Description: The article is about bidmc portal. The author interviews some people to ask about their views on current news. The most important issue they are talking about is if the patient data should be accessed by anyone. They discussed the topic a lot.
Welcome to HIT policy update with Dr. John halamka,a webinar tweet chat combo from Health System cio.com sponsored by Athena health. I am going to give you a little introduction before we get started. My name is Anthony Guerra. I’m the editor chief of Health System CIO.com and I will be your moderator today. We are having a simultaneous tweet chat hosted by Kate gamble,our managing editor and director of social media.
You can participate in a separate browser or on your phone by using the hashtag HS CIL chat or you can simply view the tweet chat in the media viewer panel on the right hand side of your screen. We’ll also use the Q&A panel. We encourage you to type in your questions as they occur to you,you can also send them in and we will take those later in the program. You can download the deck by using the URL on your screen.
Now you have realized how we’re going to spend our time today. First we’re going to spend 30 to 35 minutes with Dr. John halamka. Then we’re going to hear from our sponsor represented by Chris taylor and enterprise accounts strategy with Athena health. We’re going to have our Q&A. Without any further delay,I’m going to turn it over to our good friend and regular guest Dr. John halamka. Every time we do this I will say it’s hard to believe the policies change fast enough to fill a webinar every quarter. But in fact they do.
All you have to do is pick up the New York Times and see what policies have changed in the last 20 minutes to know what chaotic disruptive change is going on. I’m going to try to capture at least the themes of the last couple of weeks. We’ll look at what were the major new technologies what was the hype and the reality and the booth. Then we’ll go through what is ONC up to and what our folks in both HHS and CMS and ONC focused on. Given all the other changes in the administration have some consistency and those three agencies so we’ll look at that.
We’ll talk about the priorities of some of the new players. Although we ohad a change in the Secretary of the Veterans Administration at the moment it seems like there is going to be at least a quarter or two predictability on the hie policy front. With that let us begin so at him. What did we see? We had a keynote from Eric Schmidt,the former chairman of alphabet parent company of Google. What did he tell us?
He said so many of these issues that we face in healthcare IT are all about search. If we use the Google API for health and send a very large corpus of data to Google,they will be able to solve many of these healthcare IT challenges. You may wonder is it all about search? Maybe that’s an oversimplification but what did Eric tell us is probably true. In the world of ophthalmology,an average ophthalmologists maybe looks at 10,000 retinal scans in a lifetime. Verily this healthcare related subsidiary of Google has a million retinal scans on file and has built a number of applications,machine learning applications and interpretation applications from a million retinal scans.
The problem with machine learning in AI is it’s a black box. Pattern recognition is statistical. We don’t know what the machine learning is doing but it has somehow recognized patterns and retinal scans sufficient to outperform humans. What Eric said was it’s not that Google is smarter,it has much more data. I think that’s a fascinating question. I don’t believe that we’re going to see humans displaced by machine learning and AI but certainly think they’re going to be augmented.
What do I mean? I was talking to my chief of radiology the other day and he said when I was a resident we would have maybe 20 images in every MRI scan to review,I finished reviewing a single scan with 330 images in it because it had three millimeter cuts,axial and sagittal reconstructions. No human can look at 330 pictures in any detail and make some coherent conclusion.
So what we need at this point in history is some augmentation. Not means these deep learning systems are going to replace human cognition but to say of those 330 pictures,there is something funny in these 25 and then the human can focus on that. What you heard from Eric and you heard from generally Google Amazon Apple is this scale of their operations by their storage their compute their facility to exchange information and to supplement that with tools. They are able to move the healthcare IT industry forward. I think that’s probably true.
I don’t think it’s all about search but I certainly think it is all about scale. I don’t know about disrupt the supply chain and provide devices and pharmaceuticals or what JPMorgan and Berkshire and Amazon are going to do together. But I guess the reality of what I can say is that we’ve got AWSS3 and some of their other services.
