Building Unified Health Profiles - World Tour Chicago | Salesforce

Good afternoon, everyone. I am Alyssa Halcomb, the provider marketing lead here at Salesforce and welcome to our session building unified healthcare profiles. We are in a new era. There are rapid advancements in technology and a growing emphasis on personalized healthcare. We are going to show you today how data cloud is revolutionizing the way you interact and manage healthcare information. We're going to give you a firsthand look at how data cloud allows you to connect and centralize your data, enabling holistic health profiles that empower patients and you, the payers and the providers reframe how you think about healthcare engagement. And as we always do here at Salesforce, we want to start by saying thank you. Thank you for coming to hang out with us in Chicago today and thank you so much for bringing your skills, your knowledge, your expertise to the table and sharing it with others. You are the reason that you that we are making such a difference in the healthcare industry and we are so happy that we are able to partner with you. Thanks for joining us. Today. There is more data than ever before. With a world flooded with Data, guess what? 30% of that data comes directly from the health industry. And if you have all of this data and you're not doing anything with it, it's a really big missed opportunity, Right? You should be organizing the data and using it in a way that allows you to deepen your relationships with your patients and your. At the end of the day, your patients and members are customers, right? And what do customers want? They want you to know them, and I mean really know them. They want you to know their preferences, their identity, their communication preferences. They want you to know all of that about them. At the end of the day, they're getting this level of personalization in every other interaction. So why should their healthcare be any different? I am so excited to be joined today by David Samwell, one of our esteemed solution engineers here at Salesforce. He's going to show you how this works. We're not just going to talk about it, we're going to show you exactly how it works. But before we get into the demo, I want to tell you why this technology is so important, not only for you, but also for patients. Chronic diseases are becoming more and more prevalent. 95% of adults over the age of 60 have at least one chronic condition that they're managing. If we can help them effectively manage these chronic conditions, it can lead to cost savings for payers. For providers, it means increased patient satisfaction and also so better clinical outcomes. And lastly, the patient, we can help them really enhance their Quality of life. So we have so many of these high risk groups, it gets super, super hard to keep track of all them. So we created a population health dashboard and I am going to have David show you, but we have it organized by clinical risk location and more. But I'm going to let David take it away and show you what it looks like.

Thanks, Alyssa. Though this is a simplified population health dashboard. Sorry, I can hear myself there. Though this is a simplified population health dashboard. This should be something that you're fairly used to seeing, especially with those dimensions that Alyssa had mentioned there.

So David, how do I use the existing dimensions that I just talked about so we can track and calculate for each com, each patient to my advantage?

We're going to do that by taking these different dimensions and turning them into segmentation data. So when we look at our population Health Dashboard here, we're going to go ahead and filter that dashboard to look for specific types of patients. Here we'll look for those diabetics. We'll also look for those patients who are in our highest risk category. And lastly, we're gonna identify all those patients who are also smokers. Now with this list of roughly 10,000 patients finalized here, we're gonna take this list and more importantly these segments or this segment we've created and send it over to Health Cloud. There we go.

We know that patients today are not just wanting, but quite frankly demanding personalized engagement. So you might be thinking, okay, we have these groups, but how do we then start to personalize the care for them? 61% of patients do not fully adhere to their care plans. This could be for a number of reasons. This could be maybe because they don't understand it, they don't understand their unique needs, or maybe they're just not engaging. We have these segmented groups and now we can begin to hyper personalize outreach. So David, if I am a marketing specialist, how do I use this information in generative AI to start outreaching to these patient groups?

We can take this information and our segment information and start to hyper personalize our outreach to our patients. Let's go ahead and follow that thread of our diabetic smokers who are high risk. When we launch into this segment, we can see see a few of those details that we had identified previously, like the population size. We can also see things like where is this segment already being campaigned to now? These metrics are going to be what we use for that long term success of the segment and the outreach to that segment. Let's go ahead and launch our campaign wizard to see how we can use Salesforce automation as well as generative AI to really accelerate our concept to outreach. We'll go into our campaign assistant here and in the brief generator, we'll give it a simple prompt that is telling it that we want to identify those patients who are diet or interventions for those patients who are diabetics who are also looking to stop smoking. And we also want to identify the programs at McConaugHealth that would be most beneficial.

