resonates that. You know, I think as leaders, we slow things down. If we have to be involved in every step outside of kind of the high level direction that we want things to go, we just slow things down. We often have our hands in too many things to have enough continuity to really affect the day-to-day operations. So end goal is we'd like to shut this down by no later than August 31st, having administered [00:56:30] all the vaccinations that we had targeted. So actually I should qualify that we actually did start with a number target. This is how many we think we need to do. The other principles that I think we help that it set context for the operation were things like we're going to go scheduled only. This is too complex of an operation to do, just an open thing. I think you've seen that across the country where people just had open access. And we made it a principal decision early on that given the complexities, the two dose regimen and so on, that only going to go scheduled only. That was [00:57:00] the only way we were going to operate this.
And so I think we outlined some higher level principles. And then, Gene to your point, we unleashed the creativity of the teams who were involved in these things, the technology team with the EMR. Okay, how are we going to do this? The teams that are used to scheduling in our healthcare operations, okay, how are we going to schedule these patients? Are they going to go in and schedule their own appointments? We unleashed all that creativity with the overarching principles that we were trying to achieve. [00:57:30] And only when they encountered a big roadblock or needed a big decision on something, did they come to us. And we tried to not overreach into operations. We really did try and allow the experts. And this was actually made more complicated I think in this instance, because as I mentioned, we had four health systems that came together.
The other principle we said to them is, look, we are not going to tolerate infighting between the health systems. We do compete on a day-to-day basis. This is in the best interest of our community. There is no competition. The only competition here is [00:58:00] the best idea wins. So best solution, best process, best, whatever. We're not taking the Kaiser process for this and the Legacy process or arguing about who better at this, that, or the other thing. And so I think with our overarching principles and then unleashing of the teams to come up with ideas on, okay, how would we accomplish this goal? That's how we were able to do what we've been able to do.
Gene Kim: Maybe we can segue from that amazing story about the vaccination process. And by the way, on behalf of a grateful populace with Steve [00:58:30] and I we're expressing our appreciation for everything that you've done, let's go to the broader healthcare system. And so Trent, I remember you and I were hanging out at the bar at [inaudible 00:58:38] December 2019. And you mentioned the story that I literally have been pondering ever since. And I'm not sure if this was entirely joking, but you described that you were standing up the safety meeting every day and that you had to be there because, if I understood correctly, you were the only sanctioned interface between say nursing and transport or pharmacy. And in other words, for integrated problem solving to [00:59:00] happen between those areas, you had to get involved. So it was like up eight down eight. And so Steve and I talked about this for months because it's so much resonated with his own experiences in healthcare systems. So Trent, can you talk a little bit more about the story? Did I take that too literally? And even if it's an exaggeration, why do you think it's so important?
Trent Green: Yeah, you probably took it a little bit literally, but not much. I mean, the reality is oftentimes if there's an IT issue in one of our... It just so happens [00:59:30] that our chief information officer's office is right next to mine. And it wouldn't matter whether it was right next to mine or three doors down. What often happens is an operational issue comes all the way up to me. I have to go knock on the door next door or text my colleague and then it goes straight down in his chain. I think that is not uncommon in our organizations. And it's certainly not uncommon historically at Legacy. I think I've been involved with other healthcare organizations, I think it's not uncommon at all. We are actually trying to break that, as [01:00:00] I mentioned earlier, with we're trying to implement an operating system that has a very clear help chain process that is totally separated from a tiered huddling process.
One thing I am learning is that I think in a lot of organizations, people have gotten excited about safety huddles and so on. It's often a crutch in my view. A tiered huddle is often a crutch for an ineffective help chain. Meaning you shouldn't have to batch your problems and come to the daily tiered huddle in order [01:00:30] to let somebody in supply chain know that you had a problem yesterday with these gloves, or these masks, or whatever. There should be a process by which you can activate a help chain real time if you encounter a problem. And there is some of that, but I would say it's largely immature in complex organizations like ours that have multiple sites. And so we are spending some time right now trying to understand and design a help chain process that enables us to actually [01:01:00] have people who are in the work elevate their help chain.
And only if it is a really significant problem that no one else has been able to solve should it ever land on my desk. I shouldn't have to be involved in the day-to-day help chain process for problems that happen in our organizations. I might hear about them the day after at a tiered huddle, which would be a good process. I could learn about what the problem solving effort that people undertook. So I might get updated on it, but I wouldn't necessarily be involved in the help chain process. And so we're actually [01:01:30] trying to design that now as a system, because that's not how we have historically operated. People go straight up to the top, they cut over to their colleague and then it goes down the other. That was the story that I was describing to you and that still largely exists today. But we are trying as an organization. We are trying to break that because we think it actually slows the overall problem solving process. And ultimately, oftentimes, ends up not having us solve a problem at all. We just put more band-aids on it.
