Aaron Conant 0:18

Alright, let's go ahead and kick it off. Happy Thursday, everybody. My name is Aaron Conant. I'm the Co-founder and managing director at BWG Connect. We're a networking and knowledge sharing group 1000s of brands who do exactly that. We networking now and share together across multiple organizations and multiple verticals, to stay on top of the newest trends, strategies, pain points, whatever it is that shaping the industry as a whole. We connect with I connect with 30 Plus organizations every week, just to do exactly that kind of pick their brains and what's going on. And when the same topics come up over and over again, we host an event like this. So a couple of housekeeping items as we get started here. The first is, you haven't we want this to be as educational and informational as possible. So if you have any questions along the way, any comments whatsoever, drop them in the chat, you can drop into the question section there, or you can always just email them to me, Aaron, aaron@bwgconnect.com. And we can get those questions answered in real time. So don't hesitate to submit any as they come up. The other thing is, is we're starting this at three to four minutes after the hour, just you know, we're gonna wrap this up with three to four minutes to go in the hour as well, we're going to give you plenty of time to get on to your next meeting without being late. And so that I want to kind of kick off this conversation as a whole. A lot of different industries, sectors, organizations across the globe being affected by the pandemic as a whole is drastic changes that have been shaping it. And so a lot in the patient health space as a whole and patient engagement. You know, it's really interesting, you know, how it's, you know, a lot of times we think of, you know, the digital effective eCommerce as a whole and sales and Amazon and Walmart and Target. But in reality, there's multiple other industries that have been drastically affected. And so we got some great friends, great partners, supporters, the network over a Perficient as well as IBM, they agreed to jump on the phone today, and kind of walked through a study that we did together across the series of over 100 different organizations and what's going on in this space. And so, as we kind of kick off the conversation, you know, Eric, I'll kick it over to you first, if you want to do, you know, intro on yourself organization, we can jump over to Tom, and then we can jump over to Brandon, who's on as well. And after that, we'll kind of jump into some of the content sounded.

Eric Walk 2:30

Great. Thanks, Aaron. So as Aaron said, I'm Eric Walk, I'm a Director of Perficient in our data solutions practice, you know, Perficient is a global digital consultancy, we work with clients sort of across the whole stack on their digital transformation, modernization and data problems. You know, my area of expertise is really around data content and knowledge, focusing on document processing, AI solutions, natural language solutions, with industry focuses in healthcare and financial services. Really excited to be here today and talking about this and I'll pass it over to Tom.

Tom Lennon 3:10

I want my name is Tom Lennon. I am one of the leaders of Perficient's health care team. At Perficient, health care is little over a third of our business. I've been with Perficient for gosh, around 16 years, and 14 of those I believe is really helping to understand and define and implement healthcare related solutions in the IoT space for our customers. And when we say healthcare, we mean predominantly payers and providers. As part of that, that a lot of work with IBM during that timeframe, as well as many other vendors. Brendan Fowkes is with us today. So Brendan, do you want introduce yourself?

Brendan Fowkes  3:55

We'll try a video one more time. Oh, it looks like it's stabilized. So fingers crossed. Good afternoon, everyone, Brandan Fowkes, and I'm our Health Care Chief Technology Officer inside of IBM software. 25 years in healthcare it this is the only industry I've ever worked in both provider health plan and sort of government social services. So it allows me really a bit of a unique view, and how we try to approach problems, software architect by background. And one of the ways we're all successful is by leveraging a Partner Network with a great partner like Perficient, Tom and I have known each other I think too many years now to count on working on delivering successes outcomes for our provider clients. So we're really looking forward to this discussion today. If I give a nursing point of view, it's because I was raised by one. So I was taught to think like one so if you hear things that sound like your nurses, it's because that's how I ended up in this industry in the first place.

Aaron Conant 5:00

I was just thinking that I was raised by a nurse as well. So tons. And it would be a great conversation. So just to get a reminder here, if you have any questions or comments along the way, don't hesitate to drop them the questions the chat section there, or just email them in Aaron aaron@bwgconnect.com. So I think we're first going to kind of roll through a little bit around this study that we've done, and kind of give you a baseline for the conversation as a whole. And so certain survey, just over 100 different organizations, really, you know, the meat of it, there was over the over 50%, mid size or regional providers as a whole. Here's the whole breakdown, I'm not going to read through every single point that's on here. But 72% of respondents represented large enterprises, 69% organizations generate more than 100 million annually. So over index a little bit to the larger organizations as a whole. And what's really interesting, a couple of things that we found out as we're doing this is when we think about, you know, how they're prioritizing budget as a whole towards patient acquisition, huge focus, you know, shift in the focus as a whole. But what didn't change was, you know, if you look down here, that meaningful uptick in budgets for acquiring high value patients heading into the next year. So there is this idea that, hey, we all do want to over index on patient acquisition, but there's not really a change to this, you know, specifically around high value. Organizational maturity, I don't have to read, you know, everything on these slides. I don't necessarily want to but you know, people are looking at how mature is your organization as a whole, you know, organizations report to be most mature the CMS CRM EMR integrations, I think that's that's not surprising as a whole. It's kind of the first part of that people kind of buy off. But, you know, as we as we look in these other stacks over here, kind of not a big difference as a whole. I don't know. I mean, I can kick it over to you here at Eric, Tom, Brendan. I don't know if you guys have any thoughts here. But we can pause here briefly, if anybody wants to jump in. I think this was these ones,

