This transcription is provided by artificial intelligence. We believe in technology but understand that even the smartest robots can sometimes get speech recognition wrong.
Today on This Week Health.
We never talk about wellness data. Or we never talk about patients who are healthy. This data is quantifiable and is really a good predictor of a number of health metrics, whether it's behavior, whether it's mental health, whether it's physical health. And today our health system in general is not necessarily taking into account any of it. And that's where I think the future is.
Thanks for joining us on this keynote episode, a this week health conference show. My name is Bill Russell. I'm a former CIO for a 16 hospital system and creator of this week Health, A set of channels dedicated to keeping health IT staff current and engaged. For five years, we've been making podcasts that amplify great thinking to propel healthcare forward. Special thanks to our keynote show. CDW, Rubrik, Sectra and Trellix for choosing to invest in our mission to develop the next generation of health leaders. Now onto 📍 our show.
All right, today we are joined by Nasser Nizami, CIO CDO for Jefferson Health out of Philadelphia. Nasser, welcome back to the show.
Always good to be on your show, bill. Thank you for having me.
Well, I appreciate it. We caught up with each other when I was up there in Philadelphia. Went to that little coffee shop outside your office. You guys are, you guys are meeting again at the office, aren't you? You're bringing people into the office these days?
That, that's right. So we are in what we call a hybrid now. So two days in the office, three days remote for most of. Team members. Of course there are team members like in clinical engineering, in desktop support, telecom support that have to come every day of their job. But overall on balance, we are two days in the office. Three days remote.
Well, we'll, we'll get into that and some other topics as we go through. My listeners want me to make sure I ask this question first for everybody, and I've had you on the show a couple times, so they may have heard this before, but I'm gonna ask anyway. Tell us about Jefferson Health.
Right. So Jefferson Health, I think last I, I, I was on your that podcast was in sometime in 21, early 21. Yeah. At that time we did not have a insurance plan. So now we are in three businesses. We are an 18 hospital health system. We are a U two campus university, and we have a peer plan now. So altogether we are a 10 billion organization. With about 43,000 employees and three businesses.
Well actually when we were talking, you guys were at the early stages. You, you'd made some announcements on acquisitions or even had made some acquisitions, but it was still highly not integrated yet. So I imagine over the last two years, you probably have done a fair amount of work on Integr.
So these are actually newer integrations. So we acquired, so three hospital health system called Einstein Health in October of last year. And then in November of last year we acquired Health Partners Plan, which is a peer system.
So, and I may have my months reversed, but both within, within a two month period. So Einstein is primarily on a Cerner emr and they have multiple revenue cycle systems. Jefferson as you is primarily an epic shop. So we are in the process of integrating your right. Things like mpi, radiology, lab interfaces, so on and so forth.
Some of integration work has already been completed, although we continue and it's, it's a journey. As integrating EMRs even today is not an easy task. It takes a. Partner, peer business, insurance business. We are Newman like myself, my team are learning about it. We have a great team IT team at Health Partners Plan.
We are fortunate to have some really talented individuals. We are working with them to just pick some low hanging fruits in infrastructure and cyber security and data center management and so forth. Of course, their suite of applications is totally different from Clinical Suite of
applic. So Philadelphia is interesting. Are you guys primarily focused on the urban area or have you pushed out into the mainline area of suburban areas around Philadelphia?
Oh, no, we, we absolutely serve greater Philadelphia area and southern jersey, so we, we absolutely are on, we have hospitals that are outside in Abington and other areas and certainly we have hospitals in Washington Township and Cherry Hill and Stratford and new.efferson today and going into:
I'll share with you some of the enterprise priorities and then how we are working within it to support them. So enterprise-wide access. Cost cutting as healthcare in academics are facing some significant headwinds. So just efficient operations is a big initiative at Jefferson. Patient engagement, patient experience. So these are the top enterprise initiatives supporting those initiatives.
We have our what we call tongue in cheek digital funder strategy. We are working on AI automation to again manage efficiencies and cost. We are working on integrating different businesses. Einstein, I mentioned, health partner plans. We are moving to cloud. Cyber security is an ongoing thing. So some of the initiatives are top business plan and then within it are, we are structuring structured to support these initiatives going into 23 and beyond.