Gene Kim: We just exercised how quickly can I get to Trent?
Trent Green: [01:02:00] Yeah, exactly.
Gene Kim: So one of the things that you shared with me, after three amazing hours at the vaccination clinic, was this intuition that there were lessons being learned in the operations of the vast vaccination clinic that were relevant to this. And you had mentioned things like cardiac surgeries or hip replacements or something. Can you say a little bit more about that? And then Steve, I'd love your reaction because it's some of this seems eerily, uncannily similar to conversations we've had.
Trent Green: Yeah, I guess what I was struck with is in this vaccination process, we've demonstrated how [01:02:30] we've been able to solve problems and improve processes so rapidly. Look, we were only able to do 200 vaccines an hour, and actually that's probably even overstated because that assumes, as I said earlier, that we actually finished in 10 hours time. And very often we did not finish in 10 hours time. So that means we were only able to do maybe 150, maybe 175. We are doing well over a thousand an hour consistently and there's [01:03:00] no breaks in process. And so what I have been struck with is as we have implemented some of these processes is we have pushed the real-time problem solving to the people that do the work, outlined the overarching principles, only get involved in decisions that... [inaudible 01:03:16]. Like the decision to move from the ballroom to the exhibit hall.
That was a key strategic decision that we made. A decision a little bit on design, getting away from the two waiting areas, and some of those sorts of design concepts. We have [01:03:30] been able to demonstrate how you really can improve a process and improve a process very efficiently and effectively. And so I see... [inaudible 01:03:38] This has really opened my eyes as a health system leader, back to Steve's introduction, to really the true benefits of this work. And as he mentioned, healthcare is a target-rich environment for a daily improvement with so many systems. And so now my challenge is just how do I take everything that I have learned and try and bring that back in terms [01:04:00] of principles and implement it on a daily basis in our health system?
Gene Kim: Steve, this is making me recall the stories about how in a smaller setting, teams are able to reorganize batch similar work and double the throughput. Steve, can you just react to Trent and maybe tell a little bit about those stories?
Dr. Steven Spea...: Yes, so three reactions. First, earlier Trent had said something like come the fall this mass vaccination central will be closed down and hopefully it will never be needed again. But if so, they've been diligently [01:04:30] copying lessons learned to open it at the 800 rather than the 200 level per hour. The one thing I desperately hope and encourage is that the behaviors associated with going from X to 4X, that you guys don't put those away, but every day you find some other platform on which to practice those. And for a couple of reasons. One is the sort of the practical, which is there's something else right now, which is X. Which if it were 4X, people would be delighted. [01:05:00] And not only would your patients be delighted, but your colleagues would be smiling rather than frowning. But the other part is that unlike the cliche about never forgetting how to ride a bike, these skills do get forgotten.
And I would actually, if I wanted to be really adamant about it, say that by the end of today that you make a list of the next 10 things that can be platforms for continuing to hone these practices. Anyway, that's the [01:05:30] first. They've proven their worth so why throw them out? So that's one thought. The second thing I wanted to add was it comes back to this theme of microsystems and to Gene's point about if something goes wrong, it has to go up eight across and then back down eight to get fixed. Again, the cause of that is microsystems not threaded together into a real system. And so you're sitting inside a stove pipe and you have to go to the top of the stove pipe to find a connection down to another stove [01:06:00] pipe. So this ability to avoid that comes from doing exactly what you all did, which was building this system of systems to keep things moving along. And that's also really just outstandingly fantastic.
Gene Kim: You've seen this also play out in the healthcare system, right? The batching similar work?
Dr. Steven Spea...: There's actually another thought with this Gene is that we typically batch similar work by function, right? I've seen this sort of the relatively small scale of a primary [01:06:30] care practice where the medical assistance is treated as one unit and the MDs and the RNs and the administration is another unit. And I think what we see with Trent's example is this idea of thinking of the batch not vertically by specialty type, but thinking it horizontally by work type. There was this period, and it's a temporary period right, it's in a convention center and at some point you're going to dismantle a thing but it's temporary. That temporarily you had a focused flow of [01:07:00] work of getting people through these vaccinations with all the criteria that were necessary. The super cold storage and the two shots, et cetera, et cetera. Let me just offer, I keep coming back to this idea of building systems of systems, rather than having people isolated in microsystems.