Brendan Fowkes  7:04

the best part of the conversation, you have to count. But these are ones that we needed that baseline of understanding, okay, they tackled Meaningful Use in some basic integrations and started to think about potentially a CRM that goes beyond your core EMR. That's everyone's thinking. But as you just was saying, Aaron, is everyone liked to go get it, but just not sure how to go to what to go do next? Yeah.

Tom Lennon 7:30

And you know, what things add to that is, you know, it's interesting in the survey that so many people are focused on that CMS, CRM EMR integration, because what that really gives us the opportunity to integrate that data from, from the clinical events which occur and integrate that into the CRM, and then use that to help message directly to patients to make sure that they're compliant, make sure that they're understanding, you know, when their their follow up visit is make sure that they're they're doing those things they need to do to help keep healthy, but also to reduce readouts and things like that, to really look at how do we change the behavior by actively, you know, going out and driving information to those patients?

Eric Walk  8:18

And how do we start that with sort of a baseline of rules, right, that are sort of fixed and hard coded, and add intelligence on top of that, to start making predictions, right? We're making predictions about what's the best way to get a response from someone? What's the best channel to talk to them over? You know, which which services are most likely to come actually come and get? Right and come receipt? So it gets into some really interesting questions that I think are only, you know, we're only seeing the tip here. And it's going to be the beginning of a pretty big transformation looking forward for the industry.

Aaron Conant 8:53

Yeah, I think across the board, they're gonna kind of look at the data as a whole. We're at, like Digital 2.0. Now, right, we have the baseline, we have the collection of the information, we have the basics, you know, kind of, you know, starting to uptake what we're doing. That's not just those are just table stakes now, right? We're at a point now, where you need to leverage everything you're getting in all that information that you're you're collecting to get over here to a personalized experience as a whole thing. super interesting. We jump in here, CRM usage as a whole. So on average response, were two, you'd report it to us 1.3, you know, on average, CRMs 81%, though, said they'd like to go to a single one. Right. And so I think picking that apart. Everybody would agree, right, is that that's an ideal state of single but how do people get there? Is it a reality is the other side of the equation? I don't know if anybody has any quick comments on this one, and we can keep rolling through here. I know there's a good shot of that conversation but this idea of jumping to a single one

Eric Walk  10:01

I think there's an opportunity to save some money there, right? And to optimize your operations there, but it is, it is a question of how realistic it is and how much you're going to spend to get there. And so I think it's, it's definitely can feel a little pie in the sky and feel like the ROI is too far away. But there's some interesting opportunities, as you start to think about that, and, and, you know, replacing some some sort of blackbox technology that that comes with some of the, you know, some of the less less fully functional, less fully capable, but industry specific platforms, and looking at the the more enterprise, you know, full full spectrum platforms that may not have those sort of baked in features that you need. So it's definitely gonna be an interesting journey as folks go down that way.

Brendan Fowkes  10:48

And there's definitely value there. But it's hard to quantify, as Eric was just saying, the value of going to one electronic medical one integrated platform really made sense. That was kind of an easy. There's nothing easy about putting in an EMR. I used to write them, I get it. But, but it's an easy vision to articulate, right, between engagement marketing with a little was a little more abstract, the ROI statement, is the cost benefit Delta really going to be there. So the 81% that see value in it, you know, our advice, then is understand articulating what your savings by retiring legacy applications like some of these more blackbox, pre configured things that lack flexibility, as Eric was describing, what are you paying a year in that one might go so that might be the easy way to get to an ROI. Because if you have one CRM, when your patients call or you try to engage them, it's from the people on their own, the phone only have one place to look. So that could be a huge savings and a great experience for not only your call center employees, but the your patients when they call in.

Aaron Conant 12:05

100% Arvind here to write just the chat here, in a single view of the customer.

Brendan Fowkes  12:11

I guess that's exactly what it is. You know, it's but you know, there's only so much you would love to have a single view of the customer. But if you can't get there, this is ways around it.

Tom Lennon 12:22

Yeah, and we did see a lot more, a lot more emphasis, some years ago, maybe five to seven years ago, we're really using multiple systems almost strategically to because of the features and functions which are available, robust best of breed. But we've seen that going away. Now I'm really the maturity of the systems out there are able to handle more and do things in that single that single product.

