The June 2018 Gallup survey of measures faith and trust in institutions shows that Medical Institutions has dropped 38% over a 40 year period. Plus we discuss putting healthcare data to work.
Bill Russell: 00:10 Welcome to this week in health it where we discussed the news information and emerging thought with leaders from across the healthcare industry. This is episode number 42. Today. We venture into the world of data governance and we take a look at a story that is really a call to action for a for health care and health care institutions. This podcast is brought to you by health lyrics. Health systems are moving to the cloud to gain agility, efficiency, and new capabilities. Work with a trusted partner that has been moving health systems to the cloud since 2010, visit healthlyrics.com to schedule your free consultation. My name is Bill Russell, recovering healthcare cio, writer and advisor with the previously mentioned health lyrics. Before we get to the show, I wanted to to shout out to all CIO is attending the Chime fall forum in San Diego. I’m going to be.
Bill Russell: 00:56 I’m going to be there and doing one of the special episodes that we do for this week in health it where I sit down with cios for about 10 minutes. I have four questions. In fact, this time I’m going to do it a little different since there’s so many cios and cmio’s going to be down there. I’m actually going to have three different sets of questions and you get to choose which sets of questions you want to answer. It could be on technology, execution or culture are the three topics I would like to go deeper on. So I have four questions in each one of those categories. If you have some time and you would like to be on the show, uh, we, I will just sit across from you, we record on the iphone and then I spliced them together and I will ask a bunch of people the same set of questions.
Bill Russell: 01:37 So, um, if you’re going to be there and would like to be able to show, please drop me a [email protected] In one last thing, uh, we now have a, an alexis skills enabled for the podcasts. Uh, you can just say, Alexa, play this podcast, play the podcast this week in health it, or you, uh, you can enable the, any, any pod skill on your, uh, echo and say, Alexa, ask any pod to play this week in health it, uh, and I think it’s fitting that I would announced this on this show since our guests today, and I discussed the transformational power of voice in health it on our last episode. So today’s guests was gracious, gracious enough to step in when one of our guests wasn’t able to, uh, follow through based on the scheduling conflict, a friend and accomplish cmio. Lee Milligan joins us once again with Asante health. Good morning, uh, Lee and welcome to the show.
Lee Milligan: 02:32 Afternoon. Happy to be here
Bill Russell: 02:36 but we’re, we’re ruining the illusion because this gets. This will get released on Friday morning. I don’t want people to see the beautiful sunlight behind me and recognize that we’re actually recording in the afternoon.
Lee Milligan: 02:47 Good day.
Bill Russell: 02:49 Exactly. So you’re, you’re just getting back from the, uh, the HIMMS conference up in Washington. Uh, give us a little idea of what went on up there.
Lee Milligan: 02:58 It was great they have a semi annual conference up there and they invite a variety of folks to, to come in and speak. They had a folks in the clinical realm. They had vendors, uh, they even had some folks who represented the government. There was, there was a guy there who was from hhs, Mr Graham, and there was a lady who was our representative from the, the state who was there as well. Uh, I thought they did a terrific job of executing on the conference.
Bill Russell: 03:24 That’s great. I just did a podcast with Sarah Richardson the socal Himms podcast and following me, she was interviewing a gentleman who was running for Lieutenant Governor here in the state of California, uh, who is an ophthalmologist who was talking about all things health it. And it’s pretty exciting. In fact, she just released that episode, the Socal Hymns podcast today. So a lot of it’s political season. A lot of good conversations have been happening and I’m glad to see, you know, the, the politicians starting to weigh in and become a part of the, uh, a dialogue. So we’re getting a two way dialogue going. It’s kind of Nice.
Lee Milligan: 04:04 It’s interesting seeing the nurse who is also a state representative, a nurse, a state rep, feels kind of a part time deal. And so she really could represent from the clinical front Well. And she was well spoken.
Bill Russell: 04:18 Yeah. That’s awesome. Alright. So it’s been a little while since you’ve been on here. And so one of the things we like to do before we get going too far is, you know, what, what, what are you working on these days and what are you excited about?
Lee Milligan: 04:31 Uh, there’s a lot. I wouldn’t say the one word that comes to mind right now is prioritization. A, it sounds a little bit on the dull side, but trust me, getting this right is key to our survival and our sanity in this world, particularly around analytics. We found that we had a, just such a large amount of requests coming in, we really had to rethink our strategy. We’re maybe three years into our original roadmap and thought we did a pretty good job of pairing that down a bit and having some prioritization. But we realized that despite our progress, we still had an impasse, and so we’ve worked very hard to go from a whole number of cues that we had previously down to a single cue a for work. We added a cue recently to address population health and some of these new business models that are coming out.