There are a set of things which I will call them a Lego blocks. That are available from Amazon. Any innovator can glue together at very low cost in very short time to do something amazing. Let me read an example of something we’ve done with Amazon. Have you ever had surgery? You don’t have to answer that question. It’s a HIPAA violation. If you did we need to have a consent on file before we open the scalpel.
How is that consent delivered to your doctor today? You sign it probably some consent in a outpatient setting with a pen and that faxed to a surgeon or hospital. They lose it a lot. But we will open the scalpel until we validate that you have a consent.
We asked Amazon can you use the machine learning Lego block to listen to our fax traffic? When you see a consent coming over a fax line,can you put a checkbox in our EHR by using a API that says the consent was received? You can put a digital copy of that consent in the EHR under the right patient.
That works totally fine. We’ve removed the manual paper-pushing of humans and now we have faxed consents going into the EHR and they eliminate the cancellation of operating room procedures. This is perfect. It didn’t take us a year it took us a weekend to do this because Amazon says “we got a machine learning components,we got an image recognition component,you can lift those and use them as you will.” You can script things together.
Now we also heard about some of the work that Apple has been doing. In iOS 11.3,there’s going to be a standard Argonaut compliant fire API so that any patient can get their record from an episode of care as they’re walking out of the office.The initial deployments across 13 hospitals will include Surtur sites and epic sites and afina sites.
I mean imagine that the day that we get 11.3 released,you’re going to have EHRS that are able to send your clinical summary directly to your phone without a huge amount of complexity. They are all using national standards. Apple has produced 800 million iPhones then you suddenly have 800 million containers under encryption and biometric control by the patient that can receive data.
I’ll summarize this early keynotes and events of our EHR vendors who are here to stay epic Cerner Athena manutaki Club works not going anywhere. They’re going to be augmented by a series of cloud hosted services available from Google Amazon Apple. That is going to fundamentally change the way we develop applications and interpret data. There is other example of machine learning and AI. It is not real.
I am going to do diagnosis and treatment. This is very prosaic. We gave Amazon a million operating room records and we said Amazon,now that Anthony needs his appendix out,can you tell us how much time we should give them in the OR? And Amazon can say Anthony’s a young healthy no comorbidities person who’s also thin.
It needs no more than 25 minutes for an experienced surgeon to remove his appendix. We went live with that component and we looked at 15 surgeons workload and we were able to free up 30% of their OR time by sending historical schedule data to Amazon. This is the theme over and over. It’s augmenting the existent EHR with external services via API and cloud. Now we certainly see more and more mobile and more and more applications,some patient facing some provider facing. It is clear that in a world that’s post meaningful use.
We’re going to enhance the usability otherwise we know that more than half of doctors want to quit because they are so disenchanted with the burdens of having to do data entry.Where you get these creative apps whether on Athena’s more disruption,please program epics,orchard Cerner,smart on fire saw lots or lots of emerging highly usable apps running on mobile devices connecting to the EHR,I believe cloud with this it has come to its own. This cloud is secure enough and reliable enough and cost-effective enough. It’s prudent to move there. I’ve been a CIO for 22 years.
Do you think I want to wake up every morning thinking whether or not it is a servers booting? I don’t. I would much rather push the Microsoft Tuesday updates to somebody else. The maintenance of the power and the cooling and these things I think is this broad realization. Whether its infrastructure as a service platform as a service software as a service,that cloud is the way we need to go. CIOS becomes agents do procuring of cloud hosted services and the plumbing to integrate them as opposed to provisioning hardware and software.
The era of on-premise hosting and software licensing is dying fast. We also saw interoperability. Somebody is going to ask a question about if blockchain will solve a interoperability problems in the world. I would never do that. I am now the editor and chief of the a new peer reviewed open access journal called blockchain and healthcare. Why did I agree to become the editor in chief of that? It’s because I want to find the good use cases where blockchain might help us but also describe where it won’t. We saw some block drain but interoperability mostly.
As we talk about it in a minute focus on APIS and you will keep using the old hl7 ADT and lab rag. We’ll use this CCDA clinical summary. But with these new API,this is the notion that we can have a patient or a provider query data in real time and get the structured data elements in a highly predictable way.