Aaron Conant 12:50

Awesome. So jumping down this side, CRM usage as a whole, the dominant use cases for CMM, you CRM has continued to revolve around basic communications and tracking, you know, we're seeing this, you know, a meaningful rise in rates or cases, I think that's kind of like this move to digital 2.0. Like, we have to look at this differently. We have this in place, I mean, kind of like who you are, you're saying that there's all these different little pieces as a whole that we kind of used is because they were resident experts. But now there's there comes more holistic solutions, we have the opportunity without all these different integrations, right to actually make everything we're doing more powerful. So let's jump on to the next one as a whole as we keep going through here. This is kind of like the summary slide as a whole patient advocate. So from the first couple slides, patient acquisition, personalization, engagement are becoming increasingly critical initiatives within healthcare organizations. 55%. I want to jump down to this, you know, this one is 55% planet already utilized AI or machine learning capabilities for patient acquisition and personalization with different degrees of maturity. And do you have any plans to use or you're already using AI or machine learning capabilities for patient acquisition and personalization? I mean, this is a huge jump from my guess, you know, probably a year and a half ago, where it was closer to probably 25, 75. Is that something you guys are seeing across the board as well in the AI and machine learning space? And then there's a lot of questions that come out of that when you remove humans from the interaction.

Eric Walk 14:31

There's definitely a lot of variation in the answer to this kind of question industry by industry. The the more heavily regulated the industry, the more concerns there are about privacy about sensitive data, the more likely it is to be lagging. Right in, you know, the retail industry, right? This is let's charge forward as fast as we can. Right in healthcare. It's certainly behind, you know, lagging behind because of some of the concerns that you've listed here. around data protection around privacy, and you know, most importantly, around consent, right? You know, when we start to use PHSI, in these kinds of modeling activities, we're talking about using it to market to a patient, right? If the patient hasn't consented to that we've got problems. And so there's a number of, of issues that we sort of have to deal with. And it is a challenge. But they're all deal with a bowl as it were. Right? If you if you put the, you know, put your mind to it, and work through it as an organizational change problem.

Aaron Conant 15:36

I mean, if we jump into that, and so this is kind of as we get into the open discussion here as a whole, so if anybody has any questions drop in the chat or the questions section there, we can get them answer comments, and we can address them. But you know, specifically around that a lot of people have questions around, you know, the PHSI. Right? How is it protected? You know, when you're using AI, when you're using, you're training it right, and it's machine learning? What are you protecting it there? How do you see it? Is there any not in the CRM? Is it, you know, is available? How do you protect it? There's, it's, it's an interesting space.

Brendan Fowkes  16:15

But yeah, I think. And the key thing, the way we always hear about it, and we sort of more used the term trustworthy AI. Because if there's any violation of a patient's trust, you've lost everybody, whether you lose the nurses who were supposed to use it, the call center agents, the patient's themselves. If we start with that foundation, and the core belief of everything has to be trustworthy, from the start, there's no hidden. Whatever else, you're not borrowing the data to go do something else. You start with that foundation at your core belief, you can start to overcome any of these objections to Eric's comment about deal with people like that one. Because you're going to design a system with the proper security, privacy, consent, and almost borrowing a little bit of what we've done on the academic side. I mean, you really think about the IRB protocols and things along those lines to do a research study, you can borrow lessons learned from life sciences, from academia, as you're rolling out something on the production side. So that's a great thing about you know, having a diversity across the healthcare life sciences ecosystem, you can borrow other examples to make sure that everything's secure when you design a trustworthy story. Yeah. And I think the other thing, too, is, what AI ml? Yeah, is how we're just going to build an algorithm to see who we should reach out to, are we putting conversational AI techniques into an interactive experience? All of them are good options. But that's another thing is what technology? Where do you want to start? Because there's different ways to attack this?

Eric Walk 17:58

Well, and and how are you going to control the deployment to control the data? Right? You know, you can, there are CRM options out there that that have the proper certification security controls in place to allow you to load your pH I in there and, and security, control it appropriately. But even so that may not be an option for any given organization? And how can we take the AI and the machine learning models and bring them to where the data is? Right? So we, we have fewer concerns about shipping the data to other places for training and other places for for executing those models. And we can, you know, more tightly control how that data is being sent, where it's being sent in and how it's being used?

Tom Lennon 18:42

Yeah, that governance is important, especially as you put it, you know, where we're going to send that data, making sure that we're able to limit the exposure in one place and send the information, it doesn't have that exposure all over the place, because trying to manage that when it gets throughout your entire organization can be quite tricky, right?