Lee Milligan: 05:24 And that new queue I’m really, really proud of. It has representation from a variety of departments. It’s got population health, it’s got quality, got analytics, it’s got finance, um, it’s got operations and they worked together to come up with a process around prioritization that’s really healthy and transparent. Um, and works well. They developed a tool that has a weighted score based on different categories and they run the requests through that tool and then a number of pops out. And that’s kind of a starting point for where it’s prioritized. The great thing about it though is when folks want something done and they look and they say, hey, it’s not making any progress, they can clearly see on our sharepoint side how it shook out and what the discussion looked like. And what I’ve noticed is that folks have a lot more literacy about that process and very quickly understand kind of where their individual requests lives.
Bill Russell: 06:25 Uh, I’m looking forward to this because the whole series of questions that I have for you this afternoon is really all around data governance and really getting pragmatic about it. I mean, how do you just. One of the things I remember of course it’s been, um, what has been about three years since I’ve been cio, but one of the things I remember is just the sheer number of a regulatory compliance. The ones you actually had to do that lists was pretty big just in itself.
Lee Milligan: 06:54 I agree. And the problem with that is it shuts down everybody else. So if you’re a nurse manager, uh, in the, in the hospital and you have a great idea about how to improve care or throughput or something, you bring it to a committee and all those regulatory ones get scored really high and you get kicked out. So coming up with a mechanism to be able to support those folks who are doing great work and really want information to inform their work while not letting down the regulatory folks has really been the trick.
Bill Russell: 07:24 Yeah, and it’s interesting. You almost wish you could say no to some of the regulatory because you’re looking at it going, what are we doing? Does this pass the sanity test of should we even be measuring this?
Lee Milligan: 07:36 So true. Well, we have done what ones the things we have done in terms of the, uh, the payers we’re interacting with. And this is a very recent thing we decided to tier our payers based on whether we’re in a novel contract with them. So our verbal we’re in an MA contract with a payer that is at a high end deal. We both have skin in the game, a true partnership and those folks are entitled to a whole bunch of work for my analytics team. If you’re a typical payer and we have just a regular contract and maybe there’s a little bit of pay forbodance as part of it, that doesn’t allow you to get additional efforts. So in other words, our team has some standard reports that can come out, but extra stuff requires extra commitment.
Bill Russell: 08:23 Awesome. Are you a, are you writing any of this up in like white papers or stuff that you’re going to share or is it just mostly speaking engagements?
Lee Milligan: 08:32 I’m talking about this here with you today.
Bill Russell: 08:35 There you go. All right, well we’ll start getting it out there and see what happens. All right, so let’s, let’s get into the meat of the show because I really am looking forward to asking you questions around data governance. So our show breaks down into two segments in the news and soundbites. You’ve been traveling and because you’re gracious enough to step in here, we’re going to do one story and then we’re going to. I’m going to do a small section where we pay tribute to Paul Allen with a little blurb that uh, uh, Bill Gates wrote about them. I think anyone in health, it has to recognize what the passing of Paul Allen, one of the, one of the people that really helped to usher in the era that we currently live in, good, bad or indifferent in terms of, you know, the complexity. Uh, he was one of the people who saw it early on.
Bill Russell: 09:22 So here’s our first story comes from the Keckley report and this is going to be a little harder for us to, to talk about because it’s sort of a call to action and it’s a wake up call for us. And uh, if you want to get to this report, it’s uh, www.paulkeckly.Com. And this story is distrust in the US health system. Are we paying attention? And he has a chart in here and it’s the Gallup poll measures how people feel about a from a trust standpoint, our various institutions, military police, business, church, Presidency, Supreme Court, you name it. And in Gallup’s. So I’m just gonna read here so in Gallup’s June 2018 survey trust in the military and small business scored highest consistent with prior surveys. Not a surprise, but of the 15 institutions, only one has seen its level of trust fall as precipitously as Congress, and that is the medical system over the 40 year period, trust in Congress has fallen 34 percent and trust in our medical system has dropped 38 percent.
