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

[00:00:00] This episode is brought to you by Tido. Tido provides managed integration services, EHR, data migration archiving, workflow automation, and 24x7 monitoring to ensure seamless data exchange and improved operational efficiency.

Dealing with capacity challenges in data or integration projects and operational optimization issues, leverage Tido's AI driven solutions to reduce costs and enhance patient care continuity. Ready to elevate your healthcare IT strategy? Visit ThisWeekHealth. com slash Tido today and discover how

Tido's

innovative solutions can transform your organization

Today on Newsday.

it's not just about a person with a technology background pushing a solution through but If it's actually a physician who has the experience and who can see the benefits in the long run of using such a tool, it will make a huge difference, right? My name is Bill Russell. I'm a former CIO for a 16 hospital [00:01:00] system and creator of This Week Health. where we are dedicated to transforming healthcare, one connection at a time. Newstay discusses the breaking news in healthcare with industry experts

Now, let's jump right in.

Thanks for joining us on Newsday today. I am joined by Vik Patel from TIDO Inc., who is the Chief Operating Officer and a new partner with This Week Health. Vic's a seasoned health technology professional with over 15 years of experience.

In his role, he oversees company operations, focusing on delivering innovative solutions that enhance patient care and operational efficiency. I also love that he is the host of This Week in Health Tech with 57 episodes under his belt, where he discusses current trends and developments in health technology.

Vic, welcome to the show.

Thank you so much for that amazing introduction. And for those of you wondering what TDoS stands for, it's this is done. So Sarah, I know you guys put a lot of stress on the culture in this week's health, and we do as well, and that's why, this [00:02:00] is done. We take a lot of pride in getting things done, and, that's It's embodied in our name of the company.

So there you go. I don't know if you knew that.

I didn't. And there's a lot to be done. Even on your podcast, you think you talk about integration response, monitoring, interfaces, care, which is the thing that obviously all your practitioners are going to care about. When you joined TIDO Inc.

in 2007, what were some of those initial That really drew you to the mission of the company.

So you may not, maybe another fact, I'm also the founder of TIDO Inc. I worked in a hospital for a long time and, learned some of the challenges, around getting the data to the clinicians at the right time in real time.

And the other thing I had also noticed was around, challenges in production where we were always relying on the end users to find the issues, right? So [00:03:00] the nurse or the doctor or someone calling the IT and letting them know, hey, I haven't received my results, for this order that I placed two hours ago.

What is the situation? Can somebody look into it? Just some of these things, which you would think should be built into the applications in terms of self diagnosing and such, but no, that's not the case. So I learned some of those things, working in a hospital and that's where TIDO Came along, the whole idea of TIDO and helping develop some of these solutions from an integration standpoint, from data management.

And today, we are a team of around 30 people mostly all engineers. Our integration team, our data conversion, data archiving team, and automation team. And all of our team members are based in North America and U. S. and Canada.

And as we were saying, earlier, we are really good technically, but we also wanted to make sure that, We understand the background, we understand the systems and what it takes to make sure that the patient outcomes are not [00:04:00] affected by system issues.

Oh, it's great. When you think about the purpose by which you created something, you saw a need, you filled it and then kept growing it from there. And foundationally, as you're talking about the news, we pulled up the first article to discuss, which is Dr.

Patrick McGill urging health systems to establish AI governance early. Here's what's super interesting. And he's the chief transformation officer at Community Health Network. He talks about the necessity of establishing governance frameworks early in implementation of AI with health systems to enhance effectiveness and mitigate bias.

They have an executive steering committee. Both leaders from finance, clinical ops, IT to oversee these initiatives. When you find yourself going into an organization and maybe the data isn't super clean and organized or they have a desire to want to use AI but maybe they don't have their governance set up, how do you help them in understanding the prioritization of that clinical engagement as well as what needs to [00:05:00] happen from an ongoing perspective to maybe avoid some of the unattended biases?

Yeah, no, great question. And what community health care, and what Dr. McGill is undertaking, I think it's very important, right? Because everyone's talking about AI and the AI tools. And that's a sexy thing to talk about every conference you go to, or in all the conversations.

