Rae Woods (00:02):
From Advisory Board, we are bringing you a Radio Advisory, your weekly download on how to untangle healthcare's most pressing challenges. My name is Rachel Woods. You can call me Rae. This week I'm kicking off a three-part series on something we've actually never talked about on Radio Advisory before, and that's consumer data, meaning the use of highly specific data to market and personalize the consumer experience. It's something that's already revolutionizing business, and it's going to have a big impact on healthcare as well. And for all the potential this data has to provide a better, more personalized experience, more personalized care, there are also some legitimate concerns.

(00:48):
In the first part of our series, I want to focus on what consumer data means, and how different stakeholders should be thinking about the opportunities and the threats that come with this industry change. I've invited two Advisory Board experts to today's episode, life sciences leader Solomon Banjo, and digital health researcher Ty Aderhold. Hey, Solomon. Hey, Ty. Welcome back to Radio Advisory.

Ty Aderhold (01:13):
Glad to be here.

Solomon Banjo (01:14):
Great to be back, Rae.

Rae Woods (01:15):
You're both longtime guests, but I don't think you've ever actually been on the podcast at the same time. Why is it important that both of you are here for this conversation, besides the fact that maybe you're both data nerds?

Solomon Banjo (01:29):
Well, that's a given. But I also think we bring very complementary perspectives to this, where obviously, not to speak for Ty, but the digital health angle, kind of obvious. But what may not be as obvious for listeners is that part of what we do on the life sciences team is thinking about the different sources and use of evidence and data as well. And a lot of what we're going to be talking about is sort of a snake eating its tale, where it's both digital health and about the evidence in that way.

Rae Woods (02:15):
The three of us have all had the experience of getting an ad, whether it's on a webpage or what have you, that's ultra personalized. I'm thinking about moments when I'm surfing the web, I'm on social media, and I get an ad to purchase an item. I'm always shocked that they seem to come up with the thing that I need in the moment, which speaks to the fact that a lot of these companies know a lot about us, more than I even can make sense of. When we talk about consumer data, and particularly consumer health data, what actually are we talking about? What kinds of data are being collected? How is it actually being used?

Ty Aderhold (02:54):
Think about anything that your smartphone or your computer is collecting. So your location data, the websites you visit, the apps that you use, when you use those apps and websites, where you click, or what you do in certain situations within those products or apps or websites. All of that is collected and is able to be used. Solomon, I'm sure there's others out there I'm missing out. Anything that jumps to mind for you in addition to what I listed?

Solomon Banjo (03:28):
The other thing that jumps to mind to me is actually something you are all too familiar with, Rae, which is now with the spread of telehealth, we are seeing information that normally we would only say to a clinician in person, and therefore it would be covered by HIPAA, that is now in this gray area of, if I fill out a form for Amazon or you name it, giving them information that only I or my clinician would know, in order for them to triage and treat me appropriately, is that consumer data? And now, what is possible when you connect that data to everything Ty was just talking about?

Rae Woods (04:08):
And our listeners might not be aware of the fact that you're literally making fun of me right now, Solomon, because we were in a meeting at Advisory Board talking about, in this case it was Amazon's initial acquisition of One Medical. We were talking about what it all meant. And in the middle of this meeting I was like, "I'm just going to get on and I'm going to try to make an appointment, and I'm just going to start using the app now." And willingly gave away a ton of information, and accidentally paid money and then realized, "Oh no, what have I done?" Which speaks to the maybe slippery slope of all of this. And I think that our industry and the leaders that are listening to this episode are familiar with consumer data being used outside of healthcare. Targeted marketing for the goal of purchasing more goods. Can you give me an example of that slippery slope being used to get into the health side?

Solomon Banjo (05:04):
So you talked about targeted advertising, and so I can't pass up the opportunity for a good pun. And probably the example that a lot of people think of is Target, and headlines they made a decade ago now about their ability to predict when someone was pregnant, even before they were aware, or in this case before their parent was aware, and wondering why they got all of this messaging about maternal care when they weren't pregnant. Surprise, the story goes, that they were actually pregnant.