Focus, albeit a temporary focus, and sometimes the focus is months, sometimes it's weeks, sometimes it's just hours. But the idea of focus actually allows the building of those integrated those threaded systems. And when we don't have focus, [01:07:30] when we say oh no, we can't be bound by commitment, then we run into this trouble. Is that when we can't be committed even temporarily, then we can't create the thread through the system to create systems of systems. And again, I think that's probably a transferable lesson onto those 10 things that are going to be on your list later. One last thought Gene, is Trent kept referring to this idea of we do the work, we pause, we reflect on the work, we improve the work. And you can imagine there are some people who'll say, oh Gene, you have to understand our [01:08:00] environment is so tumultuous that we never have time to actually create standards and reflect on [inaudible 01:08:06].
We just have to be improvisational the whole time. And I just want to say, that's a profoundly stupid thing to do and I don't mean that in a judgemental way. I mean, it really is stupid right? Because back to [inaudible 01:08:18] and look, their point is that we think in two modalities, one is fast thinking, one is slow thinking. And fast thinking is great, it has its efficiencies that we have these ingrained habits and when we have a trigger, [01:08:30] we have a response. And it's both time efficient, it's energy efficient, it's brain efficient. And then we have the slow thinking and the slow thinking is where we have to be deliberative when we're burning a lot of BTUs in our brain, et cetera, et cetera. And of course we can sometimes go deep and thoughtful on that, but it's wicked slow.
Now I think what Trent was describing was a meshing of fast and slow, and fast and slow all the time. So we do something which is our standard. That's the fast thinking piece, which we've created [01:09:00] a standard, trigger response, trigger response. Then we pause and we say, trigger response, how will that work for us? Now let's get into the slow thinking, which is the deep [inaudible 01:09:10] and the creative thinking. And then let's create a new standard so we can be in the fast thinking. So that's like the best of both worlds, right? Which it's taking advantage of the slow thinking for the creativity and taking advantage of the fast thinking for the efficiency.
Now, for those who push back and say, well Trent, you have to understand that I have to always be improvisational. Basically what they're saying, [01:09:30] in other words, is that I always have to be fast thinking. What fast thinking basically says is you're not being creative. And that's why I say, for those who say that, it's really a stupid answer. Because basically they're saying we want to act only on impulse. We act [inaudible 01:09:42] only on impulse and not through the deep contemplate of creative stuff for which a human beings are uniquely capable.
Gene Kim: And Trent, before I get your reaction... Steve there's one story that I'd just love for you to share, which was about primary care practice that basically decided we're going [01:10:00] to only do X type of work one day...
Dr. Steven Spea...: Yeah.
Gene Kim: Another type of work [inaudible 01:10:03] day. Could you just share that story?
Dr. Steven Spea...: Yeah. Trent, this is a small scale example of... And actually it got into vaccinations way back when it was flu vaccinations. But anyway, here's the problem. So you have this primary care practice, Mass General Revere, and I'm giving a name because they discussed that they should be wicked proud of what they accomplished. So I don't think I'm embarrassing anybody. And their concern was that they weren't meeting their patient needs. That if someone needed an appointment, it was a long way to [01:10:30] get an appointment. Once they showed up, it was a long time. The touch time versus a time in the office was just a crazy ratio. And they started asking the question, well, if we're working so hard, right, which they were. If you looked at any of the care providers, they were all busy. How can we be so busy and so unproductive? And what they found out when they started shadowing each other for some period is that they were busy. But they were busy reconfiguring the system all the time, as opposed to busy caring for patients.
And [01:11:00] here's where the reconfiguration comes in, is that when they started shadowing each other and then they started shadowing patients through the system, they realized... And again, this is kind of interesting. That this was a quote unquote, and I say air quotes with a lot of sort of disclaimer on that. Quote unquote, that this was just adult primary care. So it was no pediatric, no geriatrics, no specialty care. And again, I'm saying this just to make the point right, as simple as it gets. Well, when they started shadowing each other instead of shadowing patients, [01:11:30] what they discovered is that it wasn't simple. That they actually had 12 different appointment types depending on whether you were a first-time patient, follow up on a condition, a physical blah, blah, blah. Right. And it turned out when they started looking at the flow of work for each of those appointment types, it was different.
