Mayo Clinic This Week in Health IT
April 24, 2020

 – Episode #

Guest Information

Share this clip:

Share on linkedin
Share on twitter
Share on facebook
Share on email

April 24, 2020: In today’s episode of This Week in Health IT News, we speak with Dr. John Halamka, president of the Mayo Clinic Platform, about the national COVID-19 Healthcare Coalition – a private-sector led response that brings together healthcare organizations, technology firms, nonprofits, academia, and startups to preserve the healthcare delivery system and help protect U.S. populations. We also discuss data-driven outcomes and digital tools like GPS-contact tracing and novel apps, as well as the challenges of potentially dealing with sensitive identified data and how to protect it. John goes on to discuss the importance of platforms in getting rid of misinformation, analyzing data, and sharing knowledge. For all this and some links to free, open-source resources on the COVID-19 pandemic, make sure to listen in! 

Key Points From This Episode:

  • John explains the work he has been doing with the national COVID-19 Healthcare Coalition.
  • How to gather comprehensive data sets from disparate sources into a central repository.
  • Using a federated analytic approach that gets beyond data use agreements to create a registry.
  • Creating and updating automated systems to comply with a FHIR situational awareness standard.
  • John explores some digital tools that have been developed regarding contact tracing.
  • The role that apps and platforms need to play in dispelling misinformation about COVID-19.
  • Mayo Clinic has put mechanisms in place to quickly de-identify data, centralize, and analyze it.
  • How much and what type of testing do we need moving forward?

Field Report: Mayo Clinic with John Halamka, MD

Want to tune in on your favorite listening platform? Don't forget to subscribe!

Thank You to Our Show Sponsors

Related Content

Amplify great thinking to propel healthcare forward and raise up the next generation of health leaders.

© Copyright 2021 Health Lyrics All rights reserved

Field Report: Mayo Clinic with John Halamk

Episode 233A: Transcript – April 24, 2020:

This transcription is provided by artificial intelligence. We believe in technology but understand that even the smartest robots can sometimes get speech recognition wrong.

[0:00:04.5] BR: Welcome to This Week in Health, IT News, where we look at the news which will impact health IT. This is another field report where we talk with leaders from health systems and organisations on the front lines. My name is Bill Russell, healthcare CIO coach and creator of This Week in Health IT, a set of podcasts, videos, and collaboration events dedicated to developing the next generation of health leaders.

 

Are you ready for this? We’re going to do something a little different for our Tuesday Newsday show next week, we’re going to go live at noon Eastern, nine AM Pacific, we will be live on our YouTube channel with myself, Drex DeFord, Sue Schade and David Muntz with StarBridge Advisors to discuss the new normal for health IT. With you supplying the questions with live chat, also, you can send in your questions ahead of time at [email protected]. I’m so excited to do this and I hope you will join us.

 

Mark your calendar: noon Eastern, nine AM Pacific on April 28th. If you want to send the questions, feel free to do that and you can get to the show by going to thisweekhealth.com/live.

 

This episode and every episode since we started the COVID-19 series has been sponsored by Serious Healthcare. They reached out to me to see how we might partner during this time and that is how we’ve been able to support producing daily shows. Special thanks to Serious for supporting the show’s efforts during the crisis. Now, on to today’s show.

 

[INTERVIEW]

 

[0:01:25.1] BR: Today’s conversation is with Dr. John Halamka, president of the Mayo Clinic Platform. Good morning John, welcome to the show.

 

[0:01:31.8] JH: Well, good morning. Glad we could do this with social distancing.

 

[0:01:37.1] BR: Well, we always do this show with social distancing. Well, except at conferences I guess. Hey, thanks for taking the time, I know that you’re busy. You’ve been working on this national COVID-19 coalition. What is the coalition, what have you guys been doing?

 

[0:01:53.5] JH: Sure, let me tell you how it started and what we’re up to. It was about four weeks ago and I realized, at about five in the morning, that hundreds of companies were contacting me saying, “What is the organizing principle by which we could collaborate, communicate and convene?”