Wow. NA sir, I don't need to go with that because you. What, what's, one thing that's interesting is we, we have financial pressure, but it doesn't sound like the number of priorities has gone down. It sounds like the number of priorities has gone up and continues to go up.
That is true. So here's the thing. There is so much room to improve operations, make it more efficient. The good news for Jefferson is, at least for the foreseeable future, We are not going to be in any, we are not looking at any implementation.
So up until last year, even this year, we were on the back to back major implementations of E R P EMR student system packs, so and so you name it, right? I think at Jefferson we are over the hump now. So the future is really good because we have now for the first time in like five years or more. Time to focus on optimization.
We always were working on optimization and improving efficiencies and so forth, but with large scale implementation, it's difficult. Whenever you are up against that deadline, your teams, yourself, everyone is focused on that goal live. So, so now we are on the other side of major goal lives, at least for time being.
We are focused on optimization and even with cost cutting and everything. Look, the, the initiatives, just simple initiatives to streamline, let's say. And periop processes. It's a pretty major initiative. We have a huge, short, short of nursing. We are launching a virtual nurse initiative to solve that problem.
Right. So you think about we are paying premiums. We are, we are paying for temporary staffing in one of the solution for both cost containment and the staff morale is . It's not something that we have done in the past. It is going to require significant technology. Investment and adoption, but at the same time, if you're successful, it's going to help us with our cost containment and the staff retention and just the staff morale perspective.
So, you're right. There's no lack of initiatives even in an environment where there is restricted capital.
as I'm listening to that, the nice thing about having the foundation in place, which is what you're describing, you're like we've done the ERP we've done the EHR implementation, the major EHR implementation integration.
Once you get that integrated stack, you get the flywheel effect, right? You can make a little change that has a significant impact across all your hospitals instead of when you have, when you don't have an integrated system. And most of our listeners are gonna understand this.
When you don't have an integrated system and you go, all right, we're gonna make this change to how we handle supplies across all these hospitals, a lot of times the IT team would have to step back and go, yeah, that's four projects cuz we have to do it like this for this hospital like this, for this hospital like this.
And so you get these, you get that flywheel going where you make that one change and boom. Now all of a sudden you're doing something different across the board.
You said it very well. I say this, that if you have not lived in a world of singles integrated system, you just don't know what you're living now.
So And flywheel effect, it's exactly the right word because a base, a standard base of set systems just takes you out of the integration interoperability mode and allows the team, my team now to focus. on Where the value is. So we are not talking about simple and your example one bpa one alert takes, look, when I joined Jefferson, we had nine different EMRs.
Okay. One simple alert was a initiative across multiple hospital, multiple teams EMRs. And it is just Having a standard based assistance really makes life easy, but more importantly, it truly allows for innovation to happen.
Right, right. Yeah. That's single platform. I wanna go in the automation direction.
It's the, I mean, I have AI automation, I've wrote down a bunch of things. Cyber I'm gonna avoid just cuz I, I don't put My guests in that position to talk about cyber all that often. Not because it wouldn't be interesting, but because other people can listen to this show as well and pick up on your cyber and, and in cloud.nversation a lot heading into: here mostly started in it. In:
And where we enabled processes in pharmacy, in supply chain, in revenue cycle, limited in clinical area. Some in, I, especially with onboarding of employees and so forth. The biggest bank for our buck, the success that we are seeing and we are focusing on are two areas, revenue cycle. And it, it doesn't mean that other areas don't have opportunity, but what we have found is that the vendors that are out there with solutions are most mature in revenue cycle.
In it, in it perhaps. We are sort of masters of our own destination, perhaps in other areas we have to rely on expertise from other departments in it. We, we see a lot of opportunity and we still have the issue of silo systems, HR system, not talking with our active directory and so on and so forth.
But revenue cycle really is the area where we are seeing roi, like clear ROI and where we believe that. A lot more opportunity, especially with our peer plan. Now, now that we have a peer plan and there just, I'm learning about a number of processes that are incredibly manual. We are literally, people are using phone calls and faxes and just very broken processes, if you will.
So there are technologies that that help there. But given that these are actually multiple different systems, I believe that the automation has a, a role to.