And so what would happen was if you arbitrarily scheduled patients based only on the availability of the doctor, everyone else had to reconfigure who went first, who went second. [01:12:00] What work did they have to do? To whom did they have to give the output at work? What was the output? Et cetera, et cetera. And I said, this is freaking crazy because why should we be spending so much of our creative energy on figuring out who goes first, second, and third, each time, and then having to reconfigure? So what they did is they said, look, what we're going to do is we're going to have temporary fixed flows of work. The key is both temporary and fixed. And they said, Dr. Weil... That was one of the guys involved, he was the medical director of the practice. Dr. Weil, eight to 10 in the morning, [01:12:30] we're not giving you any of 12 appointments. We're giving you three.
And the reason we're giving you only three, eight to 10 in the morning on a Monday, is that those three flows of work are more or less the same and won't require reconfiguration. Now for patients who need another appointment type, there'll be three different from 10 to 12 and three different from 12 to two, two to four, et cetera. And then we'll mix it up over the week. And then for the pushback that people said, oh yeah, but you know. So if the patient really, really needs to see Dr. Weil, well if they can't make an [01:13:00] eight to 10 on a Monday, maybe they can come into 10 to two on the Tuesday. They can only come in at eight to 10 on a Monday for that appointment type. Maybe they can see Dr. Amy instead of Dr. Eric.
All right. So anyway, it was crazy. By the simple act of finding out how many appointment types there were, committing to this temporary fixed flow of work, two hours at a time, they doubled their throughput. I mean, this is freaking insane. They doubled it. I mean, it wasn't like it went [01:13:30] up five, 10%. It's double. And you start thinking about what [inaudible 01:13:33]. And I just want to point out that this is a town north of Boston called Revere. They had 24 different languages they had to support in the community. So it wasn't just a matter of aligning the doctors, the nurses, the nurse practitioners, the physician's assistants. This [inaudible 01:13:49]. He also had interpreters. Like you had to make sure that on the appointments that you had the translator, the Vietnamese, or Farsi, or Arabic, or whatever else.
All right. Anyway, they doubled their capacity and made [01:14:00] the work easier. That's extraordinary. Because what happens when you double your capacity and the community hasn't doubled in size, it means that people no longer have to wait for weeks to get an appointment with a doctor. Some appointments are next day. And some appointments which might've been days worth of waiting were like, well yeah just show up, someone will see you. Oh, it was just phenomenal. And again, the reason I'm so excited by this example, one, it was really an exciting piece of work to have done. And I really [01:14:30] like the people with whom I work.
But the other part is though this was primary care, adult primary care... Trent was talking about vaccination and this was a much smaller population that Trent was dealing with... The principles are the same. Right. And I actually [inaudible 01:14:44] let me just tie it back together. So at one point, these people got so damn good at mapping out the flow of work and creating these temporary standards that it got to flu season. And they also had these 2, 4, 8 fold increases. So it started off that there was some [01:15:00] piddling number. Actually Gene, I think you know product plug here in the...
PART 3 OF 4 ENDS [01:15:04]
Dr. Steven Spea...: Piddling number. I actually, Jean, I think, product plug here in one of the chapters in my book on healthcare, we talk about this so we can check the actual numbers, but it's crazy. I think they had this crazy increase in flu vaccination where it got so fast. The rate limiter on the process was the elevator time, right? Like everything else was like... Because the problem with the elevator, because we're talking Boston area and flu season is that one, people had to park. Then they had to get into the elevator. They go up the elevator down there and the elevator only had a certain capacity, even pre COVID. [01:15:30] It had set number of people. And then when they got in the office, they had to sit down, take off a coat. So they said, "You know what? Screw that. Don't even come into the office. Just drive through the parking lot, slow down with the window open and just make sure your sleeve is rolled up on the appropriate arm, your passenger side right, and driver's side left." I would just jam you as you go through. And it was phenomenal.
Anyway, what was the storyline here is they did exactly what Trent and his colleagues did. So anyways, thank you for prompting [01:16:00] that very pleasant thought.
Gene Kim: Trent, I mean it was like when you told me that story outside of the convention center, I was like, "I got to connect with Steve." I mean, does that resonate with you at all?
Trent Green: Oh my God, it's incredible. And what I find fantastic about that story is when you get into this work, you start to uncover constraints you never thought was a constraint. Whoever thought the elevator speed was going to be constraint? We now have a constraint at the convention center parking. The parking lot is huge. I couldn't imagine that parking was ever going to be a constraint. There's only [01:16:30] 1800 spots. We process 1100, so can we get enough people into the parking lot, in and out of the parking lot fast enough in order to keep up our throughput. One thing I am struggling with actually is, as we think about this world, well, first of all, anybody who doubts that healthcare, isn't a target rich opportunity for implementing these principles is crazy [inaudible 01:16:50]. You just have to listen to Steve's story there. Our experience, I see it.