But the data governance, I think is the key. And what I usually tell CIOs when we are talking about this is, Hey, do you have any data scientists on your team to help you with this? Because even with our solution, even with our MIDRE solution, which uses AI to find and take action on issues with integrations or application issues, even that, we have our AI models, and so does all the other tools that the health systems are now deploying.

How do you keep up with that? Unless you have the right people on your team, I think it's a huge challenge, so I can see why he's putting in a whole new team, and [00:06:00] I'm guessing they'll probably hire some data scientists within that team to help, monitor the data reports, run the reports regularly and also keep up, in terms of when the solutions are being updated, is the model performance still close to the benchmark?

Are things different now after the update? That's not easy to do. Yeah, I think all the policies and procedures and the guardrails that they are talking about for the AI use definitely make sense. And again, I would say, if you, by the way, if somebody's looking into getting into a career in healthcare, IT, today, as a student, if you are in a university or, and you're still trying to, make a decision on which way you should go, go become a data scientist, right?

There's going to be many jobs available as a data scientist, not just in healthcare, but I feel like in every industry. But if you are looking into getting into healthcare, which. Health systems will always be there, and I think that's a great career if you [00:07:00] want to pursue that.

I agree. And when people ask me constantly about career direction or ideas, or even at what age can take a new career trajectory, and I'm like, if you have a background in AI and or cyber, And really getting both of those.

And there's so many ways to get the education, training, and even certification, if that's what you're seeking, those are the most in demand spaces, in addition to really understanding how the cloud and how all these different ways that you distribute your workload is really important. And you mentioned MIDRE, and I don't want people to lose sight of the fact that Managed Integration, Detection and Response System, you're utilizing AI to monitor and support interfaces because you don't want patient care to Interrupted.

And yet so much of what we've seen in the news over the last few months is the third party element creating the challenge or the problem inside of the healthcare system. How do you make sure that your products are something that people can rely on from that cyber and reliability [00:08:00] perspective?

Yeah, no, great question.

So for us personally, like our MIDRE solution is designed to eliminate the dependence on the end users to find issues, find, production, application and integration issues. It is a cloud solution, obviously. We use Microsoft Azure AI functionality for the backend of the MIDRE solution.

But for us, we make sure that EHRs, is not going into our solution, into our AI modeling, right? The way we have trained, Our model is all based on de identified data. I think that's a key. So that should be one of the questions for any AI solution, in, in terms of, hey, are you going to be using our PHI for training your models in the future?

And if you are, we need to know, and on top of that, if we do give you the approval, make sure it's de identified. The other thing too is in our case, [00:09:00] again, we are not helping the clinician, a doctor, make a decision. It's more for the back end, for the IT to keep, to alert them of any issues proactively, find these issues and then make a decision in terms of, oh, do I need to create a ticket?

Do I need to call a person? And which person should I be calling so that we can take an action as soon as possible? And, triaging and resolution and coming up with a resolution for the issue. So yeah, for us, we have. take in every action that we can so that we are not creating an issue for the health system, as you said where a lot of the third party have been at fault for health systems where, it wasn't the actual health systems network or therefore it was more one of their partners.

So when you think about the ability to establish AI governance early, in this case, Dr. McGill is leading transformation at the clinician level. How important is it That a physician is the one involving the clinicians to either increase [00:10:00] adoption and reduce resistance or to streamline clinical documentation and administrative workflows.

Does having a physician in that role change the dynamic of what needs to happen inside the system for it to be successful?

Oh, a hundred percent. I think if you think about, from a CIO telling the clinicians or the doctors or nurses versus a CIO who is, who was a doctor, who was a practicing doctor or still is a practicing physician, I think that makes a whole lot of difference, right?

So it's not just about a person with a technology background pushing a solution through and it's Hey, we need to use this. It's cool. It's going to help us in such and such way. But If it's actually a physician who has the experience and who can see the benefits in the long run of using such a tool, I think it will make a huge difference, right?