So when you talk about revenue cycle, are there specific cases like the submission of claims or I, I'm
at a loss actually Exactly the prior authorization submission of claims management of denials and so on and so forth.
So there are about 12 or 13 use cases that our teams. Identified. And, and it's interesting because again I'm not an expert in this area, but as I learned that each pair have different processes. Yeah. For things like authorization, things like know, denials and our teams today are just have to learn.
And they're experts in knowing, oh, this is this pair and this is this insurer, and this, this is what I have to do and this is, I have to verify this is I have to check eligibility, so on and so forth. And in many cases, these steps are the same repetitive task. You click here, you click here, you check the status here, you click here, you click here, you submit something here, right?
And that's where automation really I think provides a ton of value. To learn about these different I steps. Right. And basically automate them. Cause there's, there's some decision making, but it's almost like calling decision pre. So so about 12 or 13 use cases to answer your question.
it's, the decision making is generally rules based decision making. Are we seeing aI and machine learning start to get integrated into some of that automation as well.
I would be honest, and I'm hearing a lot, I'm not seeing a lot, right? So every vendor that we go to talks about AI. But when you dig a little bit, either AI is not really there and there're really talking about some advancing Avis or the AI does not have ROI to it. Now I do believe that AI will ultimately play a role here because anywhere where someone has to think and take an action, AI has a role to play. I firmly believe that. Yeah. Right. But the technologies that I have seen all claim to have AI and machine learning, but I have not seen them practically make a difference.
I'm with you on that. So some of the use cases I've seen AI actually delivering an ROI are around imaging.
Hundred percent. I can talk about imaging all day long. Not around automation though. Yeah, yeah. In the sense of revenue cycle automation.
Yep. No, and that's, I was trying to transition to AI and just say, all right, so as people are looking at AI, imaging seems to be one area where we're seeing significant movement in that area. I mean, we can actually break that down into multiple areas. You have radiology, cardiology. X-ray imaging is interesting cuz I'm seeing AI models on top of that to do, to do reads, and it's all clinically what's, what's the word I want to use?
It's to assist the clinician. It's to reduce the cognitive load on the physician and say, Hey, here are three things you might wanna look at. We've run it through the AI models, we've run it through the machine learning models, and these are the things that are of interest to us. Now, the clinician just comes in, the read becomes a lot easier. They're going. Yeah, that's something, that's nothing and away you go.
Radiology and imaging and, and look, even at Jefferson, if I were to say what is the most advanced area there are many, but radiology certainly is leading the back I think. And the reasons are straightforward in the sense that we have the data because we have tons of images, number one. The deep learning technologies are you you know, technology is there, which can be easily, not easily, but train on images. I'll give you one example and we can generalize this because we are using the solution today where we, our brain scans are uploaded to one of the cloud vendors that runs their AI algorithms and detect any patients with the high risk of a stroke.
And if their AI algorithm detects a patient with a high risk of a stroke, they real time sends a. To one of the physicians and the physicians have apps where the message pop up and they can take action in real time. Right. So the, the think about the efficiency now that this has bought us, and this is being repeated for a number of other use cases.
I gave you example of a stroke, but this is the, this is a use case that you can just multiply for different, anywhere you have an. That someone has to read, theoretically, there can be an AI algorithm doing the read, if you will. Right. So radiology is the area where we are seeing most I think use out of it.
There are other areas as well where like sepsis is an area where we actually have some real data in our quality team provide a model with our epic The, the one thing I would say is that I think there are two things in my experience with ai. One is having a model that is the good model, whatever the definition is, you can define whatever that means, but the challenge today in healthcare is how do you implement the outcomes and make it actionable?
So when an alert is going out, who is number one? Who is. Who's taking the action? Is it a nurse? Is it a technician? Is it the patient? Is it the doctor? And then what are they doing? It's not good to just say, Hey, I think that this patient is in a high risk of sepsis. That doesn't tell, that doesn't mean anything.
And even if it's say, Hey, look, 50% probably of this, 30% of this, 40% of this is still not good enough. You have to be, you have to have something that is us. By the clinicians or whoever is, and a process around it. How do you close the loop when you know the story example I gave is really cool because it's a, it's a closed loop process.
