[creativ_pullleft colour=”light-gray” colour_custom=”” text=”Episode 046″]
Breaking Health Host Steve Krupa hears some exciting news from Optum’s Michael Weintraub and talks Analytics with Mudit Garg, founder and CEO of analyticsMD.
Michael is Co-Founder and CEO of Humedica, acquired by UnitedHealth Group (UHG) in 2013. Michael has broadened his role within UHG as CEO of Optum Analytics. Prior to launching Humedica, Michael served as Senior Managing Director at Leerink Partners, a leading health care investment bank.
Mudit Garg serves as CEO and Founder of analyticsMD, one of the leading prescriptive analytics firms applying a software based “Air Traffic Control” for hospitals.
Tom Salemi: Hey, everybody, welcome back to the Breaking Health Podcast, post election edition. Steve Krupa and I are – this is Tom Salemi, your introduction guy. I’m here with our host, Steve Krupa of the Psilos Group. Hello, Steve.
Steve Krupa: Hi, Tom, how are you?
TS: I’m recovering, thank you for asking.
SK: You and everybody else in the Northeast.
TS: Yeah. I feel like we need to have to have another Digital Healthcare Innovation Summit because I believe everyone in the room probably didn’t think this was going to happen, so the tone of our discussions in some regard was sort of looking toward, I think, a Hillary administration or a Clinton administration. And it’s not going to be that. So interesting and changing times ahead, which is that a good thing for a VC or a bad thing? Because there’s going to be disruption, but there’s also just great uncertainty.
SK: Yeah. I’m not so sure it’s a good thing. I think we had a set of expectations about where a Democratic president would take sort of the government portion of healthcare, which is big, bigger than 50% including the ACA, which the president-elect has vowed to appeal and replace. And so it adds uncertainty in the area of healthcare across the board. We don’t know who’s going to run HHS, we don’t know what their ideologies are going to be. But I think we’ll find out pretty quickly where all that stuff starts to shake out in the first couple months of the new presidency.
TS: It’s going to happen quick, too, with the Republican Congress. They’ve got two years until midterms, so I’m sure they’ll do whatever the can right away.
SK: They will. I mean I don’t know, people probably have a rooting interest in one direction or another on that issue, many issues. I think there’s a filibuster is maintained for the Democrats, so they’ll still be in on any negotiations.
TS: Yeah. Although I think they’ve got the nuclear option, right? I mean we don’t want to stray too far. Well, this is a different Podcast. Let’s go back to last week, happier days when we were at the Digital Healthcare Innovation Summit in Boston. You played many roles. You did the onstage interview with Dan Burton, which we played last week. This week we’re going to dip into a few private interviews that you had with a couple of the panelists there. Both analytics, both very different companies. The first one we’ll tee up is Michael Weintraub, who’s the CEO of Optum Analytics, or has been until of late. He’s sort of transitioning into another role, and we’ll let him talk about that in the Podcast. But I haven’t heard these yet, so I can’t even pretend that I’ve heard them.
TS: What was your takeaway from your conversation with Michael?
SK: Well, you know, Michael has spent a lot of time in the analytics space in healthcare. And he’s experienced the business from startup to, as he would describe it, scale while he was at Optum Analytics. And I think he provides some interesting insight into the value of the analytics companies across that space. And he also gives us a little bit of information about what his next journey is going to be. And I think maybe I’ll just leave it at that and let the interview speak for itself.
TS: I like it being cryptic. And Optum, of course, is one of our very special sponsors of the conference, we we’re very happy to have their support. And then following the break, you had a conversation with Mudit Garg. He’s the CEO of analyticsMD. What was your takeaway from that conversation?
SK: Yeah. So back to back analytics discussions. analyticsMD is very early – well, I wouldn’t say very early stage, but probably much earlier stage than Optum, of course, and new to the analytics space over the last couple of years. But they are beginning to use machine learning across data sets from hospital operations to try to do predictive analytics and predictive modeling around operational efficiency inside a hospital. So anybody that goes into a hospital, you can’t help but marvel at sort of the massive number of employees that are doing all sorts of stuff in any given moment. And one of the tricks to being competitive is figuring out how to staff and organize those activities. And so analyticsMD is trying to provide data solutions to guide managers on how to run various departments inside a hospital. And it’s very interesting. He’s a young guy, very bright, and I think we did a couple of use cases on this product in the conversation.