Bill Russell: 10:25 Uh, these findings, parallel polling done by Pew Research, Harris and others, trust in the US healthcare system is low. Edelman called it an extreme trust loss and it goes on. It says stories about fraud and where’s, where’s is it coming from? Stories about fraud, price gouging, a coverage denial, excessive profit, avoidable error, ethical breaches and unhealthy work environments, cast doubt and uh, uh, compromising truth in our healthcare system. The frequency of news coverage and social media attention to our misdeeds is intensifying lending to the general distrust in healthcare. Uh, surveys Harris, Pew and Gallup show that the public’s trust in our systems, major institutions, hospitals, health insurers and drug manufacturers at near all time, lows and high profile disputes between insurers, hospitals and physicians lend to Publics Suspicions that are talks that our talk does not replicate or walk. Uh, so again, uh, I guess where I want to start this conversation is, um, should we be worried about this? Is this, is healthcare still local that really, uh, you know, if you’re a good local health system that the trust level’s going to be enough that we don’t have to worry about it? Or is this general polling that’s going on across the country? Should, should we be concerned? And is there something that we should be thinking about?
Lee Milligan: 11:49 A well in the spirit of good data governance? I think the first question came to mind for me is define trust. If trust is this kind of blind thing where we automatically say whatever this entity is telling me, I’m going to believe, then I’m actually happy that distrust has cropped up a bit. And I think it’s, you know, it’s part of the information age with the Internet and being privy now to all the frailties of our institutions. I think it’s just a natural outgrowth of that information is we’re going to have less blind trust, blind trust I should say. But I wonder if a better word would be confidence and confidence in my mind kind of has a portion of trust, but also allows for a little bit of nuance as a word for the system I was taking about reasons why folks might not have as much trust in the system. There’s a lot of things that are happening right now that I think plant seeds of mistrust in folks’ brains.
Lee Milligan: 12:50 One is when you just go see your doc and the doc spends the whole time looking at the computer because I think patients first and foremost, wanna know, want to feel like they’re heard and that the doc understands where they’re coming from and if the doc is absorbed by the computer, they’re not going to get that sense. Um, I also think about our process for releasing a health information to our patients. It seems clunky and cumbersome and you know, to somebody who feels like, you know, that’s my information. Why is it being withheld for me? I think it just sets up a dynamic where folks might feel like there’s a level of distrust there. So,
Bill Russell: 13:30 so what’s appropriate. I mean, so let me, let me do this. From a patient standpoint, you’re a physician, I’m a patient and I’m coming to see what’s appropriate for me in terms of my lack of confidence, lack of trust. So I’m sitting across from you and you say, Hey, I’m going to order this battery of tests and I know that, you know, I just had these tests, am I did. So I pushed back on you. If you’re looking at the computer the whole time, do I say, hey, you know, what, Lee my face is over here. Maybe you might want to look at it or I mean, what? I mean, how far does the patient go? I had a story this past week, a friend of ours who, uh, whose appendix burst and the first trip to the doctor and they did the blood work and didn’t see anything. And then, you know, 24 hours later to the appendix burst and the Qa, the someone actually said to this person, yeah, you should have pushed the doctor for the other test. And I thought, well that’s interesting. I mean, the person who doesn’t really know anything about healthcare is, is really the one responsible to push the doctor. I mean, how, how far should the patient go with this?
Lee Milligan: 14:33 Well, the appendicitis example is a good example. I frequently docs, uh, rely on some blood tests associated with that and yet we know that blood tests are awful, but it’s one of those things in medicine that continues and it’s one of my hopes with the advent of ai and different things that we do that our pathways will not allow us to go down and use these things that we know actually don’t work. But in terms of how far you push the doc, it’s a bit of a delicate dance. On the one hand you want to be informed, you want to make sure that doc is kind of doing their job the way they should be doing. At the other hand, you want to give them some autonomy to think through the situation and come to a good conclusion in terms of, you know, or here, here I am a. If they’re really spending that much time, sometimes I think it’s beneficially just call it out and say, doc, I can see you’re kind of absorbed in that trying to navigate your ehr world, but I want to make sure we have a one on one conversation here as well and sometimes that helps them reset.
Bill Russell: 15:29 Yeah, I could see that. So let’s talk about probably two different directions here. One being, have you guys, have you guys thought about this? Have you set up a team of people that’s just focused in on the physician’s experience during that visit that they’re looking specifically at the clicks and how to get the screens out from between them and the patient and those kinds of things. Do you have an a, maybe an experienced team focused on that? Is that something you guys are doing?
