The a16z Show - a16z Podcast: Dark Data in Healthcare

Episode Date: January 16, 2019

with Susannah Fox (@susannahfox), Anil Sethi (@anilsethiusa / @ciitizencorp), Vijay Pande (@vijaypande), and Sonal Chokshi (@smc90) The problem of "dark data" in healthcare isn't just a feel-good empowerment thing, but a structural issue that leads to miscommunication and extra friction, different players in the entire healthcare system not being able to collaborate with each other, and just major missed opportunities all round. And yes, it also leads to lack of empowerment for patients, not to mention doctors too (who often have less than 30 minutes on site to do their jobs). But we already know all that. What's not clear is WHY and HOW is this the case, when the very point of HIPAA -- the Health Insurance Portability and Accountability Act (of 1996!) -- is to make data portable, not private. That is, IF patients know to ask for it... and can easily get it. So what if we could have a sort of permissioned "permissionless innovation" for healthcare data, not only bringing all that dark data to light, but more importantly -- borrowing from the history of internet innovation -- letting all sorts of expected and unexpected uses be built on top as a result? What happens when data and entities can talk to each other (à la APIs) through patients at the center of the circle of data? From the Dr. Google problem (or opportunity!) to clinical trials and even the opioid crisis, we -- Susannah Fox (former CTO of the U.S. Department of Health and Human Services); Anil Sethi (CEO and founder of Ciitizen); and a16z bio general partner Vijay Pande; in conversation with Sonal Chokshi -- explore all this and more in this episode of the a16z Podcast. Let there be light! Stay Updated:Find a16z on YouTube: YouTubeFind a16z on XFind a16z on LinkedInListen to the a16z Show on SpotifyListen to the a16z Show on Apple PodcastsFollow our host: https://twitter.com/eriktorenberg Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

The content here is for informational purposes only should not be taken as legal business tax or investment advice or be used to evaluate any investment or security and is not directed at any investors or potential investors in any A16Z fund. For more details, please see A16Z.com slash disclosures. Hi, everyone. Welcome to the A6 and Z podcast. I'm Sonal. Today we're doing another of our episodes in our series on healthcare challenges and opportunities. And the topic is all about dark data. access to data, patients taking back control of their data, and what else is possible when such data is brought to light? We therefore also discuss the interesting nuances and a bit of history of HIPAA, as well as briefly touch on clinical trials, the opioid crisis, and the case of Dr. Google. Is it a bad thing or even a good thing? So, joining me for this episode, our special guest are Susanna Fox, who was previously the CTO at the U.S. Department of Health and Human Services, which oversees all the health and human services agencies of the United States,

from public health abroad to Medicare and Medicaid at home. She's also an advisor to Citizen. Speaking of, we also have Anil Seti, founder and CEO of Citizen, a patient-centered platform for helping people collect, organize, and share their own health data, and last but not least, and in fact, he's the first voice you'll hear after mine and Susanna's. We have Vijay Pandey, General Partner on A6 and Z Bio. So let's start with this whole problem of,

dark data. And I don't mean to make it sound sinister, but it is kind of sinister because it's hidden from worldview. Why are we in this position in the first place? We're in this position because most of healthcare has grown up believing that they own the data. So professionals generally collect and hold the clinical data that drives a lot of clinical care. And yet, patients now have an expectation to have access to data in the modern world. Plus, there's a whole part of data that isn't even part of the clinical record. And so how might we bring all of these sources of data together? I think about the vestigial elements of where the whole health care system came from it. You know, origins, you know, 50, 100 years ago from fee for service,

like doctors coming to your house. And data then is like writing something in a notebook. We have electronic medical records now, but it's really not radically different. It's basically a doctor, writing, or a provider taking down some notes. It's a difference in medium, not actually and kind. What are the structural reasons? First of all, I'm coming at this a little skeptically because A, I always hear people talk about data on health care. And I understand it's important, but I'm skeptical because quite honestly, as a user, I barely know when my car registration is due. I don't really think having control of my health care data is really the thing I need next on my list. Luckily, I'm a healthy person, so maybe that's part of the reason. But still, I really can't see a lot of people

really using data. Let's talk about why this matters really. You mentioned a car analogy. Let's just think about the fact that I think we take better care of our cars than we do it of ourselves. Okay, that's a fair point. You know, I mean, and you have a fair about data cars. You have, like, dashboard lights that come on when there are problems, and then you actually go to a professional and you give them a full run of what's going on. You have a lot of sense of the car because you're driving all the time. And it has sensors.

