S1E17: Internal Customers – Actionable Data, Clinical Buy-in (ft. Mark Pierce, Parkview Health)

Share:

 

Dr. Mark Pierce, Chief BI and Analytics Officer at Parkview Health, tells how he uses data to drive clinical behavioral change.

Transcript:

0:0:0.0 –> 0:0:7.990
Jordan Cooper
We are here today with Doctor Mark Pierce, the Chief BI and Analytics officer at Parkview Health. Mark, thanks so much for joining us today.

0:0:8.370 –> 0:0:9.70
Mark Pierce
My pleasure.

0:0:9.760 –> 0:0:42.330
Jordan Cooper
So today we’ll discuss how Parkview has been leveraging analytics to track and manage operational and quality of care metrics, including resource utilization, hospital bed staffing and the tracking of key patient data such as length of stay and ventilator use. We’ll start with a concrete example of your work. Mark, you’ve developed a deterioration index which uses predictive analytics to alert providers if a patient is at higher risk for a serious health event, prompting them to take preemptive measures. Can you use the example of the index to walk us through how Parkview Health is leveraging analytics to meet key?

0:0:42.390 –> 0:0:43.830
Jordan Cooper
Clinical and business objectives.

0:0:44.430 –> 0:1:14.920
Mark Pierce
Yeah, absolutely. So, Jordan, you know the the deterioration index was originally designed just as you said, right? So it’s an early warning system to let us know when someone’s about to and again to use a clinical term crash. And there are several of these early warning systems out there. But one of the cool things that we did at Parkview is we actually quickly adapted it and leveraged it with oxygen requirement in our COVID patients. So here’s a concrete example, right, analytics is analytics.

0:1:15.20 –> 0:1:47.230
Mark Pierce
But analytics and innovation too, right? So we were able to combine a high DI score with an Fio 2 an increasing oxygen requirement to be able to triage critical resources at the time of our pandemic. So that’s beds, it’s ventilators, it’s oxygen tanks, it’s all that stuff to understand who are the sickest of sick and most importantly who might then become the sickest of six. So that we could be prepared and ahead of them. And again that’s all about managing resources, it’s operations, it’s clinical.

0:1:47.350 –> 0:1:56.540
Mark Pierce
It’s everything sort of wrapped up into one there, and the deis been critical even in times of pandemic, when you’ve needed to innovate a little bit in using him.

0:1:57.220 –> 0:2:27.330
Jordan Cooper
So I’m interested. Many CIO’s may be listening to this episode, maybe thinking ohh. You know, we’ve had problems triaging or we may have had workflow redesign processes. Can you speak about these specific political or operational situation that led to the creation of this DI? And while you’re at it, I’d love if you could throw in some of your inspiration. I’m sure you didn’t just dream up a quality metric out of thin air. You may have built in an QF measure. So.

0:2:27.440 –> 0:2:28.420
Jordan Cooper
How did it come to be?

0:2:29.110 –> 0:2:36.420
Mark Pierce
Yeah. So that the how it came to be is probably a little bit more in depth, but it’s actually probably something that a lot of CIOs might want to listen to.

0:2:51.370 –> 0:2:51.750
Jordan Cooper
Umm.

0:2:37.800 –> 0:3:6.650
Mark Pierce
You know, analytics can be as complex or as simple as you make it right? And you know, I guess it was, gosh, I think it was Einstein, really. That said, everything should be made as simple as possible, but not simpler. Right? And so when you’re talking about the inspiration for the DI, it revolves around what we’ve struggled with and that is is the KPI’s here to stay or as it’s time come and gone and I’ll get back to COVID. And just a second.

0:3:6.810 –> 0:3:38.40
Mark Pierce
But but if you indulge me for just a minute, simple KPIs, boy, they’re just so hard to deal with these days because Healthcare is a complex adaptive system, right? You push into the left and it pulls out to the right. And in all these systems, there are agents, they act independently and behaviors emerge and it’s you just can’t keep your eye on one thing because Ala Waccamaw something else pops up. So what we’ve come to understand is that KPIs are dangerous.

0:3:38.140 –> 0:4:7.890
Mark Pierce
In another themselves, and they’re enlightening and freeing. On the other hand, and so we’ve tried to have a little bit of organization around the complexity. So for a KPI, we try and keep it around three things, right. So for example, you mentioned at the opening length of stay, right? So if we start with length of stay, we want to keep our eye on that metric. But you also have to look at what your two week Ed utilization after discharge, what your readmission rate.

0:4:8.630 –> 0:4:39.400
Mark Pierce
And we keep those around threes to help us have some control out of the chaos and the deep dive that analytics can become. So let’s go back to where we were with COVID at the time, right. And so operationally, we’re running short on ventilators. We’re running short on beds and we’re running short on respiratory therapists, yet another three for you, right? And so how can you best utilize resources? What is the, what’s the KPI you need? Well, we know who’s sick now.