You at every stage, you know that, alright, someone is uploading, someone is detecting, and a person is, you know who the person is, who's gonna get the alert and then take action in, in an inpatient environment where you know you're in ed, let's say, or other area. It's, it's not as simple. So it took a long time for us, but we are in a really good spot because we have sepsis deploy, we're using it.
And I think the more, a more where I see most adoption is in non-clinical areas again, it comes to mind. Cybersecurity is a great example where we are using AI algorithms but we are using it in. Start and so forth. So, revenue cycle as well📍 📍 In:
📍 📍 it's interest with ai. Like clinically adjacent areas could have an impact on clinical. So for instance, we're seeing companies like Lean Toss being used Yep. For scheduling of the surgery centers and the, the various ORs and whatnot. And that's driving a significant amount of efficiencies. It's utilizing the clinician's time a lot better.
It's making sure that the, the the rooms are ready when they need to be ready, when the physician shows up, and that kinda stuff. That's an interesting area. I'd love to see AI applied to. Discharges communication around discharges. Cause I, I remember we were looking at discharges at our health system, and the number one thing that kept people, and it, it could be up upwards of like six to eight hours additional in the hospital, was finding the right person to do the discharge.
And so we had started, I I, we didn't finish it, but we, we had started, I left when, before we finished it, we were starting to look at you. Why was it that we could have such inefficiency in that area where somebody was in the hospital and we couldn't find the right person who could actually discharge that patient, and they would actually take up room in that, in that facility for another six to eight hours.
I mean, I, I could see, I could see. If I were doing that analysis today, it would be nice to just take all the data, process it, look at the machine learning what are the trends and analysis, and then apply some sort of models around that. I'm sorry, this is just me brainstorming. I'm trying to build a new product here with you on the show.
Sorry. But yeah, I mean, there's, there's adjacent things to how hospitals operate that I think AI will start to really see a, a greenfield opportunity to help us be more efficient on all the things we do on a daily basis. Let's talk about, let's talk about cloud a little bit. Um mm-hmm. . The, the, the cloud move is interesting.
What's driving the cloud move for Jefferson? What's the motivation behind it?
Look it's interesting because we were, myself, my team were always fans of cloud, but not necessarily in the sense of cloud first, or we must move to cloud. We took a more opportunistic view and said, alright we'll look at inflection points when we are looking at new products, major upgrades on and so forth, we are going to entertain the idea of a cloud.
And if you roll back six or seven years ago, there was just a massive, I guess movement to move to cloud, no matter. What we have learned is that not all solutions are necessarily meant for cloud number one, and that cloud is, at least in our experience, is not going to necessarily is not a cost saving strategy.
If anything, it's at best it's cost neutral. In most cases, it's going to add cost, right? So, so that background, we were moving our assets to the cloud slowly And I'll say today, I would describe as probably about roughly 70% OnPrem, 30% in the cloud roughly. And that's, that's probably where more in the cloud on the university side, probably, maybe it's like 50 50 and on the health system side is more 70 30.
But we believe that now the technologies are there, cost has come down. So, and, and as we have grown we have business demands of scaling up very quickly and responding very quickly, and at the same time scaling down very quickly. Hey, how do we, we with Jefferson we work with a number of.
So we would bring in the company, we do some experimentation and look like many experiments. Not everything is success. Sometimes things fail, right? In OnPrem environment it's very difficult and expensive to reuse the assets. So you don't have the good concept of scaling up a scaling. Now cloud provides that.
So those cost efficiencies somewhat scalability, speed to market. We knew that, but I. There were two things that happened that really pushed us. Number one, during Covid, it became clear that almost all cloud vendors did really well. Whether it was Office 365, or Zoom or Teams, or our learning management system, everything is scaled really well, number one.
And secondly, of course we moved, we started the conversation with us being hybrid. So we know that our workforce is not coming back to the office in any time future. Again, I'm talking about non-clinical workforce. We, we are very much a, a health provider system. 80% of our workforce comes through the office.
And the third thing was around business continuity and disaster recovery. Our calculation is that for us to provide the business continuity in cyber security or so disaster recovery or cyber security to the level of big players like Microsoft or Amazon is going to be very expensive. We can do that.
We can stand up data centers. End of the day, we'll always have two data centers. We are not never gonna, and they're gonna be probably close proximity for us to manage them cost efficiently. So we made a conscious decision to partner with Microsoft and move Epic in a number of other assets to Azure Cloud.