 

It was just at that time I realized there wasn’t an organization quite doing it. We decided that we would put a couple of folks who were willing together, MITRE Corporation agreed to service the program manager at no cost and today, we’re at over 700 organizations participating and we have 11 different work groups working on things from you know, what ventilators are needed where of what type? What personal protective equipment can you trust and how do we route it to places where it’s needed? How do we collect data? Do antimalarial drugs work? Do we have an idea if convalescent plasma is helpful or not?  We’re helping support the National Convalescent Plasma program which now has over 1,000 sites. Organically growing things of people just doing their [bit to] help each other with no economics, right? 

 

No one pays anything, we’re just helping each other.

 

[0:03:08.7] BR: That’s great. Is it a combination of pharma, government, providers? I mean, who is a part of it?

 

[0:03:18.7] JH: The answer is yes. All of the above and so for example, Google, Microsoft, and Amazon came in early. The entire University of California came in early, Rush, we have pharmaceutical companies. All of these folks, primarily private sector, right? It’s a private sector-driven initiative and, although we may have participants from government, it is not government sponsored.

 

[0:03:49.8] BR: Got it. Alright, so the show is this week in health IT and, while I would enjoy talking a ton about reopening the country and ramifications in this, I’ll let CNBC and the rest of those guys really handle that one. What I’d like to talk about is you know, things like data-driven outcomes, contact tracing, and platforms and just see where this goes. Let’s start with data-driven outcomes. What does the clinical care data set look like that the coalition’s looking at, that the CDC is looking at, that researchers and public health officials are looking at? Have we been able to pull that clinical care data set together?

 

[0:04:34.3] JH: Let me answer this in a couple of ways, which is suppose that we wanted to get together 200 Epic and Cerner using organizations to ask a question like antimalarial drug efficacy or convalescent plasma side effects or such. Can you imagine – and I’m not talking policy or technology, I’m talking psychiatry – what it would take to get 200 disparate organizations to contribute really comprehensive data sets to a central repository run by anyone in the short term?

 

The answer is, you know, you can do it eventually, right? Registry is maybe the answer eventually. You’re not going to do it tomorrow. Here’s what we did, we went to the EHR vendors and they were incredibly helpful! They said, you know what? Because each of our EHR’s, although standards-compliant, still has different ontologies, right? The way we name a field, the data we record, that sort of thing.

 

How about this? We’ll define numerators and denominators, comprehensively for any of the measures you want for Epic, for Cerner, for all scripts, et cetera. And then we will build those scripts at our customer’s sites. The customer can now run it against their live data and report the numerators and denominators, which are by definition, aggregated, de-identified, don’t fall into any regulatory framework, and that has worked amazingly well.

 

What I’ll call ‘new real time capacity’ to do cross-organizational analytics. All that’s running over the next couple of days on the areas that I just grab antimalarials and the use of convalescent plasma.

 

[0:06:14.5] BR: That’s amazing. I mean, not to – well, I mean, I guess it’s an appropriate question from a privacy standpoint, you’re just getting analytics back, right? You’re not getting any – there’s no records moving around?

 

[0:06:30.6] JH: What? No, If you were to ask Epic, “Epic, do you own your customer’s data?” They would say, no, right? We’re a software company and although they might have a cloud hosting service, most of the Epic instances are locally run, so what you would say is even if Epic wanted to run analytics across everyone of its sites, it just really can’t. I mean, that’s just not just the way the data is distributed.

 

But, to write the queries in Epic so that each customer could then run the numerators and denominators and share them, of course they can do that. This is a federated analytic approach that really gets beyond a lot of the logistics and data use agreements and that sort of thing to create a registry.

 

[0:07:17.5] BR: Will we see that expand to Meditech and others?