TS: Excellent. And we will have both of these interviews in video form, too. These were done on camera, so those will be coming out in the coming weeks. We’ll have more details about that when I do the break in between the two interviews. So all right, let’s roll these conversations.
SK: Michael Weintraub, welcome.
Michael Weintraub: Thank you. It’s great to be here.
SK: Yeah. We’ve got a great story to talk about in terms of the analytics space. So you were – I noticed out at the conference a lot of focus on analytics. I’ve always gotta ask people what the heck is the analytics doing because baseball fan, cyber metrics, all that cool stuff, love the analytical side of things. But you really were at the beginning of this trend at Humedica. So I’m going to ask you to try to remember what you were thinking when you started to do this back when you started that company.
MW: What was I thinking? You know, analytics is a little bit like oxygen. Everyone’s got it, everyone needs it. Everyone thinks they own it. And I’ve been asked this for about 35 years, but really, it’s really only come of age in the past decade. And what we did with Humedica back in 2008 is really took a big gamble. You know, with entrepreneurship, it’s a little bit about opportunity, a little bit about risk. Otherwise put, they say, there’s a fine line between bravery and stupidity. And when Obama was still senator, when EMRs were not really where they are today, we had a hunch, myself and my two cofounders, AG Breitenstein and Allen Kamer, that there was going to come a time when healthcare data would be looked at the same way other industries look at their data, like financial services and so forth. And so we really took the gamble, raised capital in 2008 and built the first at scale manufacturing process to study clinical, financial and operational data, and really put together a longitudinal, cross-continuum view of the journey of a patient through the system, inpatient care, ambulatory care, the pharmacy, the medication, and so forth. And really to do that at a time when most systems had a handful or two or more EMR’s –
MW: – lots of different claims, data feeds, lots of different operational systems was complex. But we built the system in 2008-2009. We launched quickly, and raised significant capital at the time. Today we’re seeing stories of a lot more capital.
SK: can you tell us how much you raised?
MW: We raised $63 million. We raised 30 million in a series A to build, and 33 million in a series B to deploy at scale. And so we moved pretty quickly. The vision we had was to really allow a large academic medical center, integrated delivery network, ACO, large multispecialty medical group to really understand their manufacturing process, their inputs, their processes, and their outputs.
MW: And we grew fairly quickly. And we also then took that data and de-identified it, and utilized that to build predictive models and disease models to allow you to really start doing what we call today is predictive analytics, prescriptive analytics. So we were really the first to do it. We’ve grown quickly. And we also then took that data and utilized it in other industries such as life sciences. Totally de-identified. We call it HIPAA plus, not just PHI in HIPPA, but also this is about large scans of universe, not provider level information, not physician level information, of course. And so the markets are all blending: payers, providers, large self-insured employers. Life science companies are all trying to get their arms around the consuming patient. So grew quickly and then in 2013 we, after 4 very quick years, we sold the business.
MW: To Optum.
SK: Right. A traditional acquirer of businesses like this.
SK: And you got to run their analytics business, basically, from that point forward.
MW: Yeah. We were not really for sale. We were growing very quickly, and it was much more of a pre-emptive strategic conversation. Optum is very significant out there with their scale and their ecosystem. They’re $80 billion today. They were 30 billion when I got to know them. And they had a legacy of 20 years of really powerful analytics with claims data.
MW: And they were looking for what I call the Manhattan Project. They were looking for that clinical atom, that capability and the team. And so we joined Optum in January 17th, to be exact, of 2013.
SK: Those are the kind of days you always remember.
MW: You don’t forget those. 3:30 PM. And we made a decision together to do a bit of a reverse integration. We took the population health, big data analytics assets of Optum, married it with our assets under the leadership team of Humedica with the addition of other Optum executives. And it’s been a fast four years. Four years in a couple months. And so it’s grown very rapidly. The business was about 100 people when we started. Today it’s approaching 1000 people. And we’re in all 50 states. We’re quickly approaching touching almost 100 million patients with our data. And it’s been quite a ride. It’s been going back to school for me in a great way. You always want to contribute but learn, and learning about the scale has been phenomenal. My – I like to say my average survival index, because I’ve been involved in 6 startups, is about 18 months. And so I surprised myself staying as long as I did. But to be able to innovate at scale and to take innovation and have the impact on the system is phenomenal. And Optum’s scale and breadth and depth of access is tremendous.