All right, so these details look really good, but what if I maybe want to spruce them up or make any changes on the fly?

The benefit of utilizing this wizard like framework is that each of these fields is just a text field. It's completely editable. What Einstein is doing for us is providing us that creative starting point such that we can go ahead and get this campaign created faster.

Seems to me like these are cheat codes to get us where we want to go faster.

I totally agree. Now let's go ahead and create this draft. Now, for those of you who have worked in Salesforce before, you're probably used to seeing a template like this for a templated email. What is new though, is some of those tools that Einstein's going to start incorporating into our marketing cloud. Before I start to change any of the details of the actual or content here, I want to change the details of the message itself. So when I click on Generate, I can go into our paragraph generator. You can see some of the details are coming over from that campaign, but we're also beefing it up with a few other details like program information from our diabetes and our smoking cessation programs. Segment details, but we're also giving it some new instructions. We're asking it to take a look at deciding what the appropriate age groupings are. And we'd like it to divide it into three different age groupings to generate these messages. Then we click generate paragraphs. We'll see what it comes up with.

It looks like it's generated three paragraphs and separated them by age range. How does Genai account for some of these nuances?

Let's take a look at each of these specifically. So when we look at our young adults, we can see that Einstein's prioritizing things like cutting edge technology, virtual care management, or one on one counseling. On the other side of this, when we look at our eldest age group, we can see a little bit softer messaging for those folks who are maybe a little bit more set in their ways. We want to give them that information and really give them the power to take their health care back, and we want to communicate that message to them. Now let's go ahead and use our middle age group here. We can just select that and add directly into our content manager here or our template. Of course. I fast forwarded a little bit With a few additional detail changes, we can really make this feel like a Makana health email. And we're ready to send this out to our community.

I actually really love how this looks.

When you look at this. I also love how this looks, Alyssa, when you look at this. This is a personalized email going directly to Lisa. Now, the unique thing here is we typically, we typically market to the masses. All of the diabetics received the same message. This is not a message for the masses. It's not for all the diabetics or all the smokers. It is a streamlined message for Lisa and the people who look just like Lisa.

This almost looks like it's a personalized health newsletter for each individual patient.

Absolutely love it.

Okay, so now we have the communication down, we start to think, how can we help our patients now that we've started to communicate with them? Let's better understand them. Right. So we have patients that need integrated holistic care plans. So 61% of patients do not fully adhere to their treatment regimens. This could be for a various number of reasons that we have previously discussed. So if I am putting myself into the situation of a care coordinator, how can a care coordinator, David, leverage this information and generative AI and how does Salesforce bring the point of patient care?

Absolutely. So when we go into our care management environment, we can see exactly how our care managers can leverage we can see exactly how our care managers can leverage all the information that's in the C360 platform as well as AI to really go about their day. When we log into the care management platform, we can see things like today's tasks. We can see our unified health score, so we can manage patients of changing risk as well as those program enrollees for which I'm going to end up being responsible. Now, of course, we can see that Lisa here is our newest program enrollee for the diabetes program. And when we dive into her profile, we really start to see where that C360 data comes into play. We can see any of the device data that she started to share with our organization. We can see those social determinants of health and really start to track those against the timeline that we're managing her for. Last thing I'll show, at least we can see that United Health Score and how it's changed Unified Health Score and how it's changing over time directly for Lisa. Now, because Lisa is a new patient and she's got a relatively low risk score here, our health score, but it is rising risk. We are going to see that Einstein is recommending that we set up a a bi weekly communication plan with her as we're beginning to engage her in this care plan.

It looks like there's a few recommended options that we can take to engage with Lisa in her care plan. How does Health Cloud know which one of those actions to take?