Lee Milligan: 16:01 No, we haven’t done that, although I love the idea of we do have our decision support team that’s recently stood up that i co-chair where we’re attempting to get our arms around this insanity of all of these things that are popping up in the dock space. I will say in general, for most of the bill that comes through that I’m a part of, I do spend a lot of my time ripping stuff out and getting rid of stuff that is in the docs way sometimes that’s extra clicks sometimes that’s extra icons. And I, I, I kind of feel like it’s part of my duty is to overlook that and make sure it doesn’t happen. Truth be told it would be better to have a whole committee of folks who can do that, but we’re a medium sized system.
Bill Russell: 16:43 Yeah. One of the things that may… No. I understand that completely. So one of the things that keckly goes onto say at the erosion of public trust invites displacement by alternatives. Medicare for all being one where governments, the solution, Amazon health being another that he cites, which is really creative entrepreneurs, uh, being the solution. And uh, do you think one of the things we talked about last week was that 40 percent of millennials, according to a recent poll, do not have primary care physicians. They rely solely on clinics. So they’re just going to clinics, getting there one off visit and they’re done, and then they, they go on. Um, are we, are we inviting disruption in here where somebody figures out a way to create that environment that, that a millennial will walk into and go, yeah, now that’s healthcare. That’s how I want healthcare to be practiced.
Lee Milligan: 17:39 I think so that, that article is an interesting article, it’s not surprising that, you know, with how mobile younger folks are and how healthy they are. It wouldn’t make sense to have the same model that’s been in place for 100 years. Um, but to go back to just a couple of additional ways that I think we can undermine folks trust. I was thinking about a note I just read a couple of days ago from one of my partners and I looked at it and they’ve used dragon and there was a whole bunch of Gobbledygook in there. It was, you know, words you can’t understand at all. And, uh, I looked at it and I thought if I was the patient and I reviewed that I would think my doc is either, you know, they’ve lost their mind, they’ve done some LSD or they’re not careful enough to review their own notes before they actually hit sign. And so those are additional ways I think we’re undermining trust.
Bill Russell: 18:35 We can go back to how we’re managing physicians so that they have to sign those notes so quickly and just the, the whole Rv model and some other things, which I know that gets a lot of discussion as we go out there. Um, you know, I want to come back to the Bill Gates Paul Allen thing at the end of the episode. So let’s, let’s go into the soundbite section. I do want to spend some time talking about data governance and some of the exciting things that you guys have been doing. So let’s, let’s start. I’ll start with a set of rapid fire questions and they’re pretty basic, but I just want to level set people before we get going. So when someone asks you what’s the definition of data governance or what is data governance, the principles and objectives around it, why should, why should I care what, what, what do you tell them?
Lee Milligan: 19:20 Well, that word, first of all, I think is a really confusing word. It sounds almost bureaucratic, um, but really it boils down to exercising, of decision making and authority for data. So you just have a team of folks who are charged with looking at the decision making scenario around data and without having a structure around that, what you find is the data just falls apart, that if there is an accountability around it, people don’t put data in the way it needs to be put in and therefore can’t be taken out on the backside the way it needs to be taken out. And at the end of the day, if you do it right, your data is fit for business use, not perfect, and maybe not even a super high threshold, but fit for your intended purpose for it.
Bill Russell: 20:09 Yeah. And data data governance essentially says, begin with the end in mind. It’s saying, how are we going to use this data? Okay, now let’s step back and put together a data acquisition strategy and it would’ve been nice if we’d started all the Ehr’s with a data governance program instead of implementing an Ehr and then say, okay, we collected all this data, now what do we do with it? So data governance really is began at the, at how are you going to use the data and then work your way back,
Lee Milligan: 20:36 you know, having data stewards who really understand the workflow is key to the whole, the whole success of the program.
Bill Russell: 20:44 Yeah. And how do you, how do you embed those analytics and those things back in? So what, what are the elements of a good data governance practice? We’re going to try to make this as sexy as possible. It’s kind of interesting because it’s people hear data governance, they turn off, but this is like, this is, if, you know, as they say, data is the new oil data is what runs Amazon. It’s what runs google. It’s what runs apple at this point. Uh, and it’s really what is, is at the core of what can really transform healthcare. And we know that and data governance is just a way of us using data effectively within healthcare. So what are some of the elements of good data governance practice?