I get a tire pressure sensor every three months or six months or something. And so that data that is sort of the clinical equivalent for the car is something that the mechanic needs to be able to improve things. What I think really is missing is sort of having the alignment incentive. So you want your car to work. You obviously want your body to work. And your provider wants to be able to help. But then how can that person do that if they don't have all the information? And so I think really now the question is how can people who really have the incentive to take care of themselves finally use modern technology to let that data actually help

if we can do it for cars, we should be able to do it for people. And you're absolutely right that most people don't want to engage in their health. Most people don't want to know what's going on. They're not necessarily ready to look into the mirror of data. I call it. called standing naked in front of the mirror of data. A lot of us aren't ready for that. Yeah, it's a little too much truth. But then when something happens, you're diagnosed with something or your child or your mom is diagnosed with something, that will trigger engagement, let me tell you.

And then people should have immediate access to the data and information that they need to make a decision. And that structurally hasn't been the case. And that actually, I think, is this moment that we're living through right now where there's changing expectations about what we should have access to. And so healthcare is rather slow in taking up this new culture of access. But I think it is changing, and I think it has to change. What's the driver for the change? One I would think of the top of my head is just the mobile phone, things like push-button experiences. I think one of the things that is a driver is we've got

10,000 boomers aging into age 65 and having a different sort of user experience. And because people are living longer doesn't mean we're living healthy longer. And so they're going to be dealing with lots of issues that their forebearers didn't have to deal with because people didn't live this long. I mean, 100 years ago or 120 years ago, the average lifespan was about 48. Now we've artificially extended that to about, let's say, 83. And these are approximate numbers. But who's to say that we're really improving the quality of life? So people are going to retire and become more engaged than their parents were in their own health care.

And I know my parents go to a doctor and say, here's what I can share with you, now save my life, please. I think the next future generations are going to be much more engaged. When I think about my parents and how they narrate their oral histories, so much of medicine is about oral histories. And they're immigrants to this country. They're well-educated. But they somehow do this weird thing where they feel like they have to have this best-beats. behavior and you can't actually report the problems. And it drives me insane because my dad will call me and be like, oh, yeah, I didn't tell the doctor this. Why? Go, it didn't come up. And I'm like,

your job is to report your history. That's why they're asking you all these questions. You don't have to do this model minority thing. You're behaving well for the doctor, for God's sake. And I get really upset because I love my parents. There is a break in the existing healthcare system where there is a over-reliance on oral histories as part of the medical experience that I would think if I could just send them in all of our notes, then that alone would be a huge improvement to the doctor to have that context. Well, and there's much less friction for sharing this type of information. So you think about all the data you can quickly share on Facebook or LinkedIn or something like that. That's really trivial to do.

If you try to do this 10 years ago, 15 years ago, you could come to the doctor with a bunch of CDs or DVDs and they could like load it in. But by the time they've done that thing, your appointment's over. You have 30 minutes most times. Yeah. And so friction is a really critical thing. if you can bring the friction from 30 minutes to a third of a second, that's when the doctor now is actually able to use that information. But, you know, people have under the HIPAA right of access,

they've got loads of capabilities that they don't know how to exercise because the friction that you're talking about is still very prominent. And I don't think it's people that are trying to put friction in front of patients. I think they just haven't been taught or don't know any better. I think most people at medical records systems and hospital and HIM facilities want to do the right thing on behalf of patients. We were talking about the car analogy. One of the things that other industries have done is surface information voluntarily so that the lights on the dashboard and the accelerometer and all of that navigation engagement system, they are meant to be used. Now imagine if all of that was dark and you had to open up the hood and look underneath.

