"Digital innovation is an applied science"

Welcome to the big unlock podcast where we discuss digital transformation and emerging technologies in healthcare. Here are some of the most innovative thinkers and leaders in healthcare and technology. Talk about how they are driving change in their organizations.

Paddy: Hello again everyone. Welcome back to my podcast. This is Paddy and it is my great privilege and honor to have as my special guest today Manu Tandon, CIO of the Beth Israel Deaconess Medical Center which is an affiliate of the Harvard Medical School. Manu welcome to the show.

Manu: Thanks Paddy. Thanks for having me. It’s an honor to be here.

Paddy: Thank you so much. So just to set the record straight and for our listeners to really understand the structure. So, Beth Israel had a recent restructuring right back in March I believe you want to just quickly walk us through what the current structure of the Beth Israel health system is.

Manu: Sure the BIDMC which is Beth Israel Deaconess Medical Center as you mentioned is academic medical centers, Boston- based, it’s a non-for-profit, is affiliated with the Harvard Medical School, located in downtown Boston as a tertiary quadrant care hospital with about 670 beds. On a research side receives about 170 million in research funding. And like you mentioned back in March we just became part of a bigger system the name of the new system is Beth Israel Lahey health (BILH). BILH is bringing together about 13 hospitals, few AMCs, committee hospitals and is super excited about this new development and I will say that we are in a early form of our merger. And so, there is a lot to figure out but a lot of exciting thoughts and exciting stuff that’s happening around it right.

Paddy: Right. And I know that BIDMC also launched a Center for Health IT Exploration and John Halamka was on my podcast last week and I guess that’s also part of the overall health system. Is that right?

Manu: That’s right. John and I work fairly-closely. John’s looking at the BILH level innovation market space. He does a lot of work internationally and nationally. We at BIDMC has something called a Center for IT Exploration which is like you said a little bit more focused on the academic medical center, more tied with the annual operating goals of the hospital and you know our mission here at the CITEx, which is the short for Center for IT Exploration, is to look at innovations that are happening outside of healthcare and trying to bring them into healthcare but under the context of our operating plans. If our operating plan says hey our number one problem is to maximize our existing capacity, then we are looking always looking for solutions that can help us with doing that. I hope to talk to you about a few of those as we go along today.

Paddy: Excellent. So we met Manu when I attended the Executive Education program at the Harvard Med School where you gave a fascinating presentation on your IT environment and the thing that struck me the most was that you are the only major health system in the country to run your own electronic health records system. So, you tell us a little bit about how that came about for the benefit of my listeners and why the Beth Israel Deaconess Center choosen not to implement one of the major industries EHR platforms to date.

Manu: Yeah. So, I think I’ve been here only for four and five years. Our EHR obviously predates me, several very highly talented folks have touched it as it has gone along. Back in 1970 it started with an NIH grant to two doctors Warren Slack and Howard Ablation. They were trying to build a system which was patient centered and I guess they were more patient centric at the time before such a term even existed. And so, what they did was they had this notion of usability and workflow to be central to the development of such a system. They started out with building actually a master patient index with about five hundred thousand patients which is what we had at the time and then sort of one thing kept leading to another and over time they added you know clinical ancillary systems like labs and radiology, and the diagnostic, administrative systems like registration, admitting, outpatient appointment, charge capture systems, clinical documentation. So, this has now grown into what we call OMR. And then includes basically the entire gamut of EHR functionality in 2004. We launched web OMR, which was basically a web-based version of OMR. So that’s kind of where we are right now in terms of your question of why we still use it. It works within the four boundaries of the hospital fairly-well; it compares quite favorably and provider satisfaction surveys. We occasionally compare its functionality to the leading EHR vendors, and it seems to fairly okay on that account as well. As I mentioned is browser based so that has some advantages to it. It’s accessible from any browser running on any device anywhere. I will also add that having your own EHR allows us to control the workflows and as I’m sure you know one of the biggest adoption barriers to innovations and third-party innovation especially is workflow integration. So, we seem to not have that issue. I think this combination that we have has resonated with several innovators like Amazon, Google, and MIT with whom we have quite fruitful partnerships. I will say it does come with challenges. Interoperability is I will say as a national level it remains a challenge and you know keeping up with interoperability is a challenge, I think. You know occasionally we have to go through certifications that draw resources to it regulatory requirements state or federal. And I think now that we are part of a big system as I mentioned the top of the call on your prompt. You know I think we are taking a look at safe handoffs of patients within the system and where do we go with our EMR. But I would say we’re all where we are right now. You know the EMR works quite well like I said within the four walls of the hospital it provides a good core platform if you may for workflow integration, for innovative solutions which we see as a differentiator. And then obviously it helps that it is very friendly to the bottom line as it comes at a very tiny fraction of what it would cost otherwise. And we have been blessed to have great staff highly skilled, dedicated. Even after so many years, Paddy, of its existence we are still rolling out more than a dozen enhancements every month to it. So, it’s a very robust environment.