0:4:39.690 –> 0:5:10.570
Mark Pierce
And by sick, it’s what God, who needs a ventilator. But the real question is, are we gonna be able to meet capacity the next shift, the next day, the next week? And that’s really the motivation behind how we leverage DI and a little bit different way. But what’s more important than that one example is really the mindset around it. And that is you really in these days of analytics should be thinking holistically and at the same time, it’s a trap that you can get into where you get too detailed and your dashboards become.

0:5:11.270 –> 0:5:25.810
Mark Pierce
You know myriad of numbers that are tough to make sense of. So we try and keep it a little bit simple and we found that the number of three seems to pop up all the time, whether we’re looking at operational, financial or even clinical metrics that seem to come around.

0:5:26.380 –> 0:5:34.480
Jordan Cooper
So now that the big crisis of COVID has passed, are you finding that you’re continuing to use DI on a consistent basis or has it waned?

0:5:35.100 –> 0:6:4.870
Mark Pierce
No, it it’s still a daily part of our of our triage mechanisms, right? So obviously at least for us, I should say the need for beds is still there. We’re still operating close to capacity. So how do you understand how you’re going to meet that operationally in the next shift in the next day and the next weekend, right? So did I can help us understand how our ICU bed utilization is and what might be brewing out there on the floors for the next 3-4?

0:6:4.950 –> 0:6:12.90
Mark Pierce
5 shifts that are out there and those deltas in the DI can be just as informative as the absolute value.

0:6:13.180 –> 0:6:13.600
Jordan Cooper
So.

0:6:19.670 –> 0:6:20.10
Mark Pierce
Sure.

0:6:14.980 –> 0:6:44.40
Jordan Cooper
Let’s see here. I’d like to take it to a lighter note for a minute and then we’ll get back right into the substance. So just for the benefit of you or listeners and frankly, me, I’m ready to have some fun. So Mark, you’re also a freelance sports photographer, which I imagine requires you to find just the right angle framing and shot. We’re having to manage environmental forces forces out of your control, like lighting, weather and athletes movement. How do you translate your efforts behind the camera to your role and responsibilities at Parkview?

0:6:44.690 –> 0:7:12.570
Mark Pierce
Oh my goodness. Wow. So great question and you’re you’re putting left brained and right brain things together and what I’ll tell you is you’ve actually already done the secret sauce and that is if you look at a lot of the analytics together and you take it from an artistic standpoint, understanding the end goal of what you want your customer, those who use your artifact, what do you want them to think and feel?

0:7:13.220 –> 0:7:43.790
Mark Pierce
That sort of right brained exercise of framing angle, how much light you want is directly applicable to the artifacts that we make, and one of the things that my team grows tired of me saying is I want rapid at a glance cognition. When you look at an artifact that our team produces, it should slap you across the face like a dead fish, right? What in the world should I be doing next? If you have to squint, shrug your shoulders, and sort of lean into the keyboard or lean into the screen a little bit, we’ve missed our mark.

0:7:53.610 –> 0:7:54.100
Jordan Cooper
Mm-hmm.

0:7:43.930 –> 0:8:13.860
Mark Pierce
So that one frame that that one catch with the eyes on the ball and the fingertips or that that one dunk with the gorilla face as they’re just going to pound it down, that says 1000 things you feel, you know, it’s all right there and our artifacts need to be just as emotive so that you know what to do next. We hear the term actionable analytics and I would say that analytics at at most provide insights.

0:8:14.470 –> 0:8:32.370
Mark Pierce
But you have to understand those insights in terms of possible actions or not. Otherwise you end up with that last mile effect where you have great insights and everybody kind of looks and says, oh cool numbers, bro, and nobody changes their behavior and no values on leased from them. It very applicable funny bring that up.

0:8:32.800 –> 0:9:3.340
Jordan Cooper
Yeah. So, so it sounds like there’s a lot of benefits to kind of pulling together left and right brain. And I think as I’ve listened to many interviews, as I’ve engaged in many interviews, many episodes of Healthy Data, Podcast A founded general trend towards consumerism or keeping understanding the reason why we’re doing it and trying to make it as user friendly as possible. Now those consumers may not be external patients.

0:9:3.500 –> 0:9:25.180
Jordan Cooper
Right. It may be an internal executive board. It may be the chief of the Department of XYZ, but ultimately you’re serving your customers when you mention that your analytics dashboard needs to be emotive and needs a slap you across the face. I’m kind of getting a sense of almost a consumer patient engaging, but it’s customer engaging product.

0:9:26.250 –> 0:9:39.10
Mark Pierce
You’re exactly right. And and when we give analytics to folks, our hope is that those insights directly prime behavior changes that release value and there’s.