Lee Milligan: 21:27 So we looked at a whole variety of different areas we can spend our time on and recognizing we have finite resources. We boil it down to four main pillars for us, uh, and we stuck to those pillars for the last two years and that’s been pretty effective. Um, we started out with accountability, um, and then we also have proper use a quality and then movement, both movement within your system and from your system to another system. And that was plenty of work to get our arms around. On the accountability front, we ended up setting up a structure that has four layers to it. At our top layer we have our data governance steering committee, which is made up of our primarily our c suite. They meet every other month and they set strategy about where we really should focus our efforts and reporting to them are both vps and directors who make up our data, our data governance council and the data governance council really have a specific area of focus for their domain, so it might be finance or revenue cycle or quality or nursing, et cetera. And then below that we have our community of data stewards and those folks are the ones who really understand the workflow associated with the individual question at hand. And then below all that we have our office of data governance that supports those efforts.
Bill Russell: 22:50 How often do you get all the data stewards together?
Lee Milligan: 23:00 Yeah, and that’s, it’s been interesting because the data stewards, we thought that would be kind of an even cross section, but it’s been variety of different levels of folks who’ve been involved to really understand the workflow in a way that I didn’t anticipate going into this. Uh, but if you get the right people understand the workflow, they can very quickly resolve a problem. Um, but that’s the key is getting the right person and the way we structured it around the accountability front is that each layer has a reporting relationship and that was the key because we didn’t have a extra folks, we can just kind throw it this effort. These were folks that have full time day jobs and we’re asking them to do extra stuff on top of that and we recognize it despite the fact this is a institutional priority in general and we talk about it at our meetings, etc. Unless you have an accountability built into that, we felt like it would be a, you know, prone to failure.
Bill Russell: 23:58 You’re steering your c suite as your steering committee. Are they just setting more or less the guidelines or the principles of the, of the program and then essentially acting as the escalation point for those things which can’t be solved at those layers just below that.
Lee Milligan: 24:21 Yes. Both of those things. Yeah, for sure. In the beginning it was really about educating the folks around the table about what we’re trying to accomplish and we, we planted some seeds at that time. We asked them to start looking at their own reports because they are all getting reports from different areas, but really pushed back on the people who are generating those information products and handing it to them and asking them to ask the people who gave it to them, what’s the quality of this data, how do you know the quality of the data is what you say that it is, and that has generated some interesting discussions. Uh, and then on top of that they help evangelize what we’re trying to accomplish and support the folks who were actually doing the work. The data governors they meet, they meet monthly and they meet for an hour and a half and they have four subcommittees that work on individual topics and they’ve been very engaged in this process. But to have vp level and director level folks spending this much time on something, there was no way we were going to do this unless there was an accountability structure that accounted for that.
Bill Russell: 25:20 Right. So when you say c suite steering committee, how many people from your c suite do you have on it?
Lee Milligan: 25:28 It’s a, it’s a portion of it. So I’m probably guessing it’s a roughly half.
Bill Russell: 25:33 Wow. So that shows a significant commitment from your organization around the value of data and the value of the insights that comes from data, I would assume.
Lee Milligan: 25:45 Yeah. I feel like that’s the case. Our current ceo roy vineyard, he really did a great job of understanding why it mattered and evangelizing the importance of this. and I think he set the tone for the rest of the organization.
Bill Russell: 25:57 Right. So that was the next rapid fire question, which was how do you get started? And I would assume you have to start at that ceo level and get. And if your ceo understands the value of data, then it’s a matter of coaching them on, okay, here’s how we’ll put together a program, here’s what you can expect from it and here’s the communication that really needs to come from the top down to, to get everyone really on board. But it sounds like if you had a ceo that was there maybe was, was the ceo the instigator or was it sort of a, a dual kind of thing?
Lee Milligan: 26:33 There was a preparatory period where, um, I spent a fair amount of time talking to him about it. Also, our cio has spent a lot of time talking to him about it as well and he’s a very detail oriented leader and he really wanted to understand a lot of the details and once it clicked it became clear to him. Then he really started to evangelize why it mattered and that again, really set the tone in terms of starting. Um, I would say call us up and I’m half joking when I say that, but I will say that our system, when we put our system together. We spent time talking to a whole bunch of other systems out there, that have been down this road before ended up collaborating with like 13 different, uh, health system across the country. And it was reallY helpful to talk to other folks who are waging the same, the same battle and their institution.