SK: So very cool. I gotta ask you, though. It’s a whole different ball game, right, when you get to be your own boss, when you’re running your own company, and you get a acquired by what is a very large company. I can’t even keep track of how big Optum is, it’s so big. Was it fun to stay? What made you want to stay?
SK: Was it sort of like, Oh, I don’t have to worry about raising money anymore?
MW: Yeah, you know –
SK: I don’t have to worry about all these things you have to do in a startup.
MW: Well, yeah, actually, I think raising money is exciting.
SK: Well, it should be exciting.
MW: When it works. And I had great investors between Northbridge, now Flare, and General Catalyst and Bain and Leerink and several other strategics in the B round. But there is a lot of energy and passion and emotion that goes to building a company. Quite a few of the people on the management team and below have been with me for over 30 years in multiple startups, so it’s a bit of a family. We know how we compliment each other. But you get to a point in your career after 35 years where you need to ask the question, what are you trying to accomplish? And when I started seeing the scale of Humedica as Optum Analytics grow from where it was to where it is today, and then you sort of step back and say, You’re actually impacting the healthcare system at scale, it’s pretty rewarding. And so you’re working with the who’s who in the healthcare system. You have tremendous access, and you’re actually touching more lives and more patients, and you’re growing faster. So for me it was a little bit about a different journey. Having flexed the startup muscle over 35 years, I actually did this with the intention of playing it out as opposed to flipping in and doing another one.
SK: Right. Most guys get out of there in about 6 months to a year, go do it again.
MW: Yeah. And the team stayed. The team stayed.
SK: So when you think about innovation at scale, and I love that term, when you think about what you’ve accomplished, give me a sense of what you think the great innovation that you were able to sort of bring out of this thousand-person team, where you had large data sets, obviously, to work on, large computing capabilities to draw from.
MW: So you know one of my cofounders and I balance each other. I like to say that what happened was 51% luck and 49% vision. And she kicks me under the table and says, Speak for yourself. I think that in many ways, what she’ll say, and I agree, is that we spawned an industry because there were a lot of point solutions at the time, right? But this wasn’t being done at scale. And the country and the healthcare system needs it to be done at scale. Since then, you’ve seen what’s happened in the proliferation of other companies. They’ve been born, they’ve been acquired. Some are still independent; some have IPO’d. And so we took a gamble before health reform, before Obamacare, before the Affordable Care Act, and it was a belief system we had that this was coming. So I think that we’ve contributed to the industry. We’ve got a phenomenally talented workforce. We’re building and innovating. We’re on a path to continued growth, and we’ve become a strategic asset within the Optum family, in a business that is approaching $100 billion, just Optum.
SK: I was searching for that number, 100 billion. I’m glad. Glad you backed me up
MW: It’s 80. And the 100 is not a projection. We’re a public company. But as I watch the growth being a data guy, I smell it coming. And it’s big. And so we tend to now have conversations across the country with organizations that are consolidating, that are integrating, that are having significant impact. And it’s really meaningful.
MW: And at some point, the journey’s gotta be meaningful, not just valuable.
SK: Right. I know you want to talk about what your future is, but I want to ask you sort of a question that I get asked a lot. And it goes something like this, so don’t take it personally. I get asked this question as well. It’s like Oh, you got all this computing, you got all this data, you got all this analytics. And how come healthcare costs keep going up? And I always say, Well, imagine what it would be like if we didn’t, right?
SK: So are we just sort of just catching an ever rising tide, and really not able to bring it down? So maybe we’ve had an impact in dampening it, but we don’t know it because it seems like there’s a regression to, say, 6, 8% medical inflation? Every now and then we have a good year, then we have a bad year.
MW: Yeah. So that’s a great question. There’s a lot of answers, and I’ll give you several. And I don’t think it’s any one in isolation. I think the first answer is we’ve actually just begun. The capabilities out there are not yet creating scalable return on investment the way they will in the next 5 to 10 years. That’s number one. This just is beginning. Number two, the move to value versus fee for service is being talked about a lot, but if you actually study the data and look at the percentage of the healthcare system that’s shifted to at-risk value based, outcomes based, it’s really small. That’s a ten to 20-year journey, not a 1 to 2-year journey. And so in fact, the majority, and I don’t mean 50%, I mean way above that, of health systems in this country are still fee for service.
SK: Yeah. If you take California and Southern Florida out of the equation –
SK: – it’s probably 90%.