Paddy: That’s a fascinating story actually. So, can you talk a little bit about the overall technology stack and the ideas we see that you manage and a little bit about the high-tech governance model.

Manu: Sure. So, the tech stack, we spent some time talking about it. The core of our tech stack is our homegrown EHR system. But then we have all the typical applications you would imagine you know PeopleSoft; we use PeopleSoft software for HR finance, supply chain, several websites, revenue cycle system. We are in the middle of a multi-year project to migrate to the Amazon Web Services public cloud. So, at this point we are about 50 percent migrated. Over the course of the next year and a half we expect to complete that project. I will also say that besides having a core EHR we have been, as we have been thinking about our digital strategy. We have been thinking about expanding, if you may, our core EHR platform by adding three additional platforms that work fairly closely with our core EHR platforms. I would say those are our mobile platform where we are able to add mobile apps that work well with our EHR. We have added a location services beacon/ IoT platform and we have added natural language processing NLP platform and all these additional new platforms that I mentioned are all hosted on AWS. So, we have sort of a hybrid data center model going on. I will say that culturally we are kind of like a product shop within a hospital, if you may, because we have a full stack of as you can imagine we have developers, architects, designers, analysts, testers. So, we are more like a software shop that way more traditional software shop within the hospital. That’s what our tech stack looks like. In terms of your second question on governance which by the way I want to say that’s a really key question and sometimes often overlooked. So, I congratulate you for asking about it. It’s probably a session that I can talk forever on, but I will just summarize by saying that we take governance very seriously. We have four components to it. We have a steering committee with about 24 members that owns the strategic plan. The second component of our governance is a fairly straightforward way for anyone to make any IP requests. In fact, since we have put that new process, we have seen about a 300 percent uptick in the number of distinct users that make the request. The third part of our governance process is a robust upfront triaging system which is really where we put our smartest people, I call that part of the secret sauce of governance. In the final leg of our governance is a transparent system for reporting progress to stakeholders. I think the reason is critical because the governance process allows us to what I will say establish trust, transparency, and accountability and I think those three are prerequisites before one can launch innovation and digital solutions.

Paddy: Right. So, you mentioned digital strategy in my podcast. We talk a lot about digital strategy and digital transformation. Our research and everything that we see in the market and all the CIOs and everyone else that I talk to suggests to us that healthcare is still in early stages of a digital transformation. The definition of digital varies from health system to health system. So, can you talk to us a little bit about how you’re approaching digital and what are some of your key priorities. You talked about a couple of those earlier but how do you treat digital differently.

Manu: When I think about this, I can go back to like what is digital innovation. Is it the development of new capabilities or is it really something else? And I guess you know different folks have different opinions about it but where I personally land on this, I think that pure new capabilities I think emerge from real science or pure science. And I think you know I call them innovations and discoveries. To me digital innovation is an applied science. It’s about making connections. It’s about finding new ways to perform existing functions. It’s about seeing a business problem from a 360-degree view. Understanding the people process and technology components that are playing into it and then really creating a solution that leverages either new or in fact just sometimes traditional technologies in a new connected way which to me is the innovation part of it. In terms of how we handle it internally we don’t necessarily differentiate between the innovation stream and the more, let’s just say, conventional stream. We have one common process for our users to engage IT regardless of whether they are looking for a traditional or a digital platform. Like I mentioned earlier we are a common triaging process that really acts like a matchmaker if you may between business problems and the right technology platform whether that’s a new innovative technology or existing technology. And I would say that organic growth of innovation within our otherwise product-oriented development shop has served us fairly well and that’s how we see ourselves continuing down this journey.