From a doctor to doctor messaging, I would think should be a lot more trustworthy than a technical person driving a decision in [00:11:00] a health system, which, you know what, that brings up another I feel like that always comes up in, in our conversations and even when designing any solutions any products, I feel like there's always, I don't know how many product developers make sure that they have enough clinical background, enough clinicians involved in that design process, right?

In that decision making in terms of how the application will work. I find many times it's technically driven, developers who are not actually. At the front lines dealing with patients, making some of these decisions. I don't know. I want to ask you like, from your previous role, how do you see that in terms of bringing some of these tools to the CMIO or to the clinicians in terms of, hey, we.

This will help you guys, but how do you make sure that the adoption happens?

I love that you asked it that way because I've been in a couple of different camps where my last role, CIO, we did a ton of software development and I always would tell the teams like, Software development, you need [00:12:00] to have the IT team at the table, the developers, clearly, who can iterate what needs to happen.

But you also have to have a business owner and a product owner. And you can choose if the product owner is a clinician or how involved they are in the operations and the business owner really understanding the strategy of what needs to happen. But if those three things aren't true, you're going to create products That are going to be twice as difficult to adopt.

You're going to have a lot of things to fix after you bring these different sprints live. And it's really key to make sure that those three are aligned going in. Reciprocally, I've worked in organizations where the CMIO, it either is or isn't on board with certain aspects. And at the CMIO, if their role isn't to increase physician adoption of technologies, then clearly delineating what their role is and making sure that you have other champions on the clinical side of your teams for.

Those workflows and those products to be adopted, because if you don't do all of the change management ahead of time and have the right executive sponsorship, you're going to spend so much time and energy [00:13:00] retooling the things that need to happen. So I'm a huge fan of making sure that the people using the product are the ones that are helping to develop and deploy it and refine it on a consistent basis.

That was amazing. I can see why you're a seasoned CIO. That was really good. I love now that we get to

help partners and CIOs come together to solution it. So as you think about what we do at This Week Health is bringing people together to find solutions, create communities and solutions. And it works.

You'll walk out of an event with us and be like, I just met 15 CIOs or CISOs or CMIOs. That I now have a relationship with and the sale comes after. You have to create a bond with humans before they really want to work with you. And that's true everywhere. So as you think about having strong, governance in place to make great decisions in an organization, And more and more people are being asked to do application modernization, which can be a road map to better outcomes in the article that you and I discussed.

It's really talking about the significance of application [00:14:00] modernization in healthcare industry and the need to update and optimize digitally connected devices such as the EHR, telehealth, and various operational tools. What's interesting about modernizing these applications is that it's vital for improving care delivery, operational efficiency, as well as meeting regulatory requirements. So what's interesting, though, is The consumer demand for streamlined data interactions and how much it costs to make this happen when systems are outdated and the need to be able to adapt and secure things. And generally there's about four things that need to happen, Vic.

The assessment, migration to the cloud or other appropriate environment, updating the overall technology, ongoing optimization. Too often CIOs and organizations get stuck in the ongoing optimization, the expectation that if I'm going to modernize this application or this environment, it's a one time thing.

And that debt and that ongoing support will sprawl out of control and then you're just redoing the same thing you fixed a couple of years [00:15:00] later. What are some of the ways that you have been most successful with your clients in keeping their technology current or migrating them through app rationalization, archiving, etc.

that allows them to make really sound decisions that are financially feasible in a long term environment.

Yeah, no, absolutely. And like you said, the goal of The modernization is to make sure that the needs of the users, both the internal users and patients, are met and one of the ways we do this.

So . Our approach is using archival solution.

So we work with the health systems on archiving older EHRs, older clinical applications that are no longer being used, maybe they're used. every now and then to look up an older record. But you're paying so much in licensing fees, like you said, and support fees that budget is taken away from some of the modernization that you can do or, purchasing a new tool that [00:16:00] would actually help the clinical team.

But, however, the archival solution should be, such that it's easy to import the data from those legacy systems and also easy to access the data from your new EHR. And our application, we make sure that within two clicks or less, you are able to get to your legacy data that you need.