One thing I am struggling with though is this concept of focus. So as you articulated, [01:17:00] there's only one product that we're delivering at the mass fax site. It is a single shot. Where I am struggling in my learning and application is you think about our most complex medical center, a manual medical center, we do hundreds of surgeries a day. They're not all different, but there's a lot of hips. There's a lot of knees, there's cardiac surgery and there it happens to be a burn center. So there might be a few burn patients. I mean, it's [01:17:30] not single-threaded.
I would love to hear your thoughts, and I think your story about the primary care is a good one because you can start to break these down into simple things that can be improved for all. I've seen, I am a believer. I am trying to now drive these improvements or this philosophy throughout our entire enterprise, but I will encounter resistance to your point, not just on the fast thinking or the slow thinking, but just the, "Oh well, we're different." And yeah, you did all that at the convention, but it was one thing [01:18:00] that you had to do. We have a heart surgery followed by a hip surgery, helped me in my learning is to how to bring people along. And maybe some of the principles are that I can help common platforms that I can utilize.
Dr. Steven Spea...: Yes. So I'll give with the sarcastic cancer first, which is why do you follow a heart surgery with a hip surgery? Yeah, no, but let me unpack the sarcasm. That was just to make the point and be a little less dismissive. So first of all, there are the places where [01:18:30] you have limited predictability, limited control. So emergent care exists for the stuff you simply don't schedule. You know, there's the slip on the ice, the bang of the head, the ingestion of the foreign object, et cetera, et cetera. Those are situations where... And actually we did some work with emergency medicine as opposed to surgical stuff, where they actually were able to create a flow of work and get patients in the front door. From the time they got in the front door to the time they were [01:19:00] examined with their first round of medical orders was 18 minutes. I mean solidly 18 minutes. So this whole idea of what you did at the vaccination center of mapping out the flow of work and seeing how you could standardize, stabilize it, even in some emergent work, some portions of that can be done.
But let's say there's the other stuff you can't anticipate and nothing looks... All right, so that's why you have pods, which is put into a room, a patient with a particular, let's say idiosyncratic or even unique set of conditions, you put [01:19:30] the specialists in there around that person and they work out what they're going to do in a way that they're disconnected and otherwise buffered from the rest of the system. So the tempo of the rest of the system is not corrosive of their efforts inside the pod. And what they're doing inside the pod is not corrosive of the temple overall.
So that's that special case. But then you get into the question of these different threads, which is presumably the work during, before and after a hip replacement is [01:20:00] substantially different than the work before and after and during a valve replacement, and even a knee replacement for that matter. And if that's the case, then it does beg the question, can you run hips, knees and hearts in parallel? Or can you, depending on the number of cases and the amount of the facility, can you run some period of temporary focus for hearts and then some period of temporary focus on hips and center period, temporary focus on these before [01:20:30] you loop back and start doing hearts again?
This issue... Look, just one more thinking out loud, is that I appreciate the pushback you'll get, and you get it in a lot of industries, whether it's a service industry, not healthcare, but another... Oh, every client or customer is different. You get it in manufacturing, oh, we have such high variety, low volume manufacturing. But the thing I would offer is just do the peer to peer observation of how much of a person's day is spent actually doing something a patient [01:21:00] would care for and how much is spent actually reconfiguring. And the more time committed to reconfiguring, the less capacity you have to actually meeting patient need. So that's part of the persuasion.
Now, one last thought on this is, as you were giving your narrative about the vaccine center, I think what was coming through was that the amount of time you ran before you could take a breath, reflect and reconfigure, at first which you had [01:21:30] to run a lot because of how slow you are relative to the demand on the system. But as you got the system better and more skilled, the amount of time you had to run before you could pause, reflect and correct shrunk and shrunk and shrunk. And to an outsider coming in and looking at this today, they'd probably say, "Wow, they keep just reconfiguring." Because you're so good at the temporary focus and the reconfiguration and the next temporary focus.
Anyway, carrying that over [01:22:00] to hips, knees and hearts is a... I bet once you start looking at how much time people spend reconfiguring, then you'll identify the things that take a long time and a lot of effort to reconfigure. Figure out how to Kaizen those two. And when you do that, even if you have to run hearts, hips and knees, sequentially, your batch size can go down.