Eric Walk 19:02

Yeah, yeah. And having tools that you can deploy, where you can get them is really important, right? Knowing you know, it, especially as you're looking at Cloud strategies and those sorts of things, right, you can, you can choose solutions that are cloud native, you're going to have more challenges in terms of just making sure you understand the way in which each of those solutions controls pH i, and secures pH i, because a good number of them are compliant in various ways. Or you can look to other options that allow you to maintain trust through the whole chain of custody of a data by using, you know, possibly more complex and more nuanced solutions, but that allow you to take your your modeling and do it where you need to do it when

Aaron Conant 19:50

there's that. So, Roger writes in, you know, 55%, you know, seems high unless we're talking about automation versus AI, which is the kind of a little bit you're getting to there's Hey, how are we Meyer, we actually just using a we actually using AI to engage in a conversation which gets to the patient engagement side of things. But I would like flip that a little bit and said, I think 55 seems low at the point that we're at right now that 45% would say they don't have plans to or don't already to not be looking at it. Like I should say, 45% seems that's high, it seems like 55 would be 85, or 90, would be considering it.

Eric Walk 20:33

At least if this were any industry other than health care that we're looking at, I would agree. But it doesn't shock me for healthcare.

Tom Lennon 20:41

Because that security mindset also is going I think, keep that the yellow side the 45%. Where, you know, it's not actively being planned to use a higher, I actually think in the in the 55%, that's still more people from from our experience with customers is probably more along the lines of people planning and beginning to use it as opposed to, you know, actively using true AI in it, unless you start to count where there's AI built into some of the tools that their purchasing, which could have some of that included. But I think that's what we've seen more of so far as less people actively using AI and building their own models more using, you know, what's built into the vendor tool or planning to getting their infrastructure in place to be able to start leveraging AI in a safe and manageable way.

Brendan Fowkes  21:34

Yeah, and I think too, it's a great point, was it automation or AI from the person from the audience? I think if we ask them, would they like to automate some of this? I think the answer would be 100%. Now, where there's AI, there's automation with this automation, that can be AI. That is a very symbiotic story. I wonder if the number the no plans number, which is I agree on, it's probably still pretend too high at 45. Without having plans, we have to remember that of our respondents, they have been through a perfect storm tsunami, just they're probably pretty drained at the moment, you know, from the past 18 months in the recent uptick, you know, our provider community, we have to be very thoughtful about approaching them about helping them with things because they're just under this wave. I live in Massachusetts, I'm actually here. And I'm actually in an office today in New York City of 590. Madison. But in the state of Massachusetts, we just had a bill go through that the hospitals agreed to to push back elective surgeries due to the rise in cases. So there may be a COVID Hangover to this 45% number being so high as people don't want to take on anything new, until they know what's actually going to be what's going to happen. I don't know if that's the case. But that could be one of the reasons why people still don't have plans around innovation on patient engagement. Because there's still just their teams are still just drowning and trying to survive. You know, as we get hit again, it's

Aaron Conant 23:15

a great guy comes in a number of telco or his healthcare organizations are using AI for RPA applications to you know, this kind of a different purpose. I don't know if there's any comments there. But they have another question that comes in as well, they'd like to jump out too. But any thoughts on you know, using AI for RP applications?

Eric Walk  23:37

I mean, that those two sorts of things are synonymous, right? You don't really have RPA. Without AI RPA is a it's a cool set of technologies that could be really powerful in a lot of cases and, and give you sort of quick time to value on on certain kinds of quick wins. But you know, I think it's uh, you know, it's, it's an interesting, interesting space and an interesting way to start automating right? You can get going quick and and there's always though, the next stage right where you can, you can get better results, if you if you put in a bigger investment, look at some some different kinds of technologies and automated, a more thorough, more thorough way.

Aaron Conant 24:20

About a patient engagement as a whole, right is the initial outreach side. So you know, questions around how you can use pH AI to make predictions for marketing purposes. You know, we still have the concern about storing that in a CRM, right, or whatever marketing automation platform you might have. Any thoughts there about how we can utilize it to do outreach? Everyone, these are the top ones, you know, I mean, the first slide we did the shifting, it's a major focus going going forward. It was probably a top indicator there I'd love to hear your thoughts there.

Brendan Fowkes 25:03

So I think it's a little bit of back to the trustworthy AI and having problems that are over comfortable or whatever we really call it, Derek, but doable. But the thought is, your where you go to build these techniques, I think is where a separate platform somewhere fits in. Because you can put all the security and privacy in your data science environment, data can be obfuscated, masked, still with relevancy enough so that the models trained properly, you know, you learn how to design that program. And then that scoring model, you just push the scoring model to the CRM, you don't try to do that inside the CRM, because that's not what it's meant to do. It's not a data science platform. It's not an AI platform, machine learning platform, they are very, very good at what they do. This is where augmenting them with skillsets people like Eric and his team, and the right piece of technology then augments the work you do in the CRM. But to that automation piece, I think that's so key is, what can we automate, to allow your high value employees free up to do high value things? And yeah, we do it in operations we're talking about specifically this survey two is about engagement. And I don't think we've really touched on where we could free people up to provide more bandwidth for better engagement.