Paddy: So, the way you’re defining digital innovation as you take emerging or existing technologies or traditional IT, you find new ways to do things that are better faster, cheaper, better experience. Is that a reasonable summary?

Manu: That’s a perfect summary.

Paddy: Very good. Now are you also the Chief Digital Officer by definition? Is there is a formally named title or is the CIO also the Chief Digital Officer at BIDCM?

Manu: We have one unified team I would say. So, the quick answer is no we don’t have a Chief Digital Officer. We have what I would say one unified team all under the CIO. So, when I say unified team, I am including our core operations, we run a 24/7 hospital after all, as all likely needed to support that. It also includes our entire EHR product stack that we talked about a little back. And then it also includes our Center for IT Exploration that we refer to the top of the call which is really our innovation arm. It’s all under one team, one triaging system, one way to engage, one budget if you may. And a lot of the staff that works into these three streams intermingle and sometimes move on from one role to another quite a bit. I would say Paddy, that’s a conscious choice we made it to do it this way. Sometimes doing it this way may seem like it can slow things down. I know a lot of organizations have taken a different sort of part to this where they have created the center of innovation sort of outside of operations. We chose to not do that. We chose to keep it in operations because I believe in the long run that’s more scalable and plus the solutions that come out of the innovation side. If you may get absorbed better in the operational side and are sustained longer and in a more informed way by our staff when you do it in a combined unified way like I defined.

Paddy: Right and I think you are a little bit unique in that regard because as you pointed out what I see in the broader world of health systems is that the innovation group tends to be a standalone unit at least among the larger health systems. The digital and the IT function are kind of rolled into the same role. Is that also your observation? Do you kind of agree with that statement?

Manu: Yeah that’s exactly right. So, we thought hard about that and made the choice that we made for the reasons that I explained.

Paddy: Right. Now let’s talk about digital strategy itself within your health system. What we are seeing in other health systems is that there’s multiple models for driving digital strategy. At one end we have a model where they say our EHR system is our digital strategy. So, you know fill in the blank with whatever EHR vendor name you want to put in it. That’s their digital strategy. At the other extreme we have enterprises that are taking a step back and looking at an enterprise level digital roadmap and prioritizing all the initiatives and aligning them with enterprise level strategy and goals. What we see mostly though is that digital is being driven as a portfolio of individual projects each of which is evaluated on its own merit and its own unique business case. Do you agree that is the predominant model today? Where do you see yourself in that continuum as far as an overarching enterprise level digital strategy is concerned?

Manu: Yeah, I mean I would talk to this question just based on my own experience. Like I mentioned, our story starts with having our own EHR which as we talked previously. We have paid a lot of attention on governance, a common governance process and on a conscious effort to build more platforms that support that core EHR and I feel that with those three things in place a core EHR, a strong governance process, and an expansion of platforms. I think beyond that to me the journey needs to be driven by the priorities of the hospital and so in our example our President Peter Healy and his senior management team which I am a part of. You know we every year sit down and set up our operating plan and whatever comes out of that is essentially our charter for our digital strategy that we build off. And maybe I can best explain this with a few examples. For example, one of our operating goals was to maximize our existing capacity. And one of the things we found out was the discharge process could be expedited for example if we were able to notify physicians of the availability of a result that perhaps is the last thing waiting between a patient getting discharged. So, you know we could use our some of our mobile platform in that particular example to build a mobile app that notifies the clinician smartphones or even their watch that results back and with one click. Given that it’s a browser-based system takes them into the EHR where they can put in the discharge note order and then the patient can be discharged. Why is that important because that maximizes capacity, it doesn’t make someone stay here any longer than they need to. Another example is you know in hospitals like ours the operating rooms are the biggest focus of both resources and opportunity. And I don’t know if you know how operating rooms or blocks are utilized but it basically boils down to every surgeon having sort of a slice of time if you may where they can book their patients into. Now, if a surgeon is not going to be there say two months from now or let’s say two weeks from now because they’re going to go on vacation or something and they say release their block. How do you make sure that release block gets rebooked? And there’s a rare use case where we thought that a simple mobile app that could notify unexpected availability of blocks to other surgeons could help improve our backfill rates and we have seen that. So, these are just examples to sort of support the point that to me the digital strategy is having common governance, having common platforms, having a dedicated team but then really be solutions driven based on the top business priorities.