It is powerful. I would tell you you’re starting to see this year the Argonaut open source fire specs disseminated into products and services both applications and EHRS. They use the cases for blockchain. I’m seeing in the industry we are more about proof of work,proof of data integrity. Blockchain is slow. It’s seven transactions a second if you’re using etherium as opposed to G. I move 22,000 transactions a second in my EHR. Blockchain is never going to help so much with that. But it helps us do things.
This is a note that I wrote. I signed it. And maybe I put a hash and blockchain and therefore I can guarantee the notes that have been altered. There’s data integrity. Nothing’s been deleted from the medical record or altered. Maybe I can use blockchain as a means. I’m not going to say anything about the care. I’m going to say you were there.
Therefore we can look to the blockchain to figure out where your records might be found or potentially we have all issues in every state with consent. Could blockchain be used with smart contracts to hold information about your consent? The answer is sure. I agree to share my data for these purposes with these entities. That’s in a public ledger. So it’s clear what data exchange I’ve consented to. It’s not going to solve all the issues.
We have to be very tactical and understand its limitations. That was three days of hymns in about 10 minutes. I saved you talking to 45,000 people. We are starting to see on the policy side coming out of ONC CMS and HHS. At the moment ONC is focused on two things,21st century cures and the trusted exchange framework and common agreement or TEFCA.
Teske was something required in 21st century cures which was a bipartisan piece of legislation at the end of the Obama administration. What supports that are always good. They are accelerate innovation and drug discovery,interoperability,get rid of information blocking. But they didn’t say how. Teske has two sections. I’m going to save them reading 40 pages of Teske in two slides.
The first part which I think is very good is the principles of trusted exchange. We should only use commonly available,well proven international standards for content and transport and vocabulary. Everything should be done openly and transparently. There shouldn’t be silos of data hidden away. Right patients providers should be able to access data as is needed for treatment payment and operations care coordination.
We know that even if you have arch enemies across town,you shouldn’t let competition,you shouldn’t let politics get in the way of good clinical care coordination. We know that everything should be done in appropriately secure fashions with low levels of encryption and auditing to ensure data integrity and that privacy is respected.
It’s certainly great for patients and families to be able to gather all the data about them and have a continuous health record on the device or after their choice make great sense. They’ll understand certain aspects of data and metadata such as who is the patient and what data does it represent. There’s a lot of challenges because we don’t have a national healthcare identifier. I’ve written a lot about this. We need a nationwide patient matching strategy if we’re going to effectively exchange data. So all of this was good.
Where I think it gets into a little bit of trouble is in Section B. It enumerates the technologies that we need to use. It was so specific about what would be a qualified health information network and an organization that approves these qualified health information networks. The rules got to be so prescriptive in Section B.
It will impede innovation. Government is good at many things. Government can be a catalyst. Government can provide incentives and penalties. It can provide us some things like common privacy policies. I’m not sure government is very good at specifying technology. I think the private sector should be told here are the goals.
Here are the customer requirements. And now I am going to talk about private sector because some of the standards that were enumerated in Part B of Teske or invented in the 1990s. They’re not these internet friendly mobile oriented cloud encompassing API. It’s good to have guiding principles,rules of the road and guardrails. But what happens over time is we have a more outcomes focused approach and the industry is given the latitude to implement those desirable outcomes as technology evolves and as customers to map at.
Let’s look at some of the characters in Washington today. That jared has been assigned peace in the Middle East,American innovation and interoperability. It seemed a little bit unusual. I don’t know what do you think which of those three is hardest. Maybe peace the Middle East is easier than interoperability. It’s fascinating. It over his plates full. He’s very engaged. I hear time and time again when people go to the White House to talk about healthcare IT.
They’re meeting with Jared. He’s carved out the time to focus on this issue and so what does that mean. I don’t know that he is necessarily going to put forward specific regulatory proposals. The current administration doesn’t necessarily do that. But probably he will show an impatience.