Absolutely. So what's great about Salesforce is that we can program those recommended actions. Einstein can learn from our actions as we're utilizing the platform and really take how we utilize that platform and deliver the right insights to the right end user at the right time for the right patient. But we can see that it's starting to learn from us. So we'll go ahead and enroll Lisa in our bi weekly communication plan here for glucose monitoring. You can see it's doing a few things in the background here as it's automating some of what happens. Now we'll go into our care plans. We can see that there's some care plan information here for Lisa. And these care plans, because they're created automatically, we can associate any of the details that we'd like directly with that care plan. The number one thing that I want to point out here is the assignments. You can see Mason's assigned here our care manager as well as Lisa Green. This is a co management agreement, if you will, as a care plan. Really giving us access as care managers as well as our patients access to that data.

Thanks, David. I appreciate you showing us how it works. Makes me really excited for AI in healthcare and what the future looks like. So what is the steel thread between all of this stuff? What makes all of this stuff work? And that's data. But healthcare data is disconnected. Remember that stat that we talked about in the beginning about how so much of the world's data is healthcare data and how it gets really overwhelming. This problem is typically twofold. Not only is the data disconnected, but sometimes the systems that are supporting these healthcare organizations are also disconnected. So with that being said, David, how do we begin to develop a holistic profile of a patient if all this data is just super disconnected?

The answer to that is data cloud. We can create that unified C360 profile utilizing something like data cloud. Now today we saw how we have calculated scores in Tableau. We have generative AI taking Context from around our organization to create new generated content. And we also have seen that C360 care management model with recommendations identified. These all use the power of data cloud to take that silo data and bring it into the Salesforce environment. Now, when I say siloed data, I truly mean data from any platform, any other siloed area. We take it, ingest it into the data cloud platform, unify that Data under a C360 model, and then we can take all of that siloed data or that typically siloed data that's aggregated in places, and give it the power of the Salesforce automation and activation platform.

Thanks, David. All right, so in just a second, I am going to bring up 2 of our Salesforce customers and leaders in the healthcare industry and we are going to do a panel interview. All right, good to have you guys here today. So I can't wait to chat. We have lots to catch up on. Right. Lots happening in the healthcare space. We talked a lot about hyper Personalization today. How are you moving towards hyper personalization? What are some of the things you're doing and is it working?

Yeah, I'll take this first. Hyper personalization is at the core of our mission at Same Sky Health. We're a health equity company that's really focused around getting the right content and the right messaging to patients and members with multiple chronic factors. So they may have chronic conditions, they may have maybe at different life stages, and they have some underlying social risk. Right. So kind of the SVI that you were demoing earlier in your. In your campaign, your population health dashboard. So Same sky exists and we've. We are leveraging this capability within Salesforce to deliver very targeted engagement and messaging in dozens of languages across all different life stages, all different social risks. And that each requires lots of aspects of marketing technology. Right. That you think about that you would get in other kind of personalized messaging in other parts of your life, but really marketing those services to each individual patient and member. Wow.

Sounds like you guys are up to some exciting stuff. What about you, Jose?

Yeah, so when we think about, you know, hyper personalization, we have to put it in the context of healthcare and how important privacy is and being within the framework of hipaa. Right. So that's the first. The first step that we take is understand what it means for us from that perspective, and then we take it to the next step, which is understanding our consumers in general. Where are they at in their health journey. We can use technology to understand that, technologies like Data Cloud to see previous engagements to see interactions that they've had in the past and other additional pieces of information that may be relevant to the healthcare journey of those individuals. And then as we're processing all that information and we utilize an omnichannel strategy to reach out to them via text, email, etc. We can enable campaigns or processes that can meet them where they are and they can receive a message that is relevant to them. So when we look at campaigns that we have enabled in the past, you know, nudge campaigns that help our patients, you know, take care of themselves, maybe close some healthcare care gaps, we've seen a significant uplift in visits scheduled after some of the messages that we've sent in the past. So that has been very successful.

Wow, Sounds like data cloud's been such a powerful tool for you.

Indeed.

I love it. Okay. And we talked about, in the presentation about the 95% of adults managing over 60 managing a chronic condition. Have you done any campaigns or anything around chronic condition management and what were some of your results from that, if any?