Paddy: Right. These are great examples of the real bottom line benefit and in the short term it’s not like you’re waiting for a long time to spend something up and waiting for the results. So, these are actually a great example. So, Manu one of the things that I do in my podcast is something got a lightning round where I ask for the top of my thoughts from my guests on a handful of emerging technologies. It doesn’t have to be necessarily what’s happening in your own health system but a general observation around the state of adoption of technology. First one is something that we’ve already talked about at length in your context. But I’d love to hear your views on what you see as adoption rate in the health system marketplace – cloud computing.

Manu: I mean you must have picked that up I’m favorable towards cloud computing. I think it’s cost effective; I think it provides flexibility. I think it can be more secure than on-prem if it is designed correctly and like we experience in our journey with AWS, it actually opened up to new innovative platforms to surround our EMR with. Prior to having this flexibility, I think you know if we had to try to do solution, we would have to set it up for the new servers and it would take months. Now we can basically try things within the click of a button, we can switch it off when we want to. So, I am overall very positive about the cloud as long as it is given the due diligence that it deserves to make sure that it is designed correctly and securely and use productively.

Paddy: Do you feel like your peers in the health system space by and large embraced cloud and they’re well on their journey towards migrating significant portions of their workload and getting into a hybrid kind of a model?

Manu: I don’t know. I mean I think there is more usage of SaaS solutions in the healthcare sector whether it’s cloud ERPs or service centers or call center solutions. I think the opportunity still exists for migration of on-prem assets to the cloud in the healthcare sector. I especially talk about this with anybody I can, and I would be a proponent for it.

Paddy: Right. Let’s move on to the next one. So, a lot of talk around this artificial intelligence, machine learning. Where are we in the healthcare space?

Manu: Well you know I think there’s a lot of attention with AI/ ML in the sort of image recognition. But I will say that I think there’s a ton of potential of AI/ML in what I call core operations and in reducing administrative burdens on clinicians. I’ll give you just a few examples that we have had some success with here. As you may know, before you got to have a few forms in line. So, got to have your consent form, you got to have your H&P form which is your history and physical form. And these forms come in all shapes. They’re called forms but they’re literally not forms. They’re just basically blocks of information that can come in and faxes and from different places and sometimes hand carried in. So, this is great for machine learning. Machine learning does well in recognizing you know images and NLPs are very powerful for this. We have implemented using TensorFlow for solution that runs on AWS with consent forms. We’ve been using a solution from Amazon called comprehend medical to detect H&P forms and they are now in production and they are saving hours and hours of time of nurses and other skilled resources that would have otherwise been looking for these forms. You know that’s an example of how I see machine learning playing a role on the operational and administrative side of it.

Paddy: Is your data and analytics organization are also part of your function or is it a standalone function?

Manu: It is part of our function.

Paddy: OK and clearly as part of your Amazon relationship it sounds like you are taking advantage of all of the advanced analytics capabilities that they offer as part of their cloud services. So, do you feel that in addition to just the cloud computing aspects of it which is scale and efficiency. You’re also getting the incremental benefit of all of their advanced tools if you will the analytical tools?

Manu: Yeah. It allows us to sort of try quickly, fail quickly, move on quickly or try quickly, succeed quickly, move on quickly. Because it has lowered the barrier of entry into bigger stacks of platforms which are otherwise not easy to get to. So that’s the beauty of it.

Paddy: Right. OK. The next one on my list – voice recognition.

Manu: Yeah, I think specifically in medical context voice recognition has great potential. So, understanding not just words but inferring context like taking out diagnosis, problem lists, and prescription lists with a high degree of confidence from unstructured data to convert that into structured data that can in turn then be fed into ML models. I think voice recognition and we are seeing solutions in that space. I think that’s to me I think the awesomeness of its potential because it can take this tons of knowledge that is locked in unstructured data and make it more structured.

Paddy: Right. The last one on my list here is something that’s kind of was in the news a bit in a big way last year but it seems to have fallen off the radar a little bit. I love to hear your thoughts on it – blockchain.

Manu: So, I’m high on cloud I don’t know what to say about blockchain. I am yet to see, I mean I’m Obviously open minded about it, but I’m yet to see a strong healthcare use case. I think I’m standing on the sideline observing it more that I’m not actively pursuing it as that’s how I would say.