Eric Walk  26:29

Yeah, I mean, you know, an example that I was thinking about is me, because one of the comments in the in the study brought up the digital divide, right, and the the idea that there's there's high value, patients that are easy to reach with an email will come consume high value services. But there's also a whole population of patients that we need to reach, who may not be your highest value service consumers, but but still need health care. Right, and still need to be contacted and communicated with appropriately. And, you know, one of the things that struck me is how do we, you know, instead of just predicting who's going to consume high value services, can we start to predict how to best communicate with someone what kind of communication and respond well to, right, there's a lot of ways that we can use this technology to make predictions. And, you know, it really is data science, right? Where we are really doing experiments, right to try to figure out, given the data we have, what kind of predictions can we make? Can we do that in an ethical way? You know, Brendan had mentioned earlier, the idea that we can learn something from the IRB process, right, as we start to go through this, in order to maintain a an ethical stance, as we go forward. And and, you know, build this in a way that's trustworthy, and makes predictions that are, you know, you think about saving time, right? If we can predict that a particular person isn't going to respond to a phone call, let's take them off the list for phone calls, right? If we can predict that someone's you know, the best way to reach out to them is, you know, knocking on their front door, then let's go do that, right, let's use our resources where they're going to be most valuable, and where we're going to see the best outcomes.

Brendan Fowkes 28:07

And it’s interesting, you just mentioned that digital divide Eric, that was the one I knew we were going to get to, but since you set it up, is we were talking about engagement strategies and technology. And it's amazing that the respondents very quickly jumped into an equity discussion, a health equity, equitable distribution of health care where the social determinants fit in. And that's something we're really seeing what did we see over the past 18 months, our most vulnerable took the worst beating? And we where does that become it was interesting this morning, I was I was reading something this morning is on that topic. It fits right in is Dr. Stephen Klasko, who is the soon to be retiring CEO of Jefferson Health, published something on the world Romic forum this morning, around broadband access, now fitting into the category of social determinants. Because we want to engage with virtual care, because it works. We know it worked, we saw that huge three 5,000% increase in telehealth start to dial back down again. It's not good if you don't have access, and it really came in. I wasn't surprised that came up. But based on the questions, we asked how quickly people went there. Yeah. It's okay to talk about Tom.

Tom Lennon 29:29

Okay, and without using social determinants to write, especially as we start to push it into AI and leverage that in AI. You know, just because you're using that information doesn't mean that you're going to drive equity because the the bias that is inherent within AI until you can manage it and normalize it, that bias is still gonna be there. So really learn how to use that adapted is important because you know that that has a opportunity for us to really help address some of that inequality but it also has the opportunity that it could make it worse because you know, AI doesn't have the same feeling and natural intelligence, right as we would have, when we're worried about different or groups within our population.

Eric Walk 30:12

And one, that's where it's really important to, you know, to think about the tools you're picking, right? Some AI tools have capabilities that help you identify biasing your models. Others don't, you know, the, those priorities are critical, as you're sort of looking at the look at the tools you're using, looking at the way you're designing your models and thinking about building models. But also educating your data scientists, right, as your you know, as you're training your people, is people build the models, people build the algorithms, people train the algorithms. You know, it's also critical to train those people to think about these things and be concerned about them and and look out for the warning signs that there's there's an issue in a model that's, you know, exacerbating historical, historical inequality.

Aaron Conant 31:04

Where do you? Or do you think that this percentage shifts over the next year to two years? Do you think it stays here? Just, you know, getting back to some of the comments around people are inundated or do you think this, this changes a lot?

Brendan Fowkes 31:21

So, I think it's going to continue to shift upwards. And part of the reason why is if you ask the health plan community, they're 100%, focused on doing this. So it makes sense, because the providers are still the trusted partner providers and pharmacists are still a patient's most trusted advisor. So I think that trust factor that they have, gives them an edge and engaging patients now maybe they should collaborate, you know, there's a skills gap, maybe they can partner with some of their health plan, you know, when they get into value based arrangements, they could do some more collaborative effort on this tip to leverage each other shared trust as a thought, to improve engagement. Because we know the health plan community is trying to do this. Every day. It's actually why I'm here in New York, we were setting together strategy for this exact problem. But from a health plans point of view. Today, there's a bunch of sitting in a room whiteboarding, the whole architecture out, but that's the thought process, they're focused on it. So providers should be because they still, when it comes down to it, you know, do you want to hear from NYU, or ambulance? I'm in New York City, some new news in New York examples as a patient. You know, it's just it's, it's, it's a real getting a lot of help into moving to try to really adjust that trust factor, but that's a topic for a different day.

Eric Walk  32:57

Yeah. You know, I agree, I see it going up, you know, it's certainly not going to go down and I don't expect it to be be static. You know, aside from the health plan, factor, right. As as things level out, right, you know, every indication is that, you know, pandemic is becoming endemic. This is a forever problem, not a problem that's going to end one day, very suddenly, it's going to whimper.