As an industry does a better job at making data liquid especially around getting it to patients and families,to that end it happened at hymns. But it has happened more since then our administrator CMS has said that data exchange going forward should be patient centric. We should be able to have aggregations of everything about you and that you should be able to have a copy of that. If you want to forward it around,if you want to contribute it to a clinical trial or clinical research you should be able to. She announced the blue button 2.0 initiative.
If you remember 15 years ago there was a meeting at the Markle foundation,at Carol diamond and a group of folks said it maybe easy to get your data about you. It seems like pressing that blue button when you get a ticket at a parking lot. It should be that easy. What came out of that discussion was the blue button 1.0 which was a text file of basically claims data explanation of benefits information. That text was good but getting a text file a blob of all of your claims is not something you can do much with.
It seemed announced at hymns is the idea that blue button 2.0 would be a Argonaut fire API with your structured data accessible via an open IDOR aught authentication mechanism. Apple has done it in iOS11.3 and it is so many industries are doing now. We’re smart on fire and other fire demonstration projects. That’s good.
You will probably see blue button 2.0 on top of the Medicare data set,giving patients access to all that Medicare data and presumably opening up avenues for new innovation. She also signaled that Meaningful Use is running its course. We’ve done what’s necessary to get adoption. We now have more than 90 percent of hospitals and doctors using electronic health records. Do we need to have multiple years of additional attestation?
The stimulus is already spent. We’re in a penalty phase. If we instead focused on API and patient family engagements and building mechanisms for enhanced data liquidity,that’s probably a better future than looking at Meaningful Use. We’ve already started to see some revision of quality measures by CMS,we’ve already seen delays in any reporting requirements for Meaningful Use and so maybe. We’ll see that entire Meaningful Use construct disappear and be replaced by a new focus on interoperability.
He feels that’s the charge of ONC. ONC was reached for the same as last year it was going to go from 60 million to 28. Now it’s back to 60. So in the budget that Trump signed,it is fully funded going forward. One expects them to focus on Tesca 21st century cures,interoperability,reduction of information blocking.
They’ll also be very supportive of the Government Accountability Office the GAO doing a study on the costs and benefits of a nationwide patient matching strategy. You also hope 2018 becomes the year we finally develop a strategy for being able to match records across disparate EHRs so that the patients can truly have a longitudinal view.
That’s hymns. That’s the trump administration. I know you asked me to describe everything that Donald Trump is thinking. I did my very best. We’re looking forward to our QA in a few minutes. Before we get there we’re going to hear from our sponsor Athena health represented by Chris Taylor who heads up Enterprise accounts strategy.
If you don’t mind going to the next slide I can take it from there. Since our inception,our vision has been to build the health information backbone that helps health care work as it should. They were uniquely positioned to deliver a better healthcare experience because of the network we’ve built and continue to expand. What does the network mean? What are we doing it for? Could you please go to the next slide Anthony?
Our purpose is to unleash our collective potential to transform healthcare. So how do we do this? To realize our purpose we live by these three commitments. We live in relentless pursuit of open healthcare. We aim to be the platform where patients providers partners data and services are wholly connected fueling unparalleled innovation. To do this we commit to fluid exchange of formation and resources connecting everyone to the information they need.
Second we multiply intelligence. We believe the best results come from the inside of all. We will bring together knowledge from every interaction expanded exponentially and concentrated where and when it’s needed. To do this we commit to surfacing the most relevant insights and sharing what we know and finally free people to do what matters. We aspire to make every human interaction more meaningful.
We strive to remove the friction from our healthcare experience,liberating us to perform at our highest potential. To do this we commit to removing work from medicine that does not require judgment empathy or clinical skill. I’m thinking of the images the scans that doctor Halong Co was mentioning earlier.
I think there’s real opportunity there. Why does this matter? 95 percent of physicians have experienced difficulty or delay delivering medical care because they didn’t have access to their patients health records. Even inside the same organizations only 44 percent of physicians reports that they can share their patients information.