 

[0:07:22.2] JH: I have every expectation that this is going to expand, because there are really two ways we know to do analytics. One is that you create a federal certification requirement where every EHR vendor submits an HL7V2 or USCDI haloed to a place, and that could be the CDC or it could be some industry-based registry. I mean, that’s one way to do it. The challenge with that is sometimes that’s identified data and it runs into a lot of interesting issues on how data is used, who can look at it, and how you protect it.

 

Whereas this federated approach leaves the data within the firewall of every organization and everybody feels pretty comfortable about doing that kind of thing.

 

[0:08:09.1] BR: Yeah, it’s a pretty elegant solution actually. Not to get too far ahead of ourselves but, coming out of this, is this going to be the way we sort of design this for future pandemics, or are there other things worth thinking about in terms of the data and how to respond a lot quicker than we did for this one?

 

[0:08:30.5] JH: That’s just one use case, we’re asking, you said, data driven outcomes, right? So that’s an outcome. But wouldn’t you like to know, here it is April 21st, how many ventilators are going to be needed in Florida, two weeks from today? How many are in use right now? What’s the delta, right?

 

The challenge with that is do you have any idea how the CDC is collecting this information?

 

[0:08:55.5] BR: I assume it’s like a little strokes of pencil on a paper.

 

[0:09:01.0] JH: You got it! And then you hand type it into Excel and email it, right? As opposed to Keith Boone, Audacious Inquiry, HL7, Chuck Jaffe, they’re working on a FHIR Situational Awareness Standard, by which there’s just an API sitting on your EHR, how many people are in the ICU on ventilators now? 12, right? I mean, it’s not spreadsheets or smoke signals or Morse Code here. It seems to me that if we are going to be ready for the pandemics of the future, not only do we need all the outcome stuff – we talked about federated query – but we need situational awareness API’s, using FHIR, that enable our public health entities to say, whether it’s – how many are immunized? How many ventilators do we need? What PPE do you need? All this stuff is not that hard to automate if you have a standard.

 

[0:09:59.7] BR: Yeah, it’s interesting because we’re not experiencing this pandemic all at once, we’re experiencing it, it’s sort of rolling around the country. We probably have enough ventilators but, if they’re in the wrong location it doesn’t do anyone any good. So FHIR API is also kind of an elegant solution for determining where that is and getting that stuff moved around the country, I would think.

 

[0:10:22] JH: Well, as well, you have to recognize that this is a very unevenly distributed infection, right? You might very well today have few infections in Iowa and lots in Boston but you know, wait two weeks. What you’re going to need to do is put together the modeling of the epidemiology with the personal protective equipment and ventilator and staff capacity and begin to shift resources.

 

You know, a great author once said, “Famines don’t exist because of a lack of food, but a lack of a distribution of the food,” right? I think we’re seeing actually the same thing with personal protective equipment and ventilators right now and this FHIR approach is something we’ll learn a lot about in the next few weeks.

 

[0:11:12.9] BR: You know, this might sound like a self-congratulatory thing for the industry, but could you imagine going through this pandemic without an electronic medical record across the board? Which we started a decade or decade and a half ago? I can’t imagine where we would be.

 

[0:11:29.1] JH: I know, exactly, right? People say, the EHR of course has had a burden which it has. The challenge has been, in the past, we haven’t returned anything to our clinicians from the EHR – they put data in, they didn’t get much out.

 

Now, the ability to say now we’ve put this data in, we’re going to get coordination of care and supplies and learning and cures. Aha, there’s a benefit, finally, at the end of the rainbow. 

 

[0:12:00.6] BR: All right so let’s start talking digital tools, and I want to start with contact tracing. We know that Apple and Google are working on solutions. Are there solutions that individual health systems can implement or are trying to implement? 

 

[0:12:16.4] JH: So let us talk about the multiple approaches that have been used. So MIT at the Media Lab with Mayo and MGH working together, this was about a month ago, created a GPS tracker, and here’s how it works. Basically, you consent and you say, “I actually would like to track my location so that if there is someone I crossed who’s positive, I can submit my GPS tracing over the last month to a registry at public health and see if there was crossover.” 