Lee Milligan: 27:23 And we learned a lot. So I’d be happy to connect folks who may be watching right now to kind of jumpstart it. But in terms of where to start, I think you really have to have two people in the saddle out of the gate. You have to have a, an executive sponsor, so that’s typically a cmo, cfo or cmio. And then on top of that, I think you have to start out with a data governance program manager, somebody who can really devote their time to understanding what we’re trying to accomplish and putting in place a framework for accomplishing it.
Bill Russell: 27:58 It was interesting, I was on a, I was on a webinar where they asked who should lead, who should be the executive lead of the data governance program and as you and when people answered the most, uh, so the cio was like 20 percent. The cmo was like 18 percent, cfo was only two percent. coo was like 20 percent and other was 20 percent is. So it was an interesting that there wasn’t a consensus on who should really own it. And do so you see across different health systems that different people own the data governance.
Lee Milligan: 28:40 Yeah, I’ve been disappointed to see a lack of c suite or executive sponsorship for it. I can tell at least so far it feels a little bit like a hot potato and they know it’s important, but they also know that it’s kind of in its early stages. It’s Almost like an early ipo and they, you know, they want to, they want to be a part of it, but don’t necessarily want their name attached to it. Um, but I think you just gotta you gotta identify somebody who really understands that and be recognizes how it’s going to benefit the organization, who can really champion the effort.
Bill Russell: 29:13 Yeah. And it’s a leadership role, right? So it’s, it’s a, you are influencing your, changing the culture around data across your entire organization. So it requires a special person and it’s not something that you can just throw over the fence. I see a lot of health systems now hiring chief data officers and saying, okay, we’re going to give it to this person. Well, okay, that’s fine, but don’t hire somebody that’s just really good with data. You have to hire a leader who can, uh, who’s respected across the board, who can interact really well with the leadership and, and, and move the culture forward with education and other things. I would think
Lee Milligan: 29:54 really well said. I would say that in our organization to beginning the first year or so, it was all about educating folks about why this matters and one of the reasons or one of the things that we did in order to accomplish that is we profiled our own data. I think I might’ve mentioned this before, but when we decided to profile our data, we were initially going to purchase a tool to accomplish that. And we were, we were trying to do this on that on a cheap budget. And we ultimately built out a tool, a sequel based tool to query our data. And the program manager of Mark Stockwell who I’m really fortunate to work with, built this out. And we applied it to some really straightforward cases. So that was the case where we looked at the number of patients we had are older than 125 and we had like 4,000 patients over 125 and we recognized pretty quick, Hey.
Lee Milligan: 30:47 We’ve got a problem here. We looked at our A1C’s and they were, you know, they were kind of all over the map, but then when we asked, you know, good hard questions like how many patients have an, A1C’s that’s physiologically impossible and we had like 300. And so when we start to find those cases, presenting that information back to the c suite, they begin to understand how much it matters to get this right. And then the other piece about this is for several of the regulatory, um, uh, programs, either the cmo and our, in our system, the cqo or the cfo have to sign off on the validity of the data before it’s submitted and showcasing some of the data issues and then showing how we’re working the problem allows them to feel confident about signing off on that data.
Bill Russell: 31:37 Interesting. So by profiling the data, what you’re doing is you’re going into the raw data, you’re pulling things out and you’re asking questions against that data and really doing more of a sniff test to say, hey, how good is our data?
Lee Milligan: 31:51 Yeah, yeah, exactly. Because there could be a lot of erroneous data. it’s underneath that crazy threshold that we set, right? We said 125. That doesn’t mean that everybody who’s below that actually has the right birthday, but it does, as you say, it does kind of provide some information in terms of the sniff test of how our data is doing.
Bill Russell: 32:11 So one of the people I respect a lot on this topic is Dale sanders, a health catalyst, a chief. What is the chief technology officer? I believe. Um, and he wrote a paper where he talked about the triple aim of data governance and I am sure he got it from somewhere else. Maybe it’s original with him, I’m not sure, but triple aim of data governance is a number one, ensuring data quality number two, building data literacy. Number three, maximizing data exploitation for the good of the community. Um, so let’s walk through these real quick. So ensuring data quality, what are, what are some of the best practices that you have seen either in your system or in the industry for ensuring data quality?
Lee Milligan: 32:50 So I think you gotta start by, by assessing current state. I think it’s hard to say, you know, you can’t really work on it unless you know where your starting point is. So that’s that whole profiling piece, but it also has to do with the aging, the data stewards and the process to better understand how it’s, um, how the data is coming in. In addition, when we report on that data, it’s important to engage the folks who are consuming the information products to ask them, how does this, how does this shake out? Um, once you do that, I think you have to identify which areas you’re willing to work on and which ones are willing to let go. There’s just so much data in a health system, it’s, it’s hard to focus on everything. So you got to pick your battles and zero in on things that you think are going to be the kinds of things that when you accomplish it and you evangelize it, it will be impactful, but also that the specific thing that you’re working on really benefit the institution.