It’s taking its toll. Over half of physicians report the symptoms of burnout. As Dr. Helen Koh mentioned,most physicians wouldn’t recommend the practice of medicine to their children. To us this is more than numbers it’s personal. When people come to work and feel like they’re a robot.But they’re not being appreciated perspective that they don’t get the chance to improve. I think a lot of people work sixteen hours day very well with a cooler water.
I hope that that message resonated with everyone. The big takeaway for me is burnout. It is not about being overworked. It’s about not being able to do your job well. It’s about not feeling the sense of humanity in delivering health care. We have studied the interactions behaviors of over a hundred thousand providers or networks try to understand and make public understand what drives capability for providers. As the technology leaders for your organization,you are the innovators,you are the change agents of a better future in healthcare.
I want to thank our our attendees for bearing with me on playing that video.If there was an echo which we hear some of you,I want to apologize but hopefully you will absorb the point of that which actors are not afraid of working hard but they want to feel that their time is being used valiantly. It was a very nice. There are questions for Dr. Milanka.
It’s about agility. It’s about security. It’s about cost and so what do I mean. Today that stores the records of two million patients,every medical record and patients in Boston from 1977 to the present. Where it’s stored? It is stored in Amazon. How many employees does Amazon have keeping my data secure and reliable? 50,000.
Let’s imagine something bad happens on the internet. I have eight of the most talented professionals in security looking at beth israel deaconess data on my side. Amazon has 50,000. Who’s going to be able to fix it first? There’s the challenge. The challenge is when you’re looking at this need where we have zero tolerance for downtime,when we need data centers that go beyond a geography.
My board said wait today we have three triple replicated data centers with a recovery time and point objective. That’s very wonderful. This is all new england-based. What if North Korea goes wacky? What will you do about that. You could say we’d have bigger problems. The reality is by using cloud or multiple clouds,these are geographically dispersed replications of your data. By computing that,it provides you enhanced reliability and security data recovery beyond anything you can do by yourself.
Here’s another strange thing for me to say about cloud. I am sure that Health System CIO is an extraordinary employer. But israel deaconess is a great place to work but our stock options are not so wonderful. We don’t have Herman Miller chairs fish,tanks,neon or cappuccino machines so how easy is it for me to attract and retain the best AI machine learning talent in the business?
It’s hard. If you say I am now instead of going to invent these technologies myself,I’m going to have a cloud provider that is good enable straightforward coupling of my data assets in the cloud to other functions. They can hire these extraordinary people and ensure the scale. I can focus on the business requirements and my customers instead of trying to hire and retain these young AI machine learning professionals.
And finally I’ll say I have 22,000 employees. In my multibillion-dollar healthcare system,my licenses for microsoft products are still more expensive through me then through a cloud provider Amazon. It is more reliable more secure or better retention of talents with lower cost. It seems reasonable to do that. Is that enough for a business case? Is that a business case? My board said I better do it. How about that? Could you have a board that was reluctant to do this thing and want a hard ROI business case?
They care about reputational lost. If you spill data or they care about the agility, the innovation,the ability to stay competitive and attract patients and keep patients and so it’d be hard-pressed for me to imagine a board. Most of my board are folks from the financial services industry or retail.
They move to the cloud years ago. They’ve tell me it’s a risk mitigation strategy. It is a mechanism by which patients and providers will achieve better usability. You’ll be able to move faster and I suppose what your metrics for business case are. But if it’s safety,quality,cost,strategic alignment and it certainly hits all those elements,what are the chances we get a nationwide patient ID? I was extraordinarily constrained in my description.
I said nationwide patient matching strategy. Why did I use those weasel words? You’re going into politics. I travel the world and I know some societies are completely fine with the idea of a birth by giving you a number that will track your healthcare data throughout your life. If you were born in Norway today,what would your national patient ID be? It can be twenty eighteen o three twenty-nine zero zero zero one two three four right.
It’s the date and your birth order. It seems straightforward. But imagine, America as three hundred and thirty million fiercely independent folks who don’t want anyone to tell them what to do or to control their data,it’s hard. It’s a cultural challenge.I think there are maybe a couple of ways we might go. We could say you got a TSA PreCheck because it makes travel easier. How about a healthcare pre-check? You can sign up for it by giving us your fingerprints,giving us your demographics. You don’t have to have one. It’s voluntary. That’s one kind of nationwide strategy.