Brendan Fowkes  33:27

There goes my Christmas thanks.

Eric Walk  33:32

But you know, it, it will stabilize right. And as the situation stabilizes, the question that folks should be asking is, how do we how do we be ready for something like this to happen again? How do we we be ready? How can we be ready for for, you know, future challenges, where we, we need to be, you know, technologically prepared to do new and different things, using these kinds of tools and technologies to engage with patients to understand patients better, right, because, you know, it's sort of two sides, right? as we as we provide a patient with a more personalized experience, and more personalized outreach when we're trying to get them to come in and consume services. That exercise in itself helps us learn about our patients, helps us learn more about where they are, how they consume services, what kind of things they you know, drive them what kind of access they have to things like broadband, like the ability to get to the hospital to get to the care provider. So how can we take advantage of learning these new things about our patients into the future, as we you know, build our preparedness plans for future crises, but also for our day to day operations.

Tom Lennon 34:50

Yeah, and it alsocomes down to resources, right, and when we have we have different size and shape organizations have different resources. And I think it's probably universal think that well, where we can use them The tools to help lighten the load of a lot of the folks within that that provider organization. Great. But it takes it takes resources to get that ball rolling to gain that momentum. And so when will that happen? Some organizations have seen the opportunity to take advantage that already more on the largest, the largest size providers, the smaller providers typically have a little more difficulty making a making, I guess getting some wind at their back and that situation?

Eric Walk  35:29

Yeah, you know, things we think about in terms of lightning load, right? We're really, you know, sometimes it's, it's less about lightning load and more about transitioning the worker to doing higher value things. Right? You know, going back to a call center situation, right? Can we use this kind of technology to divert calls away from the call center, right, and handle them in an automated way? Sure. But we're never going to get to 100%. Right? There's always going to be some need for a human to pick up the phone and talk to someone. But can we use some of this predictive modeling to make predictions about why that person is calling? What information might be relevant to that call center worker, so that we can make the experience for the call center worker more intelligent, more intuitive, right, so that they can handle that call faster, more pleasantly less, putting someone on hold while they go look something up? Or call a friend, a friend? How can we automate that process more, so even when there's still a human in the loop doing the service, right, answering the phone talking to the patient, that that's more seamless, more smooth, easier for that, that person answering the phone to be more caring, more, more trustworthy, and build a better relationship with the patients.

Brendan Fowkes  36:42

And that's a that's a really good point. Because Tom, you mentioned it for us as resources. The unemployment numbers in this country are at historic lows. So traveling nervous is through the roof. There's a skills warfare out there. And when you start thinking about innovative technologies, sometimes those people are harder to come by. And that's where we always like to compliment strengthening, find a win that your back. But that's a very good point Eric makes two is let's say, We engagement can be about humans, humans, so it's okay. So in certain cases, it's preferred. But if you've done something up front, to more educate your patient, before they call via conversational AI, they get two or three questions answered. And they call upon the third. Or like Eric was just saying, you make the agents experience that much better. So they can be more efficient. And make the members you know, get the questions answered. Especially now because call centers had to go remote. So they phoning a friend isn't you can't lean over to the next cubicle and get an answer, because there's nobody there. But your kids down the hall, you know, so it's those people? No, no. They're the ones using up the broadband in the house playing video games. But yeah, if you can better use these tools and techniques internally, to more empower those people. Well, that's going to give the patient a better experience when they call and email. That's just something simple as scheduling. What's the hundreds, like this magic scheduling thing that happens inside healthcare, so don't tell anybody I'll work with trying to make that easy for the schedulers.

Aaron Conant 38:31

Just a quick reminder, I have some questions coming in. But if you have questions drop in the question section in the chat, or the q&a, but I think it's a great, great point, that, you know, so often we think of AI as just you know, spewing out information and targeting and using it for marketing, which is good. But patient, you know, engagement, you know, the ability to use that internally as well, where still the communication happens person to person. But acquiring all that information, I'm trying to be facilitated that gems are just a question around, like, is this an ROI question then right now that people are struggling with it? What is the return on this right time, effort, money, right, and all are stretched thin right now? How do I pull the trigger on this? You know, one more thing, and what is the ROI look like is a whole area I was just thinking about that way. But you're talking about, you know, moving people to higher value. You know, Tas how changeable? Is it? The ROI in the end?

Eric Walk  39:29

I mean, it depends on what you decide to implement, right. But you know, let's say, let's say we implement a conversational AI system, right, that can handle 80% of scheduling, random appointment scheduling. Well, you know, now we take that from entirely a situation where people are talking to a human on the phone, who's looking at a calendar and trying to figure it out, right to a situation where 80% of those calls are diverted. The 20% that get through have conflict Any problems that need a human to look at them, right? And then that human, maybe you don't reduce that person's cost, but they can spend more time working on each each issue, right, maybe you can give them the ability to do a better job figuring out those complicated scheduling problems, right. So that the patient experience is improved. And now the patients want to come, you know, have have a higher higher perception of your organization of your hospital, right and are more willing to come back to you refer you to friends. But at the same time, you can also, you know, look at it as a way to reduce some headcount in different places, reducing technology load in other places, but it really depends on what you want to implement, right and how you attack it first.