MW: Correct. So I think the incentives have to get aligned to focus and harness and motivate. So that’s number two. Number three, you know, as our MIT health economist talked about this morning, one of the reasons is because this country can afford it. And so –
SK: We make all the mistakes we can afford to make.
MW: Well, I don’t mean it that way. I mean that some of the really expensive healthcare for a small percentage of the population is still in the system as a privilege and a choice.
MW: And so in – I have relatives in Canada. When certain things happen, they don’t have access. Here we have access. So I think that that’s not going away. And so do I think there’s a tipping point coming where the capability that we have, the data, the technology, incentive systems, the transition to value is going to actually have an impact? Yes. But I don’t think it’s in the next – I don’t think the acceleration is in the next couple years. I think it’s a decade plus journey. I do.
SK: Yeah. And if we’re lucky, we’ll still be doing this, right?
MW: We will definitely be doing it. I think there’s more happening in digital health now than in the last 35 years of my career. When I see these 25-year-old CEO entrepreneurs I’m like, Man I was too early. Because they have the world at their fingertips. There’s so much opportunity. And so the technology, the cloud computing, the information highway that’s been laid down, the cognitive computing, the AI, all of that is coming together in a way that wasn’t possible before that information was out there.
SK: So look, we explore this in Healthegy through our podcasts and through a lot of programming, through investing, and obviously taking a good sampling of the industry. And all we see is there’s a thousand plus companies out there that are doing things of interest. And so I know a little bit of a secret about what you’re up to, so why don’t you – you seem to have an interest in this area, so you’re about to make a little bit of a career shift.
MW: So what am I up to? So in an hour or two I’m going out to dinner. It’s my triplets’ 18th birthday. So that’s the first thing I’m doing. And then in about 9 months I’ll be an empty nester. And so between my 23-year-old who’s in digital health, and my triplets going off to college, I said to myself, what do I do now? And so what we’re doing is we brought on an individual to run the business I built over the past ten years. She started on October first, and it’s a phenomenal addition to the team. She comes out of the healthcare system, and I’m spending the next 3 months supporting the transition. Probably longer. And the conversation that I’ve had with Optum and United Health Group over the last 6 to 9 months has been to really grab hold of innovation and really participate actively in that ecosystem that’s being built out there. And so we are launching Optum Ventures, which and my cofounder and several other colleagues from Optum will lead, in addition to some outside folks we’ll be bringing on as partners. And the entire focus of Optum Ventures is to get back to what I know, which is entrepreneurism, innovation. But to do it with the scale of Optum and the support of Optum’s distribution channel, data assets, technologies and domain experience and capital. And so I’m pretty excited about it. It’s a way to both accelerate my participation in the innovation economy and also pay forward a bit and start supporting entrepreneurs more broadly in the next lap of my career.
SK: Great. That’s a great outcome for you. It sounds like it’s going to be –
MW: It’s awesome. Really exciting.
SK: So just two more questions while I’ve got you.
MW: Two more.
SK: Do you want to share with us the type of things that you’re interested in in ultimately partnering and investing with Optum Ventures, the types of ideas that get you turned on in terms of the future?
MW: So yes, but probably not yet. But we are putting forth an investment thesis that we’re working really hard on. You’re in the same space. And I want to take a little bit of time before I bring it to the marketplace. But we will do it.
SK: Well, when you’re ready to do that, will you share it with me?
MW: We will. Absolutely will.
SK: All right, well, it was great to meet with you.
MW: Thank you for your time.
SK: Hey, everyone, Tom here. I hope you enjoyed that conversation that Steve had with Michael Weintraub from Optum Analytics and da-da-da-da-da-da, Optum Ventures. It’s going to be great to have another strategic investor in the digital healthcare space. Now I’d like to introduce our next conversation. But before I do that, I wanted to tell you that we’re working on the content from the Digital Healthcare Innovation Summit. We’ll have some reports that we will compile from the Summit, some original interviews that we did, including the two that you’re hearing on this Podcast today, as well as many others that I did, and Stephanie Porcell from our staff did as well. And also wanted to tell you that it’ll be available on the Healthegy website. So go to healthegy.com, it’s the world health followed by the letters EGY.com. Keep checking that and we’ll be putting content up there. Or of course subscribe to the Breaking Health Newsletter. We’ll send it right to your inbox. You need to go to healthegy.com to sign up for that. We just need your email. And finally, we’re working on video presentations of the panels, the entire panels. If you’d like to see those, please do this one thing for me. Shoot me an email. Tom@healthegy.com. Let me know you’d like to see those presentations. We’re going to make them available to anyone who has attended the conference, but if you weren’t able to attend for whatever reason, but want to catch up on some of the sessions, shoot me an email and we’ll hook you up. So now let’s get back to this next conversation, rather with Mudit Garg, the CEO of analyticsMD. Mudit was attending the Digital Healthcare Innovation Summit, and we reached out to him because we have been very interested in the company and wanted to hear its story. So this is another sit down that Steve did with Mudit Garg at the Digital Healthcare Innovation Summit on November second in Boston.