(05:37):
Now, a few things to unpack, because in the subsequent decade there's been a lot of journalism pushing back against this. Two things that strike me are, one, we don't talk about all the people who got that same mailer who were not pregnant, which tees up the second thing. No harm, no foul. Target sends me, Solomon Banjo, a thing saying I'm pregnant, which is not something that I am or capable of being. I throw it in the trash. If using that same information, you're making decisions about resources or my clinical care and you get it wrong, different ballgame, different stakes there. And I think that is an important thing to keep in mind when we think about how and when this data will be used, because it probably will be used eventually around clinical decisions.

Rae Woods (06:24):
So how are healthcare organizations looking to use this data in order to support their business right now?

Ty Aderhold (06:30):
It goes back to this idea of targeted marketing that we're talking about, but also behavior modification. This is something we've heard used in the consumer data world as, you can take the information you capture, and very specifically market based on that information to influence choices people make. And so in a healthcare setting, you can imagine organizations, particularly ones, say, they have a high value procedure, that they earn a fair amount of revenue if they get more patients into their business to do this. It's a huge opportunity for them to identify who those consumers of that service will be, and how to effectively market to those patients, to get them to choose their organization for that service.

Solomon Banjo (07:17):
And I want to build on something Ty just said, because I think it is important to think about that behavioral modification piece, because we are so bad at that in healthcare. How different would our healthcare spend and outcomes look, if we in healthcare could successfully get people to eat healthy, exercise, take care of their mental health? All of these things that hinge on a type of behavior modification. And so I think there's also that piece of, if we had a better understanding holistically of the human, could we see some of that behavior change? Because I don't know about you all, but if I look across my desk, there is stuff that Instagram got me to buy that I wasn't planning to buy.

Rae Woods (08:03):
Oh, yeah.

Solomon Banjo (08:04):
And so, what's the healthcare equivalent of that?

Rae Woods (08:06):
And what you're getting at is the idea that, in healthcare, this is more than just targeted marketing. It can actually be used to inflect clinical outcomes.

Ty Aderhold (08:16):
The downside is that it lets existing incentives be taken to an extreme. So, something that you will often hear provider organizations talk about is their payer mix. They are always going to be worried about their payer mix.

Rae Woods (08:30):
And rightfully so, right?

Ty Aderhold (08:31):
Rightfully so.

Rae Woods (08:32):
The standpoint for providers, particularly health systems today, is... I keep using the word fragile, and somehow it gets more fragile as time goes on. That's why, from a business standpoint, health leaders have been focused on their payer mix.

Ty Aderhold (08:45):
Yes.

Rae Woods (08:45):
But...

Ty Aderhold (08:46):
But from a business standpoint, this ability to use consumer data and modify behavior can take that to an extreme. It can bring the ability to control your payer mix, to capture high value patients, I'm using air quotes here, "high value patients," maybe avoid having "low value patients" come in for your services. It takes that ability to an entirely new level when you have so much more information on what makes those groups of patients act in certain ways, what makes them come to your facility or not, what makes them choose you or not, and the ability to potentially modify that behavior through digital channels.

Solomon Banjo (09:30):
And let me tease out something Ty said as well, which is consumer data. We keep saying, the value of data is it helps us make better decisions, in theory. But in healthcare, we know that we do not have the same level of data on everyone, because of how they interact with the system, how often they interact with the systems. And so when I think about, if we take the data that is being collected and generated and start to make decisions and iterating on those decisions, to use Ty's words, we could take that to an extreme where we're codifying and even potentially accelerating the bias already within the system. Because if we could almost, Princess Leia style, have the data sources show, "Here are the people I represent," it's not going to be as diverse as what we would see in the census. We would not have the same clarity for all different patient groups. And so if we start using data to make resource decisions, who are we not giving those resources, or what assumptions are we making about them due to imperfect data that hampers their outcomes?