Tom Lennon 40:46

And that immediately measurable ROI is also should, I guess, open the aperture a little bit to and think about, the better we can engage patients, the quicker we can do that the way we can change your behavior can have such a dramatic impact on the overall margin or bottom line of the provider organization, when we think about high position, health and involvement in ACOs, in different, you know, payment models, where they're actually going to change the way that care is delivered and reduce the overall cost. It might not be so apparent at the system or project or program level.

Eric Walk  41:25

Yeah, that’s a good point about the with the accountable care organizations, and really using some of this to reduce the per member per month overall rate, right to try to achieve some savings there.

Brendan Fowkes  41:36

But that's, that's a good point though Aaron that sometimes. Why is that number 45% still so big, is that ability to define the ROI? And, you know, I always love that. So what does success look like? When you have these conversations, having a definable metric. And if we're doing call centers, Eric was saying, you know, discussing that, well, those that these metrics did that are pretty easily and widely accepted cost per call number of calls, deflections, average handle time, you can put some hard ROI around that. And people do sort of cost sharing agreements to try to pull these things off, but all other ones are sometimes a little harder. But our recommendation our experience is, let's agree to agree what success looks like before you define the use case you want to chase. Anecdotally, there's such a wealth of information in all these providers and what we can be doing better. And we don't pretend we've all been in this industry a long time, but we don't pretend to know how to run your business but our recommendation is defined what the outcome is you want to measure because you can't fix what you can't measure. So we need something we need to measure to prove it worked. And then that success will build on itself. People will do case studies you'll go out on the speaking tour circuit you know, you'll be on one of these webinars and that blue number will start to come up as they hear about their peers doing other things and that it can be done in smaller chunks.

Eric Walk 43:12

I mean it is a really good point that you can do this use case by use case right you can tackle one one challenge in one area with this kind of technology and and you know slowly expanded out right and and you you know the overhead the upfront cost of dealing with security concerns dealing with the ethical concerns dealing with all of that is it's lower than you might expect if if your use cases sufficiently simple right and and once you get a simple use case going right dealing with the additional concerns of more complicated use cases becomes incremental incremental costs not the the overhead bear is not going to be huge and it's not gonna not gonna sink your your plan to expand out into into other areas.

Aaron Conant 44:04

If are you walkingpeople through that then because I think a lot is just you know, having you know run the surveys a whole lot of people you know, think it's this massive undertaking, right that you're you're ripping everything out you're putting something new and or it's such a huge a just undertaking as a whole but I don't have the time effort money to go after this right now. So you kind of walking them through and saying, hey, yeah, there's 57 different ways you can use it. But stem you know, the first three are really easy and it's worth giving a try. Are you breaking it down for people that way?

Brendan Fowkes  44:44

Yeah, I think I, I call it crawl, walk, run. People use it land and expand, start small. whatever metaphor you want to use. There's a way to take something simple. An example for one organization was like, Can we predict For a group of people that might be be non adherent their medication plan for something simple, it was actually, it was actually middle aged men not taking blood pressure medication not taking hypertensives probably a good group to go after, right? Because other things will happen? Well, if you looked at socio economic factors in that model, and we predicted non adherence, you would just ship them a three month supply and give them a call and tell them it's coming. Like, it's, it's a simple, it's not a complex prediction, it didn't take a lot of work. And the outcome isn't, I mean, what's a generic cost for three months in our grid, Scotia $30, to keep somebody in here. So there still is true machine learning in the model, we did all scatterplot with a random forest, look at all these different techniques and pick the right model that was most accurate. But it wasn't a really complex use case. It was just a different way of using the data they have. And then helping them automate something to drive a benefit at the same time. Yeah, so it is all about crawl, walk run.

Eric Walk 46:09

Well, and you know, what action, right? Because, you know, sometimes that creative thinking about, Okay, what if we just shipped them their medication? Right, and didn't charge? How, you know, coming up with those, those little things you can do is that is often the hardest part, right? It's finding that use case.

Aaron Conant 46:30

There other things that came up, you know, as we're getting here to the last, you know, you know, five or six minutes or so, you know, Eric, I know you'd brought something up out of the study, obviously, we couldn't tackle everything on this, this summary that we had, were there other things that stood out to you that you thought, hey, you know, this was really highlighted for me, we kind of do a round table here, like, Hey, these are key things that people should be thinking about things that stood out for you in the study or things that people should be planning for for next year. We've had around Eric, I'll start with you. And there we go, you know, Tom, and then Brendan.