Gene Kim: Well, Trent, I'm so glad you're interjecting in here. It just delights me to no end. And Trent, I remember [01:22:30] telling you also at the convention center that Steve telling that story about the 60 line side store changes per day and how the VP of manufacturing for big three auto manufacturers said, "That's insane. They're probably lying or being disingenuous, right? We tried six and that didn't work so well. We shut down car production for three days because parts didn't end up to where they need to be. And we couldn't do final assembly and so forth." And for me, what that revealed was the notion of bringing down the cost of change. I [01:23:00] think Steve, what you're describing is that a high setup time, it has a very high cost of change. And one of the things that Trent is talking about is they were able to bring down the cost of change so they could dynamically reconfigure, dynamically adapt. And so that ramp of learning, it was now increased. Steve, am I... Trent, am I interpreting that story correctly?
Dr. Steven Spea...: Yeah. I like that phrase, the cost of change came down. So the frequency of change could go up.
Trent Green: And Steve's exactly right. Our plan do check act [01:23:30] or just cycles were... They're really micro now. I mean, very rapid. They were much slower. We spent a lot of time in a plan, probably not as much time in the do. And then the check-in, and I think we're getting much better at that. And that's actually a principle we're trying to bring back into our health system is we spend an enormous time on the P. We oftentimes spend very little time on the D, the do. And we are terrible, historically terrible, at the checking and [01:24:00] adjusting and what we're really trying to do, and what we tried to do, in the mass fax is plan quickly, do implement, but then spend way more time on the checking. Did we get the result that we had anticipated and, yes or no, and then how might we adjust?
Dr. Steven Spea...: You know, Trent is talking very much about this very disciplined, scientific method of planning something, declaring a hypothesis, doing something, testing a hypothesis, and then reflecting on it. It may sound like a little bit of semantics, but I just want to give credit back to this PDCA [01:24:30] thing and how some people fall into a trap. So the trap is you plan, you do, you check, what is the checking did I do according to the plan? And then how do I act? Well, if that's my model then I'm supposed to do according to the plan, and then I check and I haven't, then the action is punitive, which is, shame on you for not following the plan.
Now, Edwards Deming, who popularized this cycle, I don't believe he called it a plan do check act. He called it plan, do study act. [01:25:00] Because study means invest a little time to try and understand the gap. And the action then is to... Once you've understood the gap, it invites corrective action rather than a punitive action. And of course, what Trenton has colleagues are doing in that convention center, I think, was really true to Deming's intent, which was plan, capture your best. Demi talked so much about the generation and use of profound knowledge. And so what do you do with planning? You take whatever knowledge you have profound or not, or quasi profound [01:25:30] or partially profound, but you take your best knowledge and declare, "This is my best understanding." Then you do according to that, and then you study to generate more profound knowledge on which you can act. And everything Trent was saying is very consistent with what I think Dr. Deming intended.
Gene Kim: So Trent, one last kind of theory testing. So as we were sort of discussing you, and it was actually a great story from a pharmaceutical substitute with different functional specialties [01:26:00] for chemists and biologists and the purificationists. And you see the same dynamic, my whole air study around DevOps, right? Same thing. Development, QA, operations, security, same thing. And so kind of a recent sort of conjecture is that kind of, as the leader, our hypothesis is that one moves from this very much kind of in the act, react, expedite mode to one of... The social, I forgot the words for it, but contemplation, study, reconfiguration [01:26:30] of the system and then kind of the assessment of. So is that spectrum something that resonates with you in terms of maybe on your worst days to kind of maybe where you would ideally see yourself in five years, if you can force us into being, as you, I'm sure, will.
Trent Green: Yeah, I guess the way I've been thinking about it, the term I would use for what I think you're describing is coach. And so often now again, and I struggle with this. Here I am, I've been in healthcare for 20 years. I've [01:27:00] had all these experiences, I'm supposed to have the answer. So, okay, here's the broken system... "Call Trent, what's the answer?" And what I'm trying to transform myself into is not just... I don't know the answer. I mean, I might think I'm know an answer, but better are the people that are doing the work every day, seeing what the challenges are, so how do I turn myself into a coach and help them come up with the right solution and keep pushing and learning? Because I'm trying to... [01:27:30] We're trying to create ourselves, trying to turn into a learning organization and push people to get better every day and utilize these principles.
I love that the statement that this is not like riding a bike, this muscle needs to get exercised all the time consistently. And so what I would say, Gene, is yeah, what I'm trying to turn myself into effectively is a coach, which at times can be really, really challenging because there are times where I really think I know the answer and I just [01:28:00] want to say no, this is what we're doing [crosstalk 01:28:02].
Gene Kim: [inaudible 01:28:02].