So we were able to use the latest development technologies and with single sign on from your EHR, whether if it's EPIC or Cerner or Meditech or Athena, we would allow you to single sign on into TIDO Archive and Using the user context and patient context, it lands you right on that patient within TIDO's archive and you are able to see that patient's legacy data and the doctor doesn't have to click around to find the the right information.

Search is built in. everywhere so you can quickly find the relevant information as well within that patient record in the TIDO Archive solution. [00:17:00] So that's one of the ways that we are helping health systems modernize is, you know what, if you are not using this older system, let's make sure you can retire it, right?

here's the thing, I've been, CIO in multiple environments, and you've got systems that are parked in a closet somewhere that are just on technology, honestly, like waiting to die, or thinking about it.

How do you balance the equation or the cost analysis of moving critical applications, even if they're critical old and they're required to be saved, to the cloud for data management? And I think about. Even as technology rapidly enables us to move these archaic old systems, it might be just sitting in a rack somewhere, sitting in a closet somewhere.

How do you walk your clients through what maybe is a leave behind, a lift and shift, or a true transition to the cloud?

Yeah, and this happens all the time, and we have a few active projects where we are actually dealing with this, and one of the ways I think people should look into, especially in the cloud, the [00:18:00] cold storage is almost like it's almost free, at this point, so a lot of the data that you're not accessing every day, it's utilize the cold storage that's available in AWS or Azure or even Google Cloud, right?

That's the way, I would recommend looking into that. And on top of it, again, having an archival solution where it will allow you to easily access the legacy data. But, that is a way to have two different data sets. The hot storage and the cold storage. So if you know that, you know what, the last two years of data is probably the most relevant data that you will need access to, put that in hot storage.

Yes, you will pay a little bit more because it needs to be available right away instead of the cold. But all the other ones, the really old data, save money on it, right? And it's still in the cloud. So now you don't have this old, like you said, server sitting in a cabinet somewhere, and somebody needs to look after.

Yeah. That's the way we are recommending and working with health systems.

I have a thing. I love to joke about Vic [00:19:00] where I call it. You can't make it up because we've all seen so many things in our careers. And I'm working for an organization. When I first joined, there was a door that had Yellow caution tape across it.

I was like, Oh, that's a system that such and such that we can't, we have to keep alive and this was years ago, so it wasn't like easy to archive and it just lives there. And so nobody opens that door. And I was like, really, we can't move this to newer technology. We can't think about a way to do this in a better, different way.

And at the time it just wasn't feasible to your point of an exorbitantly expensive to try to move it over. And so we kept it alive, but it was almost like, You can't just throw a meal or two every couple of days behind the closed door, which really isn't true, as a system was running, but there was yellow caution tape on a door that was like, those are some archives.

Don't go

in there. Yeah, you

have to survive. I thought, that's terrible from a business continuity disaster recovery perspective because This building itself becomes the challenge. And it was just fascinating that was a true statement. I love that we're beyond that [00:20:00] today. And yet one of the other things that you brought up that I love is focusing on practical integration for tangible results in AI.

This is an article from Daniel Barchi, who is the CIO at Common Spirit. And him talking about how you manage the hype. By prioritizing real world impact, and then it's like tuition and education. It's an investment with returns that are ongoing, but really having clear goals and ensuring that it aligns with organizational objectives for patient care improvements.

And he says, treat it like strategic investment, focus on practical applications, start with strong governance, which we've talked about. Tell me a little bit though, Vic, your perspective on how you do have the strategic focus that brings the right efficiencies and helping organizations think about it in terms of an evolving investment.

Yeah, and but can I start like one of the quotes that he has in the article cyber security is like rent You have got to pay it, so I was laughing when I was reading that. That's You know, that's a no brainer, [00:21:00] obviously. And, thank you, Daniel, for that amazing quote. But yeah, AI, as we talked about it right in the beginning of the episode, everyone wants to talk about AI and while cybersecurity, is like your basic necessity, electricity, water, cybersecurity.

I think what he's trying to say is, don't lose your focus, right? Don't just go and every time. If you are at a conference, obviously now every vendor, by the way, has AI enabled products, which I find mind boggling. And, right after the pandemic, I remember going to HIMSS one of the first conferences after, and I was like, What?