SK: Mudit Garg, how are you?
Mudit Garg: I’m good, thanks, Steve for having me today.
SK: Yeah. The Digital Healthcare Innovation Conference, Summit. So good to have you here. Is this your first time here?
MG: My first time here. I’ve absolutely loved the day. In some ways, wish it was longer, though I know many of us wouldn’t be here if it was longer. But I’ve really enjoyed the day.
SK: We did a good job. We had the governor of Massachusetts here.
MG: Fantastic. He was – his insights – there’s very few people who probably understand the government and the healthcare as well as he does.
SK: He does. He seems to get it both ways. I don’t know if he’s going to run for president or not.
MG: I won’t be surprised. I would hope so.
SK: Yeah. So you’re running a company, analyticsMD.
MG: That’s right.
SK: And we both came into sort of knowing who one another were a couple years ago as you started to get going.
SK: We know what you’re up to from the title of the company, but talk to me about how you decided to do the crazy thing, which is –
MG: Start a business.
SK: – start a business and be responsible for paying yourself a salary.
MG: Absolutely. And many people may not – it was probably harder even on the family than it was on me to make this crazy decision.
SK: Yeah, yeah, the family is like, What?
MG: What are you doing?
MG: But you know, it was about probably a decade ago that I was working in a hospital. It was in rural California. I was shocked by the inefficiencies, the chaos, the usual stuff. But what really struck me the most at that time was the sense that we were failing the front lines, that they were being forced to surmount obstacles completely out of their control, day in and day out. And data was an extreme example of that, where we had just, last decade, asked people to enter data in to EMR systems and staffing systems and billing systems. But that data hasn’t come back to help anyone. At best, we’ve been getting dashboards and reports.
SK: It’s been good for the EHR technology providers.
MG: It’s been good for the EHR, it’s been good for the EHR. But as an end user, I’m getting bombarded by data and I don’t have time to look at it. And at best, if I look at it, it gives me a view into the rear-view mirror. So that is what inspired me to say there’s gotta be a different way. If you – that was the excitement on predictive analytics in the beginning.
MG: And the promise with the huge amount of data that we have now extracted and gotten the computing power, the algorithms. But the frustration still was that is not enough to shape the actions of the people we need to shape the actions of without adding work to them. That is what inspired me. And it wasn’t a straight linear path to filling it out.
SK: Yeah. No, no.
MG: It took several steps and several iterations along the way. But I think the right answer is now emerging, where we take cognitive load off of the end user while providing them insights that they can act on in the moment.
SK: So I look at a hospital and say this is like running like a small infantry, like an army. There’s people everywhere.
SK: There’s a lot of action going on. There isn’t often someone directly in command, although theoretically the doctors’ orders are being followed. And then there’s a bunch of data that gets collected. And it’s not clear to me what happens to that data. And it’s not clear to me that it ever gets organized in a form that it creates a lot of value until maybe somebody questions a clinical decision or somebody decides to sort of improve a process. So when you start to work with a customer, do you first assemble the data? Is that the first task at hand?
MG: Yeah. And actually one of the points I wanted to mention as you were saying this that came to mind is the other insight was it wasn’t as much the medicine that was behind the times. It’s everything else around it.
SK: The medicine is wonderful.
MG: Right. The doctor is the quarterback of the medicine. And they do –
SK: It’s the one-third that’s not medicine.
MG: Right, right. They do a great job of that. But how do you manage an orchestrate the rest of it so we can allow them to have the flexibility to do the best job they can? And so when we start with an entity, with a customer, actually we don’t start from data. We don’t start from predictions; we don’t start from algorithms. We start from decisions. What are the decisions that you want to have happen reliably, consistently that can prevent a bad situation from happening? And we’ve chosen to start with decisions that are traditionally more operationally, non-clinically. Because my belief is that if we can just take a lot of that burden off of these end users, they know the right medicine. They trained for ten plus years, every one of them –
SK: Yeah, yeah, yeah.