Lee Milligan: 33:49 So, um, You know, we looked at some simple things to begin with. We’ve looked at things like length of stay and it turned out we had 13 definitions of length of stay, none of which were actually correlated well with each other and they were used without further defining them. And so people didn’t really know what they’re getting when they were looking at the information. And that goes Back to that data literacy that, that he points out. I fully agree with his assessment that, that needs to be a key component of this. Um, but people were looking at reports, looking at information and seeing descriptions, but not really knowing what it meant and then interpreting it and doing something with it. And so, You know, getting ahead of that and putting in place a system to, uh, to provide literacy to folks who are using these reports is a key component of it.
Bill Russell: 34:37 Yeah, it’s, it’s, it’s interesting because that’s one of the places we started a lot of the things you’re echoing when we did our, when we did our data management and data governance, a solution at the health system I was at, the, one of the things we created was a metadata repository where we actually went through our data and one of the hardest things we had to do was to say, what does this data element actually tell it? What is it actually? And length of stay was a, was a great one. I mean it ended up being a, you know, multiple meeting conversation to get to an agreement on what length of stay was. And it’s uh, and, and, and there’s, there’s obviously, there’s a lot of others. All right, so second question. I’m sorry, go ahead.
Lee Milligan: 35:22 That particular case. So just to showcase how we actually went through this process. So we had 13 different definitions we identified and the data governor for that was our director of finance and she brought together the right data stewards and the right team to really work the problem to number one, assess current state and two identify what are the needs of the organization. She ultimately worked with that team to compile a list of five separate definitions. but they’re, they’re different for different reasons and they’re really defined now and they’re published and they’re transparent. So now when reporting goes out, it doesn’t just say length of stay, it’ll say length of stay a or length of stay b. And you can define that. So you actually know what that means. So a simple example, but I think really impactful in terms of interpreting the information.
Bill Russell: 36:17 Absolutely every, every role in healthcare and in healthcare at this point has some aspect of a data analyst role in it. And uh, how have you been able to build data literacy across an organization and from an outside looking in, it looks like a growing organization. So you’re constantly not only educating the people that are there, but you have new people coming in,
Lee Milligan: 36:41 you try to hit it from a lot of different angles. Um, I’ve written a number of internal articles for our organization that gets published with our newsletters to start the conversation. we’ve also educated and then asked our c suite and data governors to push back on folks, giving them information, as I said before, but our most recent efforts really revolve around an actual data literacy project. And within my team we have a principal trainer who is focused on analytics, which is a bit unique in this space and we’re kind of leveraging her abilities around, um, so I’ll call it supported self Service for the most part. But in addition to that, she’s been tasked with coming up with a program for data literacy throughout the organization. What I found in my initial evaluations of this is that information products are being delivered to folks that is really over their heads that when they look at it, the data is presented in such a way that unless folks had some specific training or skill set, they really are not probably fully understanding that. so we want to get ahead of that really quick and so she’s working on that right now. So we’re on the front end of that. I hope next year if we, if we talk again about this, I can give you a little more feedback in terms of how far we’ve moved the dial.
Bill Russell: 38:02 Yeah, definitely. We’ll definitely come back to that. so, Let’s get pragmatic on this. Uh, what are the areas that you believe can utilize data to have the greatest impact on the communities that we serve?
Lee Milligan: 38:13 Communities that we serve? Well, in our case, we’re a medium size health system and we’re not owned by a larger system and we feel like because we know the community really well, we can provide great service, but in order to do that we have to be financially solvent. And so surprisingly, I think that Would start out with some financial data. I would talk about cost accounting, um, for example, would talk about one of our projects right now is to focus on accounts receivable, rejections. So it turns out that when you get rejections that come into your big bucket of rejections of a payer claims, they have numbers on them, that correspond to the payer, but they don’t correspond to any kind of a general interpretation. And so folks end up working those problems kind of one at a time and it’s incredibly slow. So what we want to do is really have a, a mapped, um, evaluation of all the rejections that come in so that folks, when they come in, they immediately understand what the problem was and they could work to rectify it for that individual claim, but be put in place a process so it’s not rejected in the future.