Paddy: Yeah, you and several other CIOs I must add. OK. We’ve covered a fair amount of ground so far Manu and we’re almost coming up to the end of our time. Let me ask you something. What do you see as the top three challenges for a typical health system CIO today?

Manu: You know Paddy I can speak in context of my world. I think in my world demand and supply of IT services is one of my top challenges I would say. I think in general I think there’s three times the demand of what we can do. So even though we do hundreds of projects it just feels to me like it’s not enough. In the healthcare sort of business environment, you can’t expect to just keep adding resources to address that. So how do you enhance your productivity. How do you get more smarter in the way you work is something that I am constantly finding myself thinking about. We are doing things like trying to you know in IT worker space you know context switching is very costly. Programmers like to have dedicated space where they can work on things without disruptions that enhances focus. So, we are looking very carefully at how our unplanned work sometimes seeps into our work streams and I’m a bit of a fan of The Theory of Constraints. So, I try to look at it from the constraints perspective as to understand where our bottlenecks are. And we use methods like the Kanban boards to manage our weeps. And we look at it to try to synchronize our work across the department. All this to say that the challenge of demand and getting productivity enhanced is I would say at the top of my mind. Other two challenges, I would go back to interoperability. Save handoffs of patients between healthcare systems are when they arrive at a new place remains a challenge. And cybersecurity I would say is the third one. I spent a good portion of my time with my very capable CISO to deal with cybersecurity seems to be almost always under some kind of an attention or the other. So yeah those are the three demand, interoperability, and cybersecurity.

Paddy: All right. Very interesting and we didn’t talk much about cybersecurity. There’s probably a whole separate topic for a whole separate conversation. But as far as the first one is concerned, I imagine there’s a lot of technology consulting firms out there who love to come in and help you and I’m sure you have a lot of calls from them anyway. Just to round out our discussion today, want to talk about the digital health startup environment. You know we follow this a lot; this is kind of what we do as a firm in our digital transformation advisory practice. You know the broad numbers are that VCs are putting in about 10 billion a year give or take on funding digital health startups. The vast majority of them don’t go anywhere. Of course, the past week we saw a couple of big digital IPOs come out so that’s like a shot in the arm for the industry. What’s your sense of, what that whole landscape looks like today? Are you seeing real innovation coming out that you think you can integrate into your environment and accelerate your innovation. And if so what’s your advice to digital health startups looking to partner with the BIDCM in your digital journey?

Manu: Yeah I mean I think I get this question quite a bit. What I find myself end up telling folks is something actually that we talked at the top of this podcast which is paying attention to workflow is not enough to have a smart solution. It has to be put in the context of a very fast paced work pattern. The opportunity to impact lives is a short window where you can impact the decision making of a physician. So, I think thinking of the workflow part of it, I would suggest is one of my top advices I would also say that the less data you need for your solution the better. It’s an unfortunate advice to give. Data should be more fluid than it is today but that may or may not be the reality. So, if your solution depends on tons of data integration with major vendors, I think that’s a red flag to me. And then finally I would say that people and process matter as much as technology. I think innovation is not a technology only domain. So, creating new technically smart solutions is not the only way one can disrupt the space, the process side of it, and the people sign up for it it’s just as important.

Paddy: Yeah, I think that’s well said because people tend to conflate cool tech with digital health innovation and the people process the cultural side of it, the workflow integration. This is all the you know in the balls of how you actually make this thing happen very often. That tends to get overlooked. You also mentioned interoperability which is a big challenge. Even though in your environment you’re operating your own EHR system. So, for internal innovations I imagine it’s a lot easier for you to roll out innovation. But for someone else to come in and integrate with your environment I imagine it’s a little bit more of a challenge.

Manu: Yeah, indeed. I mean it is easier for us, almost think of it as a differentiator. And we do have APIs and we do integrate with a ton of _____[unclear] systems when we need to do so but I agree with everything you said.

Paddy: Fantastic. Manu it’s been such a pleasure speaking with you. I learned a lot and I’m sure our listeners are going to find this to be a fascinating conversation as well. Once again thank you so much for coming on the show and I look forward to staying in touch with you.

Manu: Thank you Paddy. Thank you for having me.

We hope you enjoyed this podcast subscribe to our podcast series at www.thebigunlock.com and write to us at info@thebigunlock.com