Rae Woods (10:38):
It sounds like even when these data are used for a clear business goal, like that targeted marketing example, if we stop there, it will result in the outcome that you, Solomon and Ty, are describing, where if we are focused on just getting those most profitable patients for our orthopedic procedure, standard thing that health systems do forever, that naturally is pushing certain populations to the bottom of the list. And that's before you even take things to the more extreme examples that you two are describing. So, what do we do to make sure that health systems are making the kinds of moves that help them get out of this place of fragility, while also making sure that we're upholding positive outcomes for everyone? It actually feels like a very delicate balance. Is anyone even thinking about that yet?

Ty Aderhold (11:29):
I guarantee you health systems are thinking about the orthopedics example you gave. I don't know if health systems have started to think about the long-term impacts, the ten years down the line, what does this look like for patients in our market impacts of those decisions. The biggest push I would have is to also use this data for other benefits. In addition to those couple examples we are talking about, there is so much potential to take consumer data, outcomes data, and to better understand how to interact and treat a wider variety of populations. And so if we only focus on how do I interact with and treat the high value populations, that affects your bottom line right away, and that is something that, Rae, you've already said, you have to focus on. But at the same time, you can also focus on uses of this data that are more around the outcome side of thing.

Solomon Banjo (12:37):
Part of the challenge too is where we speak about providers of life sciences as a monolith, but if we dive deeper, I'm sure there are people who could talk circles around us in terms of the understanding of these challenges. The question is, are their perspectives widely known through the organization, and are they in conversations where that perspective is helpful? And I say this because, Ty, I'm thinking about the example of the health systems who, through their use of cookies and other things, were sending very sensitive health information about patients to third parties, the Facebooks of this world. I'm sure the people who made the decision that ultimately led to that data breach were not thinking about that. They were just like, "This is industry practice," and did not think about how that might...

Rae Woods (13:28):
They're doing their job.

Solomon Banjo (13:30):
They're doing their job. And so how do we think, whether you are a life science provider, data vendor, health system, you name it, health plan, how do we think cross-functionally about the data assets, so we understand them and their limitations? And then also have the right structures in place to make sure that we are using them in the right ways as we're applying them to different problems, so we're not asking data to do something that quite frankly it's not up to, and/or we're thinking about the second or third order impact of how we might use this in this one domain. I think that is the challenge.

Rae Woods (14:58):
I want to talk more about the individual stakeholders here, because I'm aware of the fact that thus far we've been mostly talking about providers, and I want to make sure we're giving airtime to other kinds of stakeholders. But before we do that, I also feel like we should remind our audience that health equity is also a business imperative. And if it's not, it should be for your organization. And just like revenue and profit is a business imperative, that we can be thinking about creative ways to use consumer data, to Ty's point, we should be using these data for other business imperatives, and equity is one of them as well. It's not just a matter of preventing the worst case scenario from happening. It's a matter of making sure that your investments are meeting your stated goals, and that includes equity.

(15:53):
All right, I'm getting off my soapbox. Let's come back to the stakeholders. We've been talking about providers a lot thus far. Is there an example that we haven't talked about yet that you want to make sure our listeners are aware of, in terms of how the actual care delivery part of the ecosystem can be using consumer data?

Solomon Banjo (16:11):
One example that comes to mind, we're talking a lot and have been for the past 18 months about the Walgreens, CVS's, these retail originating organizations, and some of the benefits they might have. There are a few things that come to mind. One, their core competency includes using this consumer data. So they have a fluency in using a lot of the data Ty was talking about earlier, in ways that health systems don't. And second, for a lot of their business models, being able to keep Solomon as healthy and act early and often so that I don't exacerbate to the point I need to go outside of their ecosystems, their networks, could be something that's really beneficial. Because I bet all of us have a number with CVS.

Rae Woods (16:58):
Yeah.