That is a two year journey and our goal there is to flip the numbers from 70 30 to 30 70. There are still assets that today, like facts come to. Are going to be local. That may change as we progress and maybe two years from down the road, we may think that should be on the cloud as well. But today our goal is to really be in the cloud and it's it's a, it's a significantly different mindset and sign significantly different way to operate.
You are very familiar with this model. It changes how we think about cybersecurity. It changes how we deal about a number of things within.
I mean, the nice thing is you guys have been so busy integrating all those other systems that we actually can look now at all the mistakes people made in moving to the cloud and go, woo, man.
We can avoid those. Avoid, and I'd love you're looking for agility. You're being very selective in terms of understanding the cost models, what works and what doesn't work. It's really, I love what you guys are doing, you're in Philadelphia Access to care. We've talked to you, we talked to you and Dr.
Klasko before about how access to care is so critical. How are you pushing that forward? What are the initiatives around access? So, access,
again as I mentioned, this is a big initiative at Jefferson at multiple levels. So this has certainly technology pieces to it, but non-technology pieces to it as well.
So how do we get in the communities that we. That don't have good access to things like transportation or don't have healthcare facilities near them. Earlier, right before your meeting we were, I was with a group of physicians brainstorming cancer screen. So during Covid we outfitted these mobile events and the full technology status suite, wireless app, a everything, and deployed in communities for Covid vaccine.
And now we are using the same model for things like cancer screen. So again, there is a, there is a process piece to it. There is. Human element to it, and there is a technology piece to it. So on the technology front we are really pushing the limit here. And really our monitor for the last 12 or 18 months is, has been how do we scale, right?
So we have the mobile technology, we have the patient portal. And the question for us was, alright, how do we increase patient engagement and how do we scale? So we today are a 60% mark of all our active patients having our My Jefferson help, which is our mobile total account, that roughly number is 900,000 patients.
Our goal is to get to 80. 80% in next 12 to 18 months. So we are ahead of the curve here. We are very proud of it, but there's a lot of work to be done. Now the things that we look at are things that are meaningful to patients. So can I schedule my appointment? Can I do each check in? Can I ask a question?
Can I request a medication refill cannot pay online. So in, in the. Initiative today is around templates and making provider schedules available online for patients to schedule. And I think we are on a path to have, we have great adoption in primary care, which is the easiest one for any type of scheduling.
But the specialist can be tricky as because you have to follow certain decision trees and you don't want the wrong patient to show up at the wrong. Yeah. So there's a, but our clinical team, especially in our medical group, has done some really outstanding work and lot of work, hard work, I would say, to create decision trees that we are not taking and implementing for 50 or so specialties.
Okay. And that's where I think one of the big, one of the areas, the biggest bang for our buck is going to be. Another area I'll give is our phone system where we talk about ai. We implemented a cloud-based phone system. We turn on sentiment analysis on that. And we uncovered a couple of very interesting problems just by AI listening to the calls or, and Yeah.
So using ai we know what calls have a positive sentiments, what calls have negative sentiments, and that gives a very good data point for our colleagues who run the scheduling center to then focus on a group of calls and say what's going on. And we are learning some very interesting things in just last two or three months, but this is another initiative where we believe that using AI we can build prompts and intelligent call, call routing and really direct our patients to the right prompt. One of the things that we have consistently heard from our colleagues in, in scheduling, central scheduling is that, I don't know, 60, 70% of a very large number of calls are how to how to questions how do I get to direction?
Can I have the, or insurance verification, things of that nature. Things that we can. Automate in things where we can route patients to a good spot using some intelligence in, in the . So just couple of examples on how we are thinking about technology and using technology to improve access experience for our patients.
that's really fascinating. As, as we increase access. One of the things I just saw is Cleveland Clinic is now charging for message. So some messages that go to clinicians in their inbox, right? So they can go into their MyChart and they can say, Hey, do a refill, and there's no charge for that. Or they can say just something simple and there's no charges for a whole host of things.
There's no charges. But then there's the, the other thing which is, oh, by the way, when I got home, I have these new symptoms and these new things are going on, things that would. A clinician 10 to 15 minutes. As we increase access, are you finding that the clinician burden is the other balancing of that skill that has to be considered and are there things you guys are doing around the, just the amount of messages they have in their inbox and whatnot?