Everybody has AI now after the pandemic? What was the change? Why was it so easy? And it's not. When you start asking some of the tougher questions, under the hood, it's actually not AI, and people don't really even understand, many people, the difference between machine learning and AI, so when you ask some of those questions, it's oh yeah, no, it's just more RPA, it's just robotic automation.

It's oh, then why is it called AI, maybe they used AI to create the [00:22:00] documentation. I think, again I know I joke about it a little bit, but this happens all the time. And. For us, I many CIOs, actually, they are obviously very smart.

They know, they're not just gonna, just sign up just because somebody is pitching a product with AI built into it. But I think, back to this article, what he's saying is that, favor the practical AI applications, right? Something that's going to actually help you and maybe, what we have, I feel like our AI application is practical because there's hundreds of interfaces, obviously there's hundreds of clinical applications in the hospital, and similar to cybersecurity monitoring, we are offering real time monitoring off the functionality, off the interfaces, off the applications.

On top of our integration management and support, this is where we came up with the MIDRE solution. And, initially it was for us, but now obviously we are making it available for everyone. was helping our team so much. But yeah, that's what I would 100 percent agree with Daniel in this case, it's easy to [00:23:00] get sidetracked, looking at all the AI tools that are in front of you, but I would say definitely start with the most, basic ones, start with the AI applications that would actually, you can measure the results, right?

So you can actually see the practical results and then you learn about AI, as he says, it's like tuition, you're learning as you go. And then over time, you will be able to, once you have the right team in place with data governance and the policies and whatnot, I think it will become easier to pick out the right solutions.

and integrate them within your health system.

And not go after the areas that are higher risk. There are certain spaces where you realize that you can be cutting edge and still take a low risk approach to bringing ideas into your organization, which is why I love that you chose an article about ChatGPT over prescribing treatments in the ED.

There's an element of what we always talk about, the human in the loop. And while AI is going to be great for things that are routine, AI [00:24:00] will make the same mistake over and over again. As you think about the governance and review of the auditing process inside of AI, humans tend to not make the same mistake over and over again.

It can be more nuanced. And I would think, especially in the emergency department setting, thinking about that quickness, that call to action that really requires the human element, because that's high turnover, high volume. There's not a repetitive type of patient coming in the door with the same diagnoses.

And as you think about the risk and pitfalls of where to implement AI, what resonated the most for you with this article?

Yeah, I think it was definitely scary, to be honest, reading this article in terms of, AI and what I find. It always answers you if you ask a question, or, there's always oh, here's three things that you can do instantly and, again, there's a lot of horsepower behind it there's a lot of these answers.

It's not like it's just spitballing the first thing that comes to it, but still, [00:25:00] I just feel like, It's scary to be honest, especially with the hallucination and whatnot, because even in our case, right? So even when we were developing our solution, we use chat GPT all the time. And you use chat GPT and right away it'll give you two answers.

And then you question it. It'll go back and indicate, oh yeah, no, that was inaccurate. Sorry, let me correct myself and here's the other two answers. And another example, again, I'm just giving other examples, like what the article is suggesting in terms of the models not being accurate.

But there was recent, where a lawyer used ChatGPT to fight a case and obviously it was a huge failure because JGPT created cases that never existed. Anyway, so going back to this, I feel like I would go back a little bit to our last article where, and what you were saying, start with the solutions that are less risky.

I feel like it's not about not adopting any AI solutions. don't think you want to be on the sidelines [00:26:00] either. I think what everyone is learning is obviously we don't have the perfect solutions yet, right? Chat GPT is learning obviously constantly, but we are also learning in terms of how to make best use of AI at this point. So I would again, advise start with the less. riskier ones. Start with the more practical ones.

And then as we learn more and more sample, you can always deploy it at a smaller scale. And then once you see the practical results, once you see the reports of actually, yep, this is working as per the benchmark, and it does help our clinicians, then we, then you can actually roll it out to more users.