Solomon Banjo (16:59):
So the ability to use that information, the other consumer data, and bring it together with their healthcare delivery arms, and try and keep us out of the hospital, especially for chronic patients, I think there's a lot of potential upside to this as they get a better understanding for, you three may have the same four conditions, but actually you're very different in meaningful ways that, if we can design and provide you support, allows you to be adherent, allows you to do all the things that you need to do to live a healthy life. And they come at it from a really intriguing standpoint, because they're used to dealing with this data in myriad ways.

Rae Woods (17:40):
So you're getting at the benefits in a total cost of care business model. And you mentioned CVS, which obviously bought Aetna and is now straddling this retail and also payer model. So what about the health plans? How can they use these data? What are the upside and downside that we want to be a warning system for, as organizations are using these data?

Ty Aderhold (18:06):
One example of how it can be used in a positive way is understanding the stages of life or anticipating future needs of members or employees, and getting them information sooner. We can even go back to that Target example, a mailer about pregnancy. If you understand when patients are entering a certain stage of life, and you can actively give them information, that can improve their outcomes, their health over time, if you are giving them the information they need, right when they need it or before they even realize, I should be looking for this information.

(18:43):
Another way that this can, I think, be a positive for payers and employers is, it is the best data source you're going to have in terms of how these members or employees like to interact digitally. You will understand their preferences, when they do certain actions or take certain actions, in a way you can't otherwise. And that can really help you get closer to modifying some of those behaviors or improving chronic care, et cetera.

Rae Woods (19:16):
Because everybody's using these extremely broad categories about demographics that we know are almost always wrong.

Ty Aderhold (19:23):
Right.

Rae Woods (19:24):
With this better information, we can better interact with, in this case, members.

Ty Aderhold (19:28):
I will say there is a flip side and a downside to doing exactly that.

Rae Woods (19:32):
Bring it on.

Ty Aderhold (19:33):
Which is locking yourself into biases. So if you start to do this, and don't continue to mitigate against data bias, and actively update your assumptions based on consumers, we could be entering a world in which payers and employers are locking into more advanced metrics, but still biased metrics, same to those demographic points you just mentioned, Rae, where it's more complex, more advanced than just, these are the demographics and we'll do this for this demographic set, but it still, at the end of the day, will lead to bias and not a true personalized interaction for patients, and therefore not get at the end goal.

Rae Woods (20:18):
All right, Solomon. It's finally your time. What about the life sciences companies? You are the life sciences researcher on the call right now.

Solomon Banjo (20:26):
Yes. To channel it, and this is true for both pharma and device, but a few things that come to mind. One, I don't know if people recognize how much time these organizations spend trying to map out the patient journey. So saying, "I'm trying to inflect this specific patient type or condition, let me understand how they get there. How long does it take for them to be diagnosed? Where do they tend to be diagnosed?" Teasing out all of these pieces, similar to what Ty was saying from a health plan perspective, as you can get more nuanced in understanding the patient journey, how even for the same condition and treatment that varies by meaningful categories. You can then better understand the support approaches, how to engage patients and the relevant clinicians for the products.

(21:16):
There's also, if we're being really optimistic, the opportunity for you to design your products better. To think about an example I've heard recently from even the clinical trial spaces, as they're using more and more wearables, understanding, what are actually the fluctuations in a lot of biomarkers and digital biomarkers that are par for the course? That if we only measure it when Solomon comes in person, we only get that one data point, we make a lot of inferences there. But if I have that data for Solomon for 24 hours every day of the trial, what then might that help me understand about changes in physiological response, side effects, that I can then used to design the product more optimally?

Rae Woods (22:05):
Okay. You said you were being optimistic. Give me the pessimistic side. Ty did it. It's your turn.

Solomon Banjo (22:08):
Yes. I mean, it always has to be the questions of bias. Who are we designing for? I also think that, differently from how we've talked about providers and even health plans in many ways, they have their own proprietary data that they're able to collect through their business about those patients. For a lot of life science organizations, they're purchasing that. And so as you think about your partners, for who you get your claims data, your EHR data, now your consumer data, if you pick the wrong partner, who intentionally or not does something that gets them in the news in a negative light, that reputational blowback and the risk associated with it is very real. And so that's the other element here is, beyond just the risks of how you apply this and enforcing bias is, depending on who I partner with, it might actually end up being really bad for me as well, if they're not approaching this with the same level of conscientiousness I would want them to be doing.