Eric Walk  47:03

Yeah, you know, I mentioned the, you know, I think we all sort of were really intrigued by how quickly the issues of digital divide and Health Equity came up. But I was also impressed by the strength of the signal in certain areas. You know, I think, more people are more interested in in doing these kinds of things than even we predicted, you know, we wouldn't have done the study if we didn't think there was interest. But I think the interest is stronger than than we were expecting. And I think that's really, really compelling. But, you know, I think the something I said earlier, right, you know, active explicit and informed consent, when we work with our patients, and to use their data for these things is really important. And I think, I think that's sort of the the starting point of that trust today i and, and sort of where we're going to be going here into the future. And it helps address a lot of the concerns that we saw for the study around around compliance, right, where, you know, ultimately, when when the patients have given their consent, we have a lot of flexibility to do things. But we do have to make sure we're dealing with that appropriately.

Aaron Conant 48:14

Awesome. Tom, anything that stood out to you?.

Tom Lennon 48:20

Sure and, you know, so we talked a lot about trustworthiness of the data, right. But one of the things that really caught my eye was multiple mentions in the survey about situations where people don't are sure that they're getting the right patient, have they identified the right patient? Are they somehow mixing him up? You might have consent, right? I was I was a great point Eric made, but then if you send the wrong information to the wrong person, that's a problem. So what are the ways when we what are we going to do about that and make sure we can increase the confidence there. And there are and part of it was answered by other people's comments. And, you know, so using third party data, for instance, and we've seen a lot of that, you know, that the tooling, you know, whether it be tools that come with your CRM to do some matching or foreign, you know, MDM ENPI tooling or whatnot, and the the advances there and use of AI within those tools, but um, but one thing is well is using that third party data where you can do referential matching, starting to use bio, bio identity matching things like fingerprints, and, you know, recognition and eventually even, you know, gesture on a device, things like that and start to incorporate that to really change the game and how they can elevate you know, that matching percentage, and we have the tools now that are some of them have been around for a long time and have seen advancement, and others are relatively new or new ways of using those tools. But that is a critical measure, but it's also something that's attainable and and worthwhile.

Aaron Conant 49:57

It's super interesting. Just maybe think of like digital identity As a whole, and in other areas, your digital identity gets morphed into an avatar and into the metaverse. And where all of this could be going in reality, right is, you know, in the portal where digital identity is all locked up in the blockchain.

Brendan Fowkes 50:22

We just went to a totally different webinar, we're going to talk about network networks, and immutable records and things along those lines. But I think the thing that that stood out to me, obviously, we talked about that social divide. It was interesting, we did a study on that one, with the blues network, we did with 39 of the 50 states represented. And this idea of engagement around equitable distribution of health was top of mind, every C suite, every single one, and they were trying to use techniques to it. But what's interesting, though, too, is what the other thing that stood out was, they didn't use the word like bias, but they use words like well, I have multilingual, bilingual 40%, Spanish, what do I need to do that's non traditional to work across these, these new mediums. And people are thinking about innovative ways to do I think that was exciting. Everyone's people willing to engage across these large enterprises. Because this is a topic that matters. It is creating a pivot I, whatever the number is on the circle chart. But they're thinking about the nuances of trying to do this. And Tom mentioned it earlier, they didn't use the word bias. But we have to make sure that there's not bias in the data. As I'm sitting here and from Boston, but I love New York City, especially from a healthcare IT is a micro nowhere else on the planet. Do you have a microcosm of all of the world's populations in such a small area? And how do you engage this massive population and then treat them and care for them differently. That's the fun part. But that's why there's so many academic medical centers, because you can study anything in New York. But it's one of those things that you have to think about that the population you are trying to target to make sure there's not bias in the data, if we're trying to do some of these innovative techniques, but that comes back to trustworthy all those other things. People that have helped you do this, we can advise you, other people. This is these are important things. And the fact that they're excited about it, hence the participation people here today. Everyone's thinking about it. So let's go see if we can try to solve some of these.

Aaron Conant 52:30

Yeah, awesome. Well, thanks. As we get to the end here, awesome conversation. Eric. Tom, Brendan, thanks so much for jumping on. You know, this study was incredibly insightful. If anybody wants a follow up conversation with anybody, you know, on the line here. 100% worth your time. I think it's worth you know, doing that deep dive on a one step 123. We're just getting into it. You know, you don't have to, you know, eat the elephant in one bite, I guess you can eat it one bite at a time. And these are the, you know, these are the people out there that are leaving the space that come highly recommended from throughout the network. So 100% worth putting some time on the calendar. With that. I think we're gonna wrap it up. Thanks to everybody who said a great conversation. Great talking points. Great questions. Thanks again, Eric. Tom, Brendan. We're gonna wrap it up here. Oh, everybody has a fantastic Thursday. Everybody. Take care. Stay safe in the board. Have you had a future event? Thanks, guys. Thanks.