0:9:39.680 –> 0:9:53.750
Mark Pierce
You know, I spend my day taking data and arriving at insights, but ultimately the organization continues to invest in in analytics because of the value it can create. So Jordan, I think there’s a benefit.

0:9:54.910 –> 0:10:25.500
Mark Pierce
Role-playing is probably the wrong word, but really understanding the point, the value what you want to get out of it and absolutely from a customer standpoint, when we’re looking at patients or the community, we want to keep that in mind, but also our internal customers too. So where are we culturally? Are there certain changes in behavior or supply or physician behavior, dare I say or staffing that we are or aren’t ready to make at this point in time?

0:10:25.710 –> 0:10:48.590
Mark Pierce
And understanding that end value creation, I think is a critical part of creating effective analytics, whether you’re using an early warning tool like DI or a sepsis early warning system or even just basically giving routine operational metrics, understanding that behavior change that releases value from these artifacts I think is critical.

0:10:49.660 –> 0:10:53.500
Mark Pierce
Really, to make effective uses of the analytic resources, then organization has.

0:10:54.130 –> 0:11:8.900
Jordan Cooper
So when you come across, when you when, when you analyze your data and I guess that’s sometimes leads you to a conclusion about what is happening or what should be happening, how do you translate that into a behavioral change?

0:11:9.660 –> 0:11:39.810
Mark Pierce
Right. So you know, there are a lot of simple models that we try and use to kind of figure that out, right. So the first thing we asked when we get an interesting nugget is where are we focusing and we keep it simple. Are we focusing on input patients that go into the process, are we focusing on the process itself or are we focusing on the output? So that gives us a little bit of a lens into what we’re dealing with and then we further slice it by saying, you know, who’s really aligned with this behavior change is that the patient.

0:11:40.160 –> 0:12:12.750
Mark Pierce
The provider or the payer, or the grand panacea are all three aligned, and with that you can get a little bit of an idea of the mountain you’re trying to climb, because really the essence of unleashing new value with new insights is behavior change and it’s hard work on a good day. Behavior change is hard work, whether you’re a physician and examiner with a patient, or whether you’re an operational leader trying to change workflows from people who have done it this way for 20 years and they do it without thinking.

0:12:13.70 –> 0:12:43.610
Mark Pierce
So there’s a lot that goes into it. There’s a cultural balance, but the key piece is that relationship that you have with operations, the bridge between, hey, here’s what we’re seeing. Here are a couple of ideas of changes that might unleash them value. What do you think is it possible? Are there better ways to do this? How would they argue or how would they potentially fight the data? Are there errors in this or are there gotchas we call them? Yeah, that’s right. And physicians are famous for this. Yeah.

0:12:45.400 –> 0:12:47.60
Jordan Cooper
Yeah, yeah.

0:12:43.690 –> 0:13:9.540
Mark Pierce
That my patients are sicker. Yeah, but my patients are this. Yeah, but this right. So what are the potential pieces that we can assemble before we ask for behavior change? Because, Jordan, that really I think for any leader CIO, up, down the chain, that’s really the challenge is how do you get effective and sustained behavior change to release value. And again, analytics can prime that, but there’s a lot more that goes into it.

0:13:10.0 –> 0:13:39.740
Jordan Cooper
So Mark, we mentioned I mentioned the very beginning of episode. I introduced you that you’re actually a physician and when you’re thinking about behavior change, some CIOs may think if they are not providers, that there’s a lot of cultural resistance to providers. Would you just mentioned they’ve been doing the same workflow the same way for decades. They learned this way and residency it’s been working. Why mess with success? And yet you have this data that suggesting there should be a change. I wonder if you could address this problem.

0:13:39.940 –> 0:13:41.360
Jordan Cooper
As a provider.

0:13:42.560 –> 0:14:3.950
Jordan Cooper
What sort of perspective it does your clinical background provide you that helps you convey the the message of your analytics dashboards to providers in order to lead to their change and kind of what legitimacy do you bring to the table that that your, your colleagues without a medical degree may may wanna.

0:14:14.920 –> 0:14:15.510
Jordan Cooper
Uh.

0:14:5.130 –> 0:14:15.960
Mark Pierce
Being question right, so So what legitimacy do I bring? I you know, it depends on who you ask. I proposed, you know, I proposed to your right, depending on who I’ve slipped a 52, you’ll get a, you know, a different answer one way or another.

0:14:16.640 –> 0:14:21.30
Mark Pierce
And so I think just like in the old days.

0:14:29.60 –> 0:14:29.470
Jordan Cooper
Umm.

0:14:22.330 –> 0:14:52.800
Mark Pierce
CIOs would have a physician or a clinician when everybody who’s going live on Emrs, right. And you wanted that clinical validation. And I would argue that in today’s world, you need high level clinicians on your analytics staff. The reason why is when you start the conversation with. Yeah, but I understand that you believe that your patients are sicker than the average. Here is the prospective risk score on average of your patients compared to your colleagues patients.