Rae Woods (23:12):
And that's so important, because we're in this time where right now, consumers, patients, people everywhere want more personalized data, want more personalized interactions, want more personalized care. And yet there are also very real concerns about data privacy. Hence me going, "Oh my God, what information did I just give to Amazon?" after that meeting. How are consumers actually experiencing this industry change?

Ty Aderhold (23:44):
There is a very fine line between personalization and creepiness. And I think that will be so important for healthcare organizations to figure out over time. And it can change patient to patient. One patient may love that you sent a mailer or an email following up after they've done Google searches, to talk about what their Google search was about. Another might find that extremely creepy and not want to interact with your organization, because how do they know this about me? How could they possibly be sending this to me? I don't like this, I need to opt out and find a way to not interact with them. And so it's really going to be a key engager or disengager for organizations, and finding that fine line of, we are reaching out proactively for the right things at the right time, and not reaching out too soon or too much about things that we know, but that a patient may not realize we know or want us to necessarily know, is going to be so important for organizations to solve for.

Rae Woods (24:54):
There is one other stakeholder that we haven't actually talked about yet, and that's the government. What kind of maybe responsibility does the government have to protect consumer data?

Solomon Banjo (25:05):
I'd say one of the changes we've noticed is the role of the FTC, where in previous years, let's go three, four years ago, they were not really someone we're thinking a lot about in healthcare. But not only do they seem to be explicitly focused on a lot of what Ty and I were talking here, they have different tools and approaches that a lot of healthcare stakeholders, industry agnostic, are not used to thinking about. And so I think it actually in many ways raises the risk of not being deliberate here, because to take something out of the headlines, we have what happened with GoodRx, where as a result of the practices they had been doing, the government essentially said, "You specifically can't do this any more. As part of the settlement, you can't... Sure, no admission of wrong, but you can't do this any more." So if the data actually ends up being capable of all the great things Ty and I said, they can't play in that any more. They're going to miss out on what that could have meant for their business and for their patients.

Rae Woods (26:10):
Because they messed up.

Solomon Banjo (26:11):
Because they messed up.

Ty Aderhold (26:13):
And I think the reason Solomon brings up the FTC even is that right now, there's a bit of a gap here, where there's not a ton of regulation around the use of consumer data in healthcare, or the use of healthcare data in some of these more consumer focused apps that aren't covered under HIPAA. And so I think that is why we are starting to see headlines around this and some of that movement from the FTC.

Rae Woods (26:39):
Which is probably why we should pause here, because we have enough to say on the FTC, including the example with GoodRx, that we should do a whole other episode on that. So I will say, Solomon, Ty, thank you so much for coming on Radio Advisory. Ty, this is going to be about digital health, so do you want to come back to the next episode?

Ty Aderhold (26:57):
Sign me up. I'll come back.

Rae Woods (27:01):
All right. Thanks, both.

Ty Aderhold (27:02):
Thanks Rae.

Solomon Banjo (27:02):
Thanks, Rae.

Rae Woods (27:07):
The most important thing that I heard from Solomon and Ty is that consumer data can improve equity by painting a more holistic picture of patient health. And it can do the opposite if we're not careful. So remember, as always, we're here to help.

(27:30):
Make sure you tune into the next two weeks of Radio Advisory. We're going to continue our series on consumer data. Next week, we're talking all about the FTC. If you like Radio Advisory, please share it with your networks, subscribe wherever you get your podcasts, and leave a rating and a review. Radio Advisory is a production of Advisory Board. This episode was produced by me, Rae Woods, as well as Katy Anderson, Kristin Myers, and Atticus Raasch. This episode was edited by Josh Rogers, with technical support by Chris Phelps, Joe Shrum, and Dan Tayag. Additional support was provided by Carson Sisk and Leanne Elston. Thanks For listening.