 

And that is great, but it has one challenge. You said you’re in Naples, Florida. Well your GPS would have your home location in your GPS data. That is not exactly de-identified, right? So you would have to then, as a public health department, to redact an individual submission. You know it might be the place they work or the place they live or something that would be truly identifying to them. So what Apple and Google have done is really interesting. 

 

Would you agree that if you want to do contact tracing you would simply turn on your Bluetooth function on your phone and an API would be available to public health from your phone that does nothing more than collect the Bluetooth signatures of the people within six feet of you, right? So it is not identified and, in fact, the global universal identifier for the Bluetooth is actually changed quite frequently so that there is really no mechanism of hacking and figuring out people’s movements and that kind of thing. No GPS signal. This is just to say, “Who have you been close to?”

Which point then when a phone is identified as being a COVID-positive person, anyone who was near that phone could be notified. So it is really kind of an elegant privacy-protecting way of doing contact tracing. 

 

[0:14:21.9] BR: Yeah that is interesting. I mean would there be a situation where a health system would try to go to this length? Or, essentially, I mean, is there a need within a care delivery platform to have this kind of mechanism? Or would you just tap into Google or Apple’s solution?

 

[0:14:39.4] JH: Yeah, the challenge is that your contacts are bigger than any one health system, and so I think the answer is that many of us will want to contribute to these contact tracing efforts by providing clinical expertise, privacy oversights, maybe even create some novel apps – and we are working on some novel apps to help with return to work and things like that – but it’s truly got to be a national scale effort. 

 

[0:15:13.5] BR: So what other digital tools have you seen be effective during the pandemic? 

 

[0:15:18.2] JH: Well, I mean there’s a couple of threads that I want to work on here. Mayo has recently released what’s called Mayo Expert Advisor, which is traditionally a tool that is not available to the public. We are putting it now in a place where all physicians in the country can access it, and soon even the public will be able to access it, and it provides accurate advice. So here’s a problem: you think eating twice your bodyweight in garlic is going to help your COVID resistance? Well, there are websites that suggest it!

 

[0:15:54.6] BR: Well, John, obviously it is going to because it will keep people away from you I would assume. 

 

[0:15:59.3] JH: Well there you go! But the point being is that part of the apps have got to be just getting rid of the misinformation. So a lot of us are working on just getting the right information out, getting the data out. If you go to the website of the COVID Coalition, there are over 700 publically available datasets on that website, and data visualizations that anyone can access. There is no password needed at all. And so I’ll get you the website before we finish. 

 

So you know it is situational awareness, it is knowledge in getting rid of misinformation, and then of course I think there are a number of interesting apps that are being deployed. There is a non-profit in Seattle called Audere, funded by the Gates Foundation, and their app says, “What are your signs and symptoms?” It does a screening to determine if you need to be tested and then, if you need to be tested, arranges for you to be tested. 

 

And so figuring out who needs a PCR molecular test. As we get past the infection and getting to return to work, then there is a question of tracking the serological tests, immunize, do you have IGG? Do you have demonstrated resistance? Because you could imagine it will open the economy by bringing people, who have already had it and recovered, back to the workplace first, you know, assuming that they are immune. 

 

[0:17:32.4] BR: Yeah absolutely. All right so let us talk platforms and, if we have any more time, I mean I got like 75 million questions, but, from a platform standpoint, you’re the president of platforms now, which is a great title. Platforms help us move quicker within a crisis or outside of the crisis because the key components already stood up, and access to the data, security, it’s already baked in the platforms, so you can really focus in on the problem at hand. How have you seen platforms or even the absence of platforms come into play during this pandemic? 

 

[0:18:09.3] JH: So a number of us have just completed an article – which has been submitted, we hope it gets published soon in the scientific literature – that says we are going to look at the medical record of every COVID positive patient that has been tested in our organizations and we are actually going to start to look at characteristics about them. Who had a severe infection? What were the earliest signs? Can you predict severity based on early signs? 