0:14:53.180 –> 0:15:23.120
Mark Pierce
Here is their outcome X in your outcome X. Can you help us understand why and to be able to to to preempt some of the questions so that yet you get beyond trust in the data quickly to the issue at hand is important and what I’ll tell you Jordan, is is having these courageous conversations isn’t just somebody that you pull a doc from clinical practice 1/2 day a week and say, hey, I want to hire you to have these.

0:15:23.250 –> 0:15:53.520
Mark Pierce
It’s actually something that Parkview that we’ve invested in and by that I mean pulled folks out practice these given a separate skill set to in order to be more effective of it. And I don’t know whether you intended to or not, but you actually hit the critical thread of analytics and getting from the from insight to value and behavior change. These are the conversations that will either make an analytics program successful.

0:15:53.820 –> 0:16:24.170
Mark Pierce
Or not. And it’s very simply the ability to have the conversation, to get to behavior change. And quite honestly, as a clinician and again, I still practice a day a week. So folks see me in the hospital, folks see referrals from me, they see my patients, that piece of understanding, hey, this data is not perfect, but here’s where everybody else is. Maybe we’re off by a systematic error, maybe we’re not. But look, here’s where you are and everybody else is a little bit higher. How can we help you get there?

0:16:24.370 –> 0:16:42.680
Mark Pierce
It’s a much different story than perhaps somebody non clinically trained coming in and saying, well, here’s your numbers. Why aren’t you where your colleagues are O it does mean a lot. And again, any analytics program without understanding you need behavior change to release value is always gonna fall a little bit more short of where it could be.

0:16:43.690 –> 0:16:44.790
Jordan Cooper
So mark.

0:16:46.320 –> 0:16:50.300
Jordan Cooper
We’ve talked about behavior change we’ve talked about.

0:16:50.390 –> 0:17:6.970
Jordan Cooper
A kind of having difficult conversations and helping people trust the data, and once they trust the quality of the data and they trust that what you’re presenting them is actually real, then we can talk about how they can address the story that the data is conveyed.

0:17:8.370 –> 0:17:34.120
Jordan Cooper
At the onset of the pandemic, you launch a Community needs assessment that I presume was designed to tell a story. We have a few minutes left of the podcast, but I’m wondering if you’d like to tell us about that story and maybe to our CIO listeners or considering it. Why would that be important? And or perhaps more broadly, you may wish to take this conversation and just generally the power of storytelling, so I’ll leave it to you to see how you’d like to take this conversation.

0:17:34.660 –> 0:17:35.770
Mark Pierce
Sure, absolutely.

0:17:42.300 –> 0:17:42.690
Jordan Cooper
Mm-hmm.

0:17:37.310 –> 0:18:8.560
Mark Pierce
To kind of put a bow on this, maybe Jordan. So we’re we’re not-for-profit and as as we all know, right, not-for-profit doesn’t mean not for margin cuz right without a margin there, there’s really not a story, but we take that seriously. We’re kind of Midwestern polite and when we look at our community and and do a Community needs assessment that helps us align our focus, our strategies and our operations for the coming years.

0:18:8.630 –> 0:18:38.540
Mark Pierce
Because our community thrives together. When when the tide goes up, all boats rise. And so to the extent that we take it personally to improve the health of our organization, of our community, we’re benefiting the entire Community and that’s why we exist. So when you look at the Community needs assessment, we have a lot of needs around prediabetes. We have a lot of needs around obesity. And so when you look at some of the cool analytic things we’re doing.

0:18:38.880 –> 0:18:52.560
Mark Pierce
At least in in my neck of the woods, we really focus on early detection, on intervention and on matching effective interventions with the right people on whom it will be.

0:18:53.280 –> 0:19:24.230
Mark Pierce
Most effective, and that’s critical for a couple of reasons. Not only does it improve the overall health of the community, but look every resource we have is on loan to us from the Community it it provides resources for us and to be good stewards of that as part of the key pieces of who we are at Parkview. So that community needs assessment guides what we do and then as you drill it down into my neck of the woods, it guides the analytics, it guides what’s important to local employers.

0:19:24.600 –> 0:19:34.870
Mark Pierce
In terms of keeping their their coworkers healthy and productive and happy, and so it all fits together for us in an overarching strategy to better take care of our community.

0:19:35.950 –> 0:19:49.440
Jordan Cooper
Well, thank you, Mark. We have reached the end of this episode. So like their mind, our listeners at this has been Doctor Mark Pierce, the chief business intelligence and Analytics officer at Parkview Health. Mark, I’d like to thank you very much for joining us today.

0:19:49.780 –> 0:19:50.560
Mark Pierce
My pleasure, Jordan.