So a couple of comments. I also saw the, the announcement that Cleveland Clinic will be charging, so we are not charging. We do have a high volume of messages, and this only as physician workload. There's no we just, I think three or four months ago, we undertook an initiative to just clean up the inbox, right?
So, for instance a very large number of messages just thank you messages, right? Patients being nice and thanking the physician for whatever service, right? I asked you a question, physician responded, and I sent a thank you message. The volume is so high that this is essentially a clutter so you said a nice thing to do, but essentially it's a clutter in, in, in physician inbox.
So how do we perhaps move that into a separate folder where physicians can go time to time and look at them, but it's not going to be in their inbox. So there are initiatives in at Jefferson to improve the physician experience from an inbox perspective. Right? But look, I mean, simple answer. Any advice, any question?
Every question, every advice is someone thinking, whether it's a physician or someone in their office. So there is certainly validity to the fact that this is a work for someone. It could be a nurse, like, so not all messages are answered by the physicians. Typically, someone triages the message and then before it gets to the physician.
But even that, someone is typically a nurse. Someone in the office and I have to believe that after a certain volume there is a good case to say how do we charge for the service that we are providing. We are not charging anything.
When I read the article, it sort of took me back and I'm like, oh, I've gotta see where they're going with this.
But then when I read it, I'm like, yeah, that makes sense. It's like if an architect, if you say, Hey, what do you think of my plan? And they look at it and go, yeah, it looks good. That took two seconds, although you still got their expertise. Or if you say, Hey, can you review my plan and submit it for for approval?
Yeah. That's like, well wait a minute. I, I understand you sent this to me in email and you think it's no big deal, but that just represents a lot of schooling, a lot of expertise, and a lot And I think that's how they sort of looked at it and said, how much, how much effort is associated with this?
Without, without. All right. So you've given me a lot of time. I appreciate it. Let me give you a closing question here, and it is let's do a future question. So what technology are you keeping an eye on and you believe's gonna have an impact on healthcare over the next five years? So something that's maybe a little bit more out there than, than the normal stuff we're looking at.
Certainly ai, like, I mean, AI has not put a big dent in healthcare, right? Yeah.
We have, we, we haven't even started yet, right? We we're gonna see at of hype, I think on ERPM in:
Rpm, perhaps in chronic health patients and here and there, but ultimately long term, this is a technology where I believe that is gonna have a much larger, I think, impact on overall population. And lastly, generally look, as a, when we think about healthcare, we never talk about wellness. data Or We never talk about patients who are healthy.
Okay? And the result of data, like I'm wearing a Apple watch, I mean, I have to believe that you are doing something, I'm guessing. But regardless of whether you're logging your health information in an app or you, you have a variable or you are just using a telecon there is a ton of data that is being generated.
Your social interaction on. On social media, for instance, that is this, this all data is quantifiable and, is really a good predictor of a number of health metrics, whether it's behavior, whether it's mental health, whether it's physical health. And today our health system in general is not necessarily taking into account any of it.
And that's where I think the future. is Right now there are some health conscious, if you are a health conscious individual, you can set up your own regimens and you can use three or four different apps in your hospital data, in your lab results, and can create something. But the future, I think, is going to be a lot more integrated, perhaps 10 years from now. But it'll happen. Yeah.ard to to staying in touch in:
Likewise, my friend. Thank you for having me and have a great 📍 day.
I love the chance to have these conversations. I think if I were a CIO today, I would have every team member listen to a show like this one. I believe it's conference level value every week. If you wanna support this week health, tell someone about our channels that would really benefit us. We have a mission of getting our content into as many hands as possible, and if you're listening to it, hopefully you find value and if you could tell somebody else about it, it helps us to achieve our mission. We have two channels. We have the conference channel, which you're listening. And this week, health Newsroom. Check them out today. You can find them wherever you listen to podcasts. Apple, Google, overcast. You get the picture. We are everywhere. We wanna thank our keynote partners, CDW, Rubrik, Sectra and Trellix, who 📍 invest in our mission to develop the next generation of health leaders. Thanks for listening. That's all for now.