Trent Green: Exactly. Here's what you're going to do. But I'm trying to relearn myself, coach myself, not to deliver that answer, but to help tease it out of people.
Gene Kim: Steve, I have to imagine you will validate this intuition, right? Given the last half of your book, right? This is exactly what was embodied in your study of [inaudible 01:28:22] as a production system. But at the most senior levels of leadership, I think that very much matches how they viewed themselves. Steve?
Dr. Steven Spea...: Yeah, 100%. [01:28:30] Look, any organization, whether it's a five 55,000 half a million, that's a lot of brains. You start thinking about it in terms of BTU's, right? Some crazy number that the... What is it, Trent? You would know this better than me, but isn't it like a quarter of your BTU spent every day is by your brain, right?
Trent Green: Yeah, I think so. Yeah.
Dr. Steven Spea...: And there's some crazy numbers, Gene, like in the midst of a chess tournament, the grand masters they're burning six, 10,000 calories a day. There's Michael Phelps training for the Olympics. It's crazy. [01:29:00] All right so think about just the energy efficiency problem here. You've got all those brains burning all those calories, and as a leader, your default, what, it's going to be that you're not going to put those brains to good use? I mean, it makes no freaking sense, right? So what we're talking about is this coaching model, which Trent was alluding to, was how do we take advantage of that distributed intelligence? All those engines of creativity, which are just sitting there idling anyway, right? I mean, the engine is running, they're [01:29:30] burning calories. So put it to good stinking use.
And on top of that, if you start creating these threading mechanisms, not only do you have the distributed intelligence, you now have the collective intelligence. People having conversations about, well, here's my work, what's your work? How does that work come together? How can the two of us do the work in such a way that the whole is greater than the sum of the parts? And now the only way that's to happen though, is that if Trent and his colleagues, coach people how to behave that way and have those conversations. But what a freaking multiplier, [01:30:00] I mean, no offense, Trent, but whatever BTU's your brain is burning in a day, its creative potential is nothing compared to the 100 people in the convention center or the 200 people in the convention center, who if you could get that... tap into their creative engines, oh my God.
Trent Green: Yeah, 100%.
Gene Kim: Gene here. Steve had mentioned the flu vaccination effort. One of the many improvements made at Mass General Revere, which was indeed described in his book High-Velocity Edge. On page 330, [01:30:30] there's a table that describes the vaccinations administered during two-hour sessions. And in session number one, it was 43 vaccines delivered to 71, to 151 delivered in the third observed session. So there's your four X improvement in throughput. But it gets even better than that because the table then shows the number of clinical staff required to support them, which has reduced from three and a half people to two and a half, which drives the number of flu [01:31:00] shots administered per hour per person from 6.1 to 14.2 to 30.2. And so that's a five X improvement. If you're interested in the work that Steve described, there's a whole chapter on this and it's just a fantastic read.
Okay, number two. I had mentioned that I had spent the time with Mr. Green at the Portland Convention Center with Dr. Chris Strear, an emergency room physician and who will soon be chief medical officer at Columbia Memorial. [01:31:30] He gave a fantastic talk at the DevOps Enterprise Summit two weeks ago. And so many people mentioned that it was one of their favorite talks. He described the work that went into his book Smash the Bottleneck: Fixing Patients Flow for Better Care (and a Better Bottom Line) where, among other things, he was able to reduce the ambulance divert time from six hours per month, where the emergency department were unable to accept incoming ambulances often due to lack of availability of beds, down to 45 [01:32:00] minutes per month. But he also described how flow has never been so important in this time of COVID. He described how last year, it was the first time he felt like his life was actually in danger, driving him to make sure that his will was in order after years of procrastinating and that he was unable to see his kids for nearly a year.
But here's what hit him. We know from Dr. Patrick Cawley, CEO of Medical University of South Carolina Health Systems, described in the first break-in that [01:32:30] volumes in the emergency department are down significantly. Dr. Strear described how he went from seeing 25 patients per shift down to a mere handful. He described how in one shift, there were only three patients in the emergency department, the slowest he's ever seen, but one of whom was on a ventilator and it still took six hours to move that patient from the emergency department to the ICU. He kept thinking, "This is one patient in a nearly empty hospital. What happens when we're [01:33:00] overwhelmed with dying patients? How will we handle that?" And that's what scared him. I will put a link to this short 15-minute talk in the show notes. And I think there are lessons relevant to any technology leader.