Yeah, you think about the fact that if you are training models, large language models, while you're also training the team, that intersection of which one is more correct, or are we more willing to accept as correct, what percentage of error that goes through governance and audit, etc. You align all those components, you know exactly where your organization can thrive [00:27:00] with the usage of these enabled tools versus certain spaces you may, and it's all going to be about culture, you say, you know what, in the ED setting, we already know that we're better off with the expertise of our residents and our physicians.

So the ED at this point, As an example, is off limits for the use of this version of AI. And you say, I am going to use it for chart prep, for documentation, for the ability to understand if there's preventative care that's being missed. What are areas that you can validate that don't introduce harm both to the patient or to the organization?

Yeah, these exactly like the physicians in ED They don't have time to what Chad GPT is spitting out. So 100 percent I agree with you

And so you mentioned earlier that there are times when you can determine if it's actually AI being used or otherwise. What would be some words of wisdom from your experience and perspective that can help people better uncover whether or not [00:28:00] AI is really the engine behind the product?

Yeah, so it, that comes down to some questions to understand what's under the hood, right? So is the solution acting more like an ML, machine learning where it's actually, using the data, it's able to identify patterns, right? And it's able to help you in some of the decision making, or is it actual AI where it's taking the next step of not just identifying the patterns, but actually making a decision in terms of, oh, I need to create such and such order based on what I'm seeing, right?

So based on this result, maybe this person needs this type of follow up. So that's where or is it, in even there, that could be a difference because you could train, only certain things and it could become more like an RPA. So if it sees certain keywords, it's going to go and create.

ABC orders, because that's all you [00:29:00] taught, it's more like a robotic process automation, not a true AI. But so that's where I would look into the underlying data set that was used to train the model, right? Ask those questions. What was the data set used for training your model? What is the actual model that's being used?

There's hundreds. If you go in Microsoft Azure and online, and if you can just look up all the different models available. So start with some of those basic questions and then maybe, depending on your contract situation with this vendor, maybe they can actually show you the actual data and the algorithms and step you through it.

Again, that all comes down to their IP and, some of those conversations might not be easy in terms of the vendor may not want to share that because that's their IP, but that's the other thing. I feel like many of the AI solutions, they are almost like black boxes, right? They don't want to tell you a whole lot about it because Everyone uses reason for, no, that's our IP, we can't share, because then the compeTIDOrs can easily [00:30:00] use that, to their advantage and whatnot.

So I think definitely not easy. But I feel like if you have the right policies in place and the right questions to ask, even before you go any further into like how and who's going to use this solution, I think that'll definitely help.

You bring up such a great point, though.

Think about this. If we're going to go into business together, and I ask you those exact questions that you just shared with me, we've already gone through loads of paperwork. And NDA being a component of that conversation is more than completely appropriate. And you can walk away from deals. There's times when people can't meet their security requirements, and you're like, I'm sorry, we can't do business.

Now you're saying, Hey, The level of transparency and trust that's created when you have that kind of reciprocity with your partners, that's what creates a partnership versus having a vendor solution. And you get that better than anybody. And I'm so grateful that you had a chance to come in and do Newsday with me and also be a partner with this week health.

Our Communities will love the fact that you have all [00:31:00] of the expertise for strategic alignment and data and resources to drive the operational excellence that they are so needing in healthcare today. And by spending time with you and your team, they'll realize that can be an absolute two way street.

Amazing. Yeah, no, thank you so much. This was a lot of fun. We should do this again and another news day in the near future. But yeah, thank you again for having me. And really look forward to maybe if you get any questions, maybe we'll have some follow up questions or comments from some of our listeners.

We often do, and the best part is you're a part of our community, and so they know where to find you as well as follow up with either information about the product or anything that we discussed today. So thank you again. We'll do more News Days. We'll see you at future events. More than anything, we appreciate you being a part of This Week Health and the 229 community.

Thank you.

Thanks for listening to News Day. That's all for now.

Thanks for listening to Newstay. There's a lot happening in our industry and while [00:32:00] Newstay covers interesting stuff, another way to stay informed is by subscribing to our daily insights email, which delivers Expertly curated health IT news straight to your inbox. Sign up at thisweekealth. com slash news.

Thanks for listening. That's all for now