Lee Milligan: 39:27 And that’s really the essence of data governance versus data cleaning, data governance, really about not just fixing the problem but putting in place the process. So it doesn’t happen again.
Bill Russell: 39:36 Yeah. And there’s, when putting together a data governance program, one of the things that, it’s not wrong to focus in on the financials first and the reason that I say that was a and you know, and we, we’re a faith based ministry and, and it was a ministry and, and, uh, first and foremost, and we were about the needs of the underserved, but at the end of the day, you’re trying to get a data governance program in place to improve quality, to decrease costs and all those other things. And you need to get it funded. Well, there’s a whole bunch of areas that if you do it well, you will free up a whole bunch of money that will fund this so that you can really grow the program and continue the program. so it’s not wrong to focus in on, on money originally, uh, on, on a program like this. Um, but, uh, so let’s, uh, give us one or two use cases to highlight what you guys have been able to do at Asante health, uh, with over, with your data over the last couple of last couple of years since you put this program in place.
Lee Milligan: 40:43 So one of the ones that comes to mind is again, a simple example, but our problem list, so it was a problem list is um, I think it’s used by the inpatient docs and the ambulatory docs, it’s commonly used, but it’s commonly used incorrectly. So it ends up happening is folks are seen by their doc and they get this ever increasing problem list that becomes increasingly inaccurate and it’s one of the strange aspects of having an ehr versus a paper record with a paper record. There was this, I’ll call it a, a natural attrition that happened with the problem list. The doc would have to actually write the thing again and it, unless it was worth it, it wouldn’t actually write it. So stuff that was old or irrelevant wouldn’t make it onto the newest problem list. But in the ehr world it just gets pushed forward and it’s not really a lot of standards about what actually constitutes a problem and when a problem should be resolved or removed.
Lee Milligan: 41:44 and the reason why this is so important is because a lot of our registries are based on the problem list. It asked really simple questions like, is the patient alive or dead? And if they’re alive, is x problem on a problem level, if the answer’s yes, they go into a bucket on a registry. So getting that wrong has a lot of implications, but we put together a team of about a year ago, uh, and it has six ambulatory docs and six inpatient docs coming together to work the problem and they’ve met every month since then and they’ve put in place policy and procedure and they’ve evangelized change around what qualifies as a problem and what doesn’t. And it’s really been terrific to watch that process move forward. So that’s, I would call that a win in our, in our box.
Bill Russell: 42:27 No, that’s great. Well, we will definitely, uh, definitely come back to you on this as the journey continues. Uh, I, I do want to close out, I’ll close out with this story. Bill gates wrote a nice little note about his time with Paul Allen and in this, just to give you guys an idea of who Paul Allen, paul this, for bill gates’s words, paul foresaw that computers would change the world even in high school before any of us knew that personal, what a personal computer was. He was Predicting that computer chips would get super powerful and would eventually give rise to a whole new industry. That insight of his was the cornerstone of everything we did together. In fact, microsoft would never have happened without paul in december, 1974. He and I were both living in boston area. He was working. I was going to college one day.
Bill Russell: 43:18 he came and got me insisting. I rush over to a nearby new stand with him. When we arrived, he showed me the cover of the january issue of popular electronics it featuring a new computer called the altair 8,800, which ran a powerful new chip. Paul looked at me and said, this is happening without us. That moment mark the end of my college career and the beginning of our new company, microsoft, it happened because of Paul Allen and he has done the same thing in a lot of other, not only in the computer world but with philanthropy and some other things. So I just wanted to give him a shout out, uh, when people make that kind of contribution, uh, especially in the world of health it, we wouldn’t even have a lot of the stuff we’re talking about today without the work that those guys did early on. So, Lee Thank you again for coming on the show. What’s the best way for people to follow you? You seem to be everywhere these days.
Lee Milligan: 44:10 I’m on twitter. It’s @lee_MD_IT. Uh, and I’m on linkedin. You can just hit me up on linkedin and then my email is, is [email protected]
Bill Russell: 44:21 Uh, you can, uh, you can follow me on twitter @thepatientscio the show @thisweekinhit our website thisweekinhealthit.com. And the shortcuts of the youtube channel is thisweekinhealthit.com/video. Please come back every friday for more news, information and commentary from industry influencers.
The June 2018 Gallup survey of measures faith and trust in institutions shows that Medical Institutions has dropped 38% over a 40 year period. Plus we discuss putting healthcare data to work.