MG: – to make the best clinical decisions. But the rest of it is what we need to fix. So we really start with decisions. What decisions are we trying to make? Because every –
SK: Give me an example of a decision.
MG: So let me give you an example of a decision. In the chaos that you were just describing, the emergency room is a great microcosm of that chaos. If you ever walk into one, there are a few people who are in charge of saying I need to be the manager of making sure this does not devolve into a really bad situation. And they’re dealing with that day in, day out. What that means is you don’t want lines waiting out the door. You don’t want patients waiting, not being seen. That’s both bad risk, bad patient care, bad experience. One of the common decisions you are anticipating is am I going to run out of capacity. Am I just not going to have the capacity to treat these patients? Because that’s when the worst happens. Constantly thinking of that is very hard. Constantly paying attention to that is very hard. The decision is should I refocus certain specific resources in the moment. And they’ve had surge plans and things for a while. But how do you realize when is the right time to do it? So what our system will do is look at how cold is it outside, what’s the weather like, what is happening in the ER, Dr. Smith’s working, four patients out in the waiting room. All those factors. And if we have a high confidence that that is going to require now an action now, so that an hour from now a bad situation does not happen, then we will nudge the end user, the charge nurse, the house supervisor, the EVS lab, whoever is able to most affect the bottleneck that will prevent the bad situation from happening. And we have countless examples like that of what the system monitors in real time, ingests, learns from, and then nudges people to take action on.
SK: So is it the system by itself that’s doing the nudging? Is it the system in combination with people?
MG: The system does the nudging by itself. But it enables a social dialog with the operator to take the action. So you know, there’s a big debate in the world of intelligence about AI or IA.
MG: Artificial intelligence and intelligent augmentation. I’m a big believer that we need to augment.
SK: Well, AI is a cooler word.
MG: it’s a cooler word, absolutely.
SK: But it probably isn’t the right word for what you’re doing.
MG: I don’t think it’s the right word. We need to augment the capacity of people to take actions, to take the right actions without increasing the load. If I asked you to go straight to a self-driving car when you didn’t have a GPS that could figure out the route, when you didn’t have your cruise control –
MG: – and when you didn’t have the lane –
SK: I would get anxiety.
SK: I have anxiety just thinking about it.
MG: I still get anxiety. But –
SK: Yeah. I want to at least be able to watch where the car is driving me.
MG: Absolutely. So what I think we need to do first is let people keep their hands on the steering wheel, but do all the complexity behind the scenes, compute the traffic, compute the optimal route, compute where you should go, and just nudge you. I think now is the time to turn right. And let you make the decision. And then since the system is smart, it learns with what you choose to do. If you choose to go straight, hm, that actually worked out better. All right, I gotta make note of that and I am going to improve off that. That is, I think, what is going to make a sustainable way of using data to continuously improve outcomes. And that’s what we are hoping for.
SK: So I’m reading this book called How Not to be Wrong. Do you know the book I’m talking about?
MG: I don’t know, but I like the title.
SK: It’s basically try to make everything into a math problem, but understand what the math problem means. So you can take a computer, you can take data, right, you can do a whole bunch of interesting things, and you can come up with correlations and predictions that have no actual affiliation –
SK: – with real world activity, or no real world benefit. So how do we train computer models –
MG: To do that?
SK: – to sift through that and be logical in real world?
MG: And I think that’s a very good balance. And I’ll give you two examples on the opposite side of the coin for that. One is I always tell customers there is correlation between solar spots and the stock market.
MG: But would I invest my money into the stock market? No, based on solar spots? No, I wouldn’t.
SK: Based on solar spots. Well, you don’t know what you’re doing then.
MG: Right, exactly.
SK: There’s a lot of people making money on that strategy.
MG: So it is important in the universe of what we consider marries the human intuition, but doesn’t rely entirely on the human intuition. Because I’ll talk about the flip side of that. One of the big beliefs in any of the ER’s you go around the country is on full moon nights – and I had this. The first time we ever did a prediction and the nurse was telling me that you take the full moon effect into account, I didn’t know if he was pulling my leg or if he was being serious. He was being serious. So we had that as a feature. There is an article now. We actually just recently – there was a Wall Street Journal article on this, so we redid the math and rechecked it. Still has never been picked up as a real signal. But that’s the beauty of the human intuition, but has to be combined with rigorous statistical testing of if it actually pans out or not. We can’t do one without the other. We can’t let the algorithms train on anything and everything because they will pick up spurious correlations. But we also can’t only rely on rules based algorithms that are oftentimes only driven by our intuition, and very often an environment where the population is changing and underlying data is changing, are very fragile and break very quickly. So that is the balance of the two that I think we need to strike.