Yeah. So we have somewhat of a unique perspective here. I mean, our primary customers are large national health plans. Right. So they'll provide us populations that have many, oftentimes many comorbid chronic conditions. And one of the things that we've discovered engaging with these vulnerable populations is there's also a combination of risks where they may also have comorbid SDOH factors that are underlying that non compliance with their chronic condition management. And so we've started to layer in to our messaging and to our education ways that they can help address those underlying SDOH risk factors to help them ultimately close those clinical gaps in care. But back to kind of the personalization. What we think about it is we have the member longitudinally. Right. So we're looking at that full patient 360 experience, not just a combination of their clinical factors, but also the social risk factors to build pathways that are really targeting each of those different points in their care journey. Right. We're talking to a real person. They're not going to be able to address all of their gaps in care in one day. Right. So you really have to bring them along, educate them, build engagement, build trust in the channel. Once you've activated that channel, now you can start to bring them along as they make progress as individuals. Right. Again, we're thinking about macro population health type of processes. This is what doesn't work. We've learned this doesn't work in healthcare in the past. Right. Just blasting one message out to all of your diabetics and thinking that everybody's gonna take the right next step. Right. Somebody might take it on message one. Somebody might take it on message 12.

Right.

You really have to be able to tailor and hyper personalize each of those touch points to really activate and move your population. Wow.

There's certainly a lot of benefits for patients in there too.

I mean, ultimately, it's all about giving the right patient the information that they need at that touch point, which could be totally different than what somebody else needs. Right. Talking about gen AI, maybe it's written at a different reading level. Maybe it's in a different language. Right. I mean, there's all kinds of factors that go into those personalization elements.

Absolutely. Jose, what about you?

Yeah. So when two things come to mind from our perspective. You know, when you look at early indicators of healthcare needs and analyze all that data, then you're able to reach out to your consumers where they need to be. Right. Think about, you know, an early indicator that it tells you they might need a cardiologist in the location where they. Where they're at, an oncologist. So we're able to trigger campaigns with this information that is meaningful to the individuals in the area where they're at and with all the information that we have about them. The second one is around med adherence, a very important component. We have a pilot program right now around med tracking, which allows us to communicate with our patients at the time when they need to take the medication, and then we can send them reminders, and we can keep it as a bidirectional conversation. Would they reach out and tell us, you know, have they taken the medication or not? And if we don't receive information back from them, we're able to follow up to get them back on track. So, you know, when we think about that population importance of medications with some of these conditions, this. This is something very, very powerful.

Yeah. That is truly personalized. When we talk about personalized there, it does not get any more personalized than that. So if a patient misses a dose, you actually contact them and tell them, like, hey, is everything okay? We saw that you didn't reach out to us. That's amazing. Good to hear. All right. We talked a lot about the disparate data and how there are a ton of data silos. How are you managing your data?

Yeah, I mean, this is the real hairy problem that we're dealing with in healthcare. Right. I mean, it's 30% of the world's data, as you said, it's generated in all kinds of standard formats. Right. That's the great thing about standards. Everybody gets their own. Right. And we're dealing with integrating clinical data with payer data, with now SDOH data. And so really being able to bring that together. A tool like Data Cloud, right. Is allowing us to bring in all of these disparate data sources, normalize them in a standard ontology and then activate, we call them community health guides. Right. So it's a combination of our technology and platform doing omnichannel digital outreach.

Along.

With these community health guides that are helping real humans, that are providing that last mile human touch. So being able to bring that 360 disparate data together, map it and normalize it, build those long term profiles, now we can produce something that's both understandable by our automation as well as real humans that are having those one on one patient interactions. It's, it's really powerful. Right. I mean ultimately it's about knowing the real details of the patient, what actions they're taking. In our case, we're dealing with these SDOH risks. Right. So some of you may be dealing with social risk needs assessments. Right. Think about stuff like the Gravity Project where we have validated instruments that are collecting all of these social risk data and then we want to make that actionable in the platform. Right. So something like Data Cloud can help us bring in those well understood social risk concepts now, match them into the part of the health cloud data profile where that data would normally reside. And now it's just part of that 360 patient experience and we can fold it right into the communication strategy long term. So that's just one example of the power in the OSDH space. SDOH space.

Sounds like you've got tabs on your data. No disorder data here.

Don't take anything for granted.