 

Well imagine this: suppose you wanted to go do a study that required you to access 20,000 medical records and do a comprehensive analysis, how long is that going to take you? Well, we did it in a week! And that is because we had a mechanism in place to de-identify data, place it in a container where then we can bring analytic tools into that container, run the analysis, and not share the data but exfiltrate the knowledge, right? 

 

And so this is what the idea of the platform does, it creates these reusable components and modules, these processes and these policies that enable you do this stuff really fast, without having to re-litigate, with teams of lawyers, the process from the very beginning every time. 

 

[0:19:26.9] BR: Yeah, you talked about that at the JPM Conference and we’ll probably come back and talk more about that as we go forward. You know as sort of a set of closing questions, talk to me about testing. I know I am going to get outside of health IT here but I am more curious than anything. How much testing do we need? We have obviously the molecular testing, which is pretty quick to determine if you have it. We have the antibody test, which looks like it is going to start ramping up if they release it from Abbott, the pressure leases and things hold true, which is 20 to 30 million tests a month coming forward. I mean what is an adequate amount of testing to get us moving back into this? 

 

[0:20:13.6] JH: As you might guess there are many models here. Amazon’s model is that they think in order to get the Amazon supply chain back that they – you probably read about Jeff Bezos’ blog about this – that they need to really do very rigorous testing on each of their employees on a regular basis. I mean this could be hundreds of thousands of tests a week, just to keep that workplace safe, and that is actually not so different to what some countries have done. Kind of aggressive testing and contact tracing and that kind of thing. 

 

I think our challenge at the moment is we just don’t have easy access to either the molecular testing or the serologic testing in the numbers that are needed for diagnosis and return-to-work. So you may have seen – again, we’re not getting into politics here – but the senate health committee, last evening, has put together a right around the new, we will call it the Paycheck Protection Program Successor Bill for 25 billion, to just be routed to get us the adequate number of this PCR and serological testings wherever they are needed, to get people diagnosed, and the economy back on its feet, and that number is $25 billion needed to really ramp up our testing capacity. Just look at the number of tests we can do every day today, it is a small fraction of what is needed, and so I think the models are varied but it is still a small fraction. 

 

[0:21:46.4] BR: That is interesting. Well John, again thanks for taking the time. I really appreciate it. I know you are on Zoom calls probably from what, like 6 AM until 6 PM every day? 

 

[0:21:57.4] JH: That would be a short day. For the last four weeks I have been averaging 18 to 20 hours a day. But, you know, it is a once in a lifetime opportunity for us to do good here. 

 

[0:22:11.5] BR: Yeah, so if Zoom were smart they’d offer frequent flyer programs and, you know, they could replace all the miles you were putting on airplanes I guess. 

 

[0:22:21.4] JH: Well that’s great. Now one thing I just want to leave you with is the website for the Coalition, https://c19hcc.org, and it is all free and it is all open source. In the research library you will find hundreds of great data visualizations, and training, and all kinds of coalition-based resources. 

 

[0:22:49.8] BR: Fantastic. John, thanks again for your time, I really appreciate it and I look forward to having you back on the show. We’ll get an update in a little bit. 

 

[0:22:58.6] JH: Bill, great, you have a wonderful day. 

 

[0:23:00.4] BR: Thanks, take care. 

 

[END OF INTERVIEW]

 

[0:23:02.0] BR: That is all for this show. Special thanks to our channel sponsors, VMware, StarBridge Advisors, Galen Healthcare, Health Lyrics, and Pro Talent Advisors for choosing to invest in developing the next generation of health leaders. If you want to support the fastest growing podcast in the health IT space, the best way to do that is to share it with a peer. Send an email, DM, whatever you do. You could also follow us on social media, subscribe to our YouTube channel. 

 

There’s a lot of different ways you can support us, but sharing it with a peer is the best. Please check back often as we would be dropping many more shows until we’ve flattened the curve across the country. Thanks for listening. That is all for now.

 

[END]

Play Video