SK: Cool. So how’s your business doing? How long have you been in business now?
MG: We’ve been in business about – with the product we’re talking about, about two years. Last 12 months have been phenomenal. It all started from being able to sustainably drive outcomes. What changed, what sparked the growth was being able to show evidence that we are able to drive sustainable outcomes for our customers in a repeatable way without burning out the providers. And once that has happened that we’ve seen academic health systems, some of the top academic health systems all the way to community health systems in rural Arkansas and Ohio and Missouri, all the way to 40 bed critical access hospitals all be able to use this method of approaching data and approaching change, and have reliable outcomes. And that’s really changed for the good. And there’s plenty to solve going forward for us, but I believe if we keep that as a true north of how do we reliably drive outcomes through data, and how do we make data and influence our action, not just data for data’s sake. Then hopefully we’ll be able to continue that curve as well.
SK: Terrific. Last question for you. Sort of give us some advice on the life cycle of a startup and what the feeling is building a company and kind of get to sort of create your own world at some level. So what kind of world are you creating there?
MG: You know, I had a baby ten months ago.
SK: Oh, congratulations. That’s always a good time. You get to spend a lot of time in hospitals when you do that.
MG: That’s right. It was actually good because we delivered in a hospital where they use our product, and it was amazing to see the users. But I think in a lot of ways building a startup is like that. The joy is infinite. The high is very, very high, but it takes a lot from you as well.
SK: It does.
MG: And from a life cycle of a startup, one of the people I really respect told me it’s like a marathon, except that as you run it never ends, and as you run, you kind of just look around people falling down and dying. But the one who perseveres, has a clear, true north can win in the end. And if we learn and we stay humble, I love the point that Dan made. If we stay humble, we learn and we change along that way with keeping the core principles the same. Life cycle of the startup eventually leads to a much bigger outcome and certainly more satisfying for me personally. But there’s a lot of hard work along the way.
SK: Yeah, yeah. And you’ve gotta find people to do it with you.
MG: Absolutely. It cannot be done by myself. The number one lesson is for us as a company is intellectual honesty across the organization. Everything is transparent across the organization. Problems are talked about immediately as soon as they’re discovered. They’re not – because if that is the case, we talked about last night, it’s not me solving the problem anymore. I get amazing ideas from the entire team, everyone’s collective minds thinking about it. And we have the smartest people, so why shouldn’t we use their collective minds for every problem I face? My cognitive load, as I was just talking about, is a lot lower in that scenario and likelier of getting the right answers is way higher. So that’s probably the biggest and most important piece.
SK: OK. So how do people find out about your company? Website? Twitter account?
MG: Website certainly. Info@analyticsMD.com , mudit@analyticsMD.com . You’re welcome to reach out to me. The website’s a great place to see about us. Our customers are great to talk to as well. The website’s probably the best starting point.
SK: Cool. Nice to meet you.
MG: Very nice meeting you again, Steve. Thank you.
SK: Yeah, thank you.
TS: All right, folks, that is a wrap of this very special Breaking Health Podcast. I hope you enjoyed these two conversations. Steve Krupa, great job. Thank you to Mike and Mudit for making yourselves available during the conference, and thanks of course to our Breaking Health listeners for joining us. We’ll go back to our normal format next week. And don’t forget, if you want to receive video presentations of the panel discussions that we had at the Digital Healthcare Innovation Summit, shoot me an email at firstname.lastname@example.org . If you want to get the content sent direct to your inbox, these are our original interviews and reports from the conference, you’ll need to get the Breaking Health Newsletter. So go to healthegy.com and sign up for that. And also, while I’m asking favors or offering you things, actually, I haven’t asked you a favor yet. I will do this one favor. If you enjoy this Podcast, we’d love to know about it. Go on iTunes or any of the formats that you use to listen to this Podcast and just give us a quick rating, and please take a minute and just write a quick comment about the Podcast. We’re always trying to make this better and would very much like to hear what you have to say. So thanks again for joining us today, and tune in next week for another tale of innovation on the Breaking Health Podcast.