But yes, so far, definitely a lot of work. When we think about creating that 360 view of a patient, I mean that's the best way you can engage with your consumers. It's about removing blind spots and reducing friction when you're engaging with them. As you're gathering all this data spread across a very complex landscape, your engagements become more meaningful and your interactions are significantly better. People feel like they know what you're talking about, they understand that you're there and you can meet them where they're at in their healthcare journey. So that has been fantastic. Being able to unify all the data, expose it to individuals and make use of that data and our interactions is, is key.

Absolutely. It makes them feel like you know them. Right. Having all that data. So now for the age old question I have to ask, I know everyone is wanting to ask too, what are you guys doing with AI right now? And if you're not using it right now, that's totally fine but how are you planning to use it in the future?

So we are very carefully using Gen AI for certain aspects of our personalization. You know, I just want to stress we in healthcare maybe have a different set of rigor and requirements around what we're doing with Gen AI than some of our other industry esteemed industry panel cohorts. Right. So we use it very judiciously and with, with tons of human supervision. Right. And again this is one of the things that we appreciate about Salesforce is you very much have a focus around ethics in the way that your AI is being trained and the way that that data is being used and how it's actually being applied in the end result. So how we use Genai is we're using it as part of our hyper personalization strategy where we're able to feed in same sky has been around for many, many years before Genai. So we've generated tons of content. We have a lot of outcomes around what messaging has been effective with which populations both clinically and on the social risk side. And we're using the Genai capabilities to help us build new messaging for maybe new patterns of content, new comorbidities, new social risks where we can use Gen AI to generate many, many proposed messaging content. You may have it generate 30 different flavors. Gen AI is not perfect yet, so maybe half of those flavors are no good. But you have human supervision to make sure that yeah, you're picking the best messaging. And then also in healthcare again this is going to go through many layers of rigor and review. So think about messaging that we're producing. It might be for a plan that is operating in a state for a Medicaid line of business. You're going to have to have that go through our content review. It's going to have to go through the payer's legal and content team. It's going to have to be approved by the state. Right. So there's a lot of of filters to safely roll out Gen AI content before it ever really interacts with a patient one on one. So I think there it's more about building the automation and the, and the management of your ML pipelines. Right. For how you're going to get that AI generated content safely in front of patients. Maybe more so than like the fun interesting part of just generating Lots of, you know, AI whiz bang stuff.

So safely is the keyword there.

Yeah, for sure.

Spreading lightly.

I mean, it can be done safely, right? I mean, you just have to use the tool for what it's good for, for sure.

Jose, what about you?

And that makes sense. Yeah, you know, in healthcare it's a little, a little touchy. At Ascension, we're still formulating our AI strategy from an engagement perspective. There are definitely use cases that are, that are interesting that we could explore in the future. You know, content generation is, is definitely an area that, that looks very promising. It's not about the speed at which you can deliver the content, but more about personalizing that content appropriately. So that sounds very exciting. Summarization of data is also relevant, making sure that people can find the information that they need at the point of engagement as well, whether it's via email or a direct interaction between humans. But yeah, we're still in the process, so more to come on that treading lightly there too.

I love it. Doing it safely. Well, thank you guys so much for joining us. It was great to hear what you're planning on doing and what you're doing right now. I can't wait to maybe have this conversation a year from now and see what you're doing with Gen AI. Maybe it'll be more widely used in the healthcare space by then. But Gen AI, you know, getting there is a journey. It doesn't happen overnight. Rome wasn't built in a day. Gen AI was not built in a day, especially in the healthcare space. So everyone always asks, what can I do right now? Well, you have to prepare your data. And to prepare your data, I really just mean establish a single source of truth with your data. Stage two is sprinkling in some of those predictions and generic AI capabilities that we already have right now. And then lastly, that's adding in the Gen AI. Who knows when we'll get there. Hopefully it's sometime soon. But yeah, there is a process we have outlined for you and if you didn't get enough of us today, go check out healthcloud.com we have a lot of cool things you can look at. What's coming up, what we're releasing a lot of really good stuff, so check it out. And lastly, without further ado, thank you so much. Enjoy the rest of World Tour Chicago.