And third, lastly, on coaching. You may have heard me laugh when Mr. Green mentioned that he saw himself more and more as a coach. So I laughed for a couple of reasons. One reason is that Mr. Green was actually the coach for my oldest kids softball [01:33:30] team back in 2015. He was a fantastic coach and exhibited all of the patience you need when dealing with six year olds who are playing a sport perhaps for the first time, which was definitely the case for my son, Reed. Number two, the role of head coach seems so interesting. Steve and I have talked numerous times about coaches in American football. I'm not an expert on sports, but the role of coach seems like a very interesting analog.
[01:34:00] I am reading from an article called A Detailed List of an NFL Coach's Responsibility from the Bleacher Report. We see them in press conferences after games, during the combine and yelling on the sideline each Sunday, but very few of even the most intense fans know what an NFL coach actually does. It really varies from coach to coach and from team to team. Every team has a fit they like, and every coach has a way they like doing things, but there are some very consistent and regular things an NFL coach is responsible for.
So [01:34:30] I'm just going to read some of the bullet points off. Post-game analysis. It's a coach's job to hit the game tape almost immediately after the game is over to figure out what worked and what didn't. Number two, game planning. Implementing whatever the staff learned from the preceding game is only part of getting ready for the next week. The coach will also look at the last game or two of the team's upcoming opponent and use all that information along with the assistant coaches and the offensive and defensive coordinator to come up with a way to beat [01:35:00] the next team.
Number three, practice makes perfect. All the while the head coach is making sure things get done on the practice field. And then number four, game time. When the day of the game comes, a head coach is still tweaking the game plan. During the game, they'll continue to make adjustments based on success or failure of the plan that they have come up together over the past week, keeping that bird's eye view of the action in the sense that they need to know what is happening with every aspect of the team. How is the quarterback holding up after [01:35:30] that big hit in the first quarter? Is the middle linebacker handling that hamstring strain? And so on and so forth.
All of that information goes into adjusting the game plan that was worked on all week so that it works in the second half. Most of the major decisions come down to head coach as well. Sure, the offensive coordinator might call the plays, but the head coach is the one who pushes for that big fourth downplay or the fake field goal attempt. It's the head coach will be facing the press core after the game to talk [01:36:00] about every decision that was made. So I find this interesting because nowhere in that job description of the football coach is it to get on the field and actually play the ball. Most of the work is happening before the game. So I'm finding the metaphor of the sports team coach as a very appropriate metaphor for what the job of the leader actually is. More in the slower modes of cognitive thinking around planning and assessment and not so much getting involved in daily operations. [01:36:30] That is in the domain of the people on the team who are actually on the field. We will explore this, I'm sure, in future episodes.
Okay, I had mentioned that I had so many more questions for Mr. Green, but because of time limitations, we couldn't get to all of them. So I am very much hoping that we will be able to do this again with Mr. Green and Dr. Spear in the weeks to come. So let's go back to the interview where Trent describes how he can be reached and what topics [01:37:00] he would like to be reached out to about.
Trent, I actually had a bunch of questions I've still queued up, but I know you have a hard stop. So I'll end on this. Trent, thank you so much for being here with Steve and me and sharing your experiences, your expertise, for such, without a doubt, the most societal important problem on this planet, which everyone's health and economies reside upon. So if people want to reach out to you, how can they do that? And is there anything in particular that you want people to reach to you about?
Trent Green: [01:37:30] Well, I just want to thank you and Dr. Spear for having me, this has been fun. People can reach out to me. They can find Legacy Health. They can find my bio on legacyhealth.org. My email address, we can probably put it into your show notes and I'd be happy to have people reach out to me. And I'm trying to find other learners that have hooked onto these ideas that Steve and others have been talking about for a long time. So anybody who has, they're willing to share, I'd be more than happy to connect them.
Gene Kim: Perfect. Thank you so much, Trent and Steve. [01:38:00] I hope you learned a ton and were as inspired as I was by all the amazing achievements by Mr. Trent Green and team, who are now vaccinating over 8,000 people per day here in Portland, Oregon. Going to the Oregon Convention Center, seeing the work being done there, has been one of the highlights of my year. It is truly awesome to see the scale of the mission, and to what extent human creativity has been unleashed in service of the most important mission on the planet today, to vaccinate [01:38:30] everyone on the planet in the fastest possible time.
So please join me next time. When I will be speaking with Dr. Gail Murphy, professor of computer science at the University of British Columbia. She is one of my favorite academic researchers in all things related to architecture, modularity and developer productivity. The ideal cast is produced by IT revolution, whose mission is to help technology leaders succeed and their organizations win through books, events, [01:39:00] podcasts and research.
PART 4 OF 4 ENDS [01:39:22]