Here’s another one I hear startup pitches every day. and one that I heard recently which is pretty reasonable. Maybe you’re called Tony. Your differences are totally reliable. It was your dad named Fred. I know where you’ve lived who else has lived there,what a car you bought,what property tax you pay that’s a little bit outside of health care. That in combination could be referential data that makes the name gender date of birth match a whole lot better.
Their mom’s name was Jean. They lived in this city back in the 70s and so that might be a strategy. Then one of my favorite and one that I’m working on a lot these days is biometric. In South Africa for the Gates Foundation,we had to match the HIV data 65 million patients across the country. We deployed iris scanners. The idea is we’ll simply tag your lab data with some hash of your iris scan.
When you come back into clinic,we’ll be able to say we’ve seen these irises before. Here are the lab tests associated with them and the doctor and the patient have now fluid access to every test done anywhere in the country based on iris scan tagging. That’s a strategy. Sometimes we do multiple biometrics there may be people without irises lost their eyes,so we’ll do fingerprint and iris or palm vein geometry or other things.
We stay a bit neutral to the evolving biometric technologies. It is a long answer but I think yes. We must as a country articulate the guidelines and principles of matching data in a consistent way for high false. We’ll call it positive predictive value. It is if we’re going to have a coordinated healthcare system.
You had mentioned the idea that EHR vendors are not going away but that they’re going to be augmented. Do you have to be willing to be augmented you have to be an augment Abul frame of mind? I would imagine if we have a spectrum we’re going to have some EHR vendors that are more open to becoming platform life with Apps being plugged in and some that are less inclined to do that and play nice in the sandbox. My guess is you’re going to say if you’re not open to other people you’re going to be in trouble but I’ll let you answer that.
There are 750 certified products that exist in the marketplace today. When I said the EHR vendors aren’t going away what I meant is Cerner epic Athena metatag eClinical Works. We are always going to need transaction engines to get the bills out and to be compliant and meet with all those Medicare and met regulations.
They’ll be there. Unless they provide an ecosystem of innovation around them which enables patient data access through applications then they won’t survive. That’s why I think a lot of the smaller niche players will disappear because they don’t have that mindset exactly as you said. How do you think the newly announced program might healthy data will fit into the interoperability landscape? Are you familiar with that?
I don’t have any intent of getting an application on my phone and being the steward of my data and managing it. But you should do that on my behalf. I think patients and families are more engaged than ever before and want their data more than ever before. But it’s going to require thoughtful delegation of who acts on it.
It will either be family members like me. I would do my mothers or my wife’s or my daughters or you might even see third party services. We are the Care traffic controller for your family. Delegate access to my healthy data API to us. I think it’s good. It won’t solve every problem. There’s still going to be provided a provider data exchange but more patient family engagement is good. The whole JP Morgan effort was interesting when I came out.
If you recall,at least from my recollection. Some of JP Morgan’s banking clients that had been very good customers of it who thought they were going to be negatively impacted by some of the things coming out of here in terms of competition and then it was all kind of scaled back a bit.
I don’t think we quite know what it is yet. We’ve been talking about the scale of an Amazon,the financial resources of a JP Morgan and Berkshire,they will enable disruptive innovation and maybe we’re going to disintermediate some insurance companies then maybe we’re going to provide unique opportunities starting with some cohort like Amazon employees. For I don’t know employer,funded health plans that focus on wellness using unique tools that help you eat less exercise more and generally have a better lifestyle.
If we did that your total medical expense will go down. It’s a little early but all I can tell is those three companies are impatient and have the resources necessary to try these things. With Amazon’s testing markets where you walk in and take stuff,you probably won’t know who works. I am looking forward to the initiative as it’s a opportunity to learn something that you do not know about. If you like us,you can reduce a webinar on the topic of your choice. You can reach out to Nancy Wilcox on our team. You can go to our website to view our upcoming schedule of events.