and know how to discern that information. Our healthcare system is very much like that. So you brought up HIPAA, and that's something that everyone hears about. We talk about in the context of privacy, but you're saying it's an opportunity for people that because of HIPAA, they can have data. I usually hear it the other way around when it comes to startups and HIPAA and privacy. Can we quickly talk about what HIPAA is and how it came about and why it matters here? Well, I'll start by saying that one of the most important things to know about the acronym HIPA

is that the P stands for portability. Oh, I had no idea. What does HIPA stand for? So it's the Health Information Portability Act. The other important thing to know is the context that it was actually written in 1996. So throw your mind back to 1996 when it was about 10% of American adults had access to the internet. This was a time when we weren't sure what was really going to be the impact of technology on health care. And yet there were visionary people who said this is going to be important that people have access

and that people are able to take that data and have it centered on the consumer, centered on the patient. The idea is that patients should be able to get access to the data and redirect it the places that they want it to go. So in the modern world, we now talk about that in terms of an API, that somebody could get access to their data and directed to an app or directed to where they want to go or be the conduit that creates the interoperability between to, clinicians, which unfortunately is sometimes the case. Well, that's actually the spirit of HIPAA. The portability. Yes. Instead, it's been used kind of as a weapon as a way to say, stop what we're doing, let's get the lawyers involved. Yeah, privacy is important, but that's exactly the context I've always heard it in is a sort of, you can't do this because of HIPAA. But that's exactly the opposite what it actually is. There are very few sentences that a licensed professional, a nurse, practitioner,

or anyone can say to a patient that starts with, I can't do that. That's a HIPAA violation. There are very few mental health and sexual health carveouts. But most of the time, I am in the right one. I can ask my doctor and she can send me all of my information in any digital format I request as long as she's got it to my Gmail account and it's not a HIPAA violation. I think very few people know that the hip- I have no idea.

That's possible, actually. That's because it's driven by the patient. That is correct, right. Patients are actually outside of HIPAA. So you can ask, as a patient, a HIPAA qualified entity, to do something for you and share your information, whether it's through an API or whether it's through a PDF or email, and they have to do it. Most people don't know. They have that much flexibility. This is a very important idea.

If I were to visualize this structurally, you think about all these players in the health care system, you have insurance and hospitals and clinics and universities and universities and. I could keep going, pharma companies, et cetera. If you picture this whole ecosystem as a circle, right now what you're talking about is putting the patient at the center of that circle. And therefore, they can pull on all these pieces. But otherwise, these entities cannot necessarily talk to each other directly with the patient's data. The patient is at the center and just doesn't realize it. Ah, that's interesting. Yeah, taking your circle, most of the entities are sharing with each other, but not through the patient.

So the longest way around the circle from one side to another is around the circle. The shortest path is to the center. That's the diameter. That's where the patient is. Once you realize that the patient's at the center of it, you realize there's a huge opportunity there. If companies or startups empower the patient, then the patient can drive this whole thing. They can quarterback it. They can ask for the data.

They can send the data. Now the question is actually, what can you do with it? And what's the impact? I do want to talk about what you can do with it now. Because going back to the original question, of do people really need their data and does it really make a huge difference? I'm thinking of the analog of what happened with wearables. People have long talked about wearables.

And just like the analogy of the car, there are lots of sensors out there measuring our bodies. But the number one problem when wearables and data is nobody actually takes that as actionable data and does anything with it. So it's a case where you have a lot of data just gathering that's not being used. Are we talking about the same problem here? In the case of wearables, you're dealing with something that's very broad and very shallow and also often not very clinical like 10,000 steps versus 9,000 versus 11,000, that's largely just made up in terms of its clinical significance. Maybe let's turn it upside down, which is what are the clinical areas where you have clinical data that could create actionable

outcomes for patients? Maybe start with what is the greatest need? People, maybe that could easily die in a year or two. And, you know, people that have like mid to late stage cancer, as an obvious example. And there's data they have from numerous different areas and imaging and genomics and all of that. And it's just kind of amazing how much data there is now to get. And then now the question is, how do you gather that and how can you empower the doctors via the patients to actually make a difference? But in the case of cancer, aren't the doctors already sharing that data? So why does a patient need to be at the center of that? So the question we get asked is, what will we do with it? And the answer is you'll be able to share it with an oncologist who will know how to

operationalize that data and help treatment. For folks who know me, my little sister contracted a number of years ago, late stage metastatic breast cancer diagnosed straight into stage four because someone missed the diagnosis and blew it a year earlier when it was stage one. So the point is, in her last year of life, Tanya was seen at 14 facilities, all using, not all, but many of them using different EHR or medical record systems across multiple states, which have their own transmission of health data cross-state lines issues, and she was deemed by 23 oncologists.

So that sets the groundwork. Oh, my God. Every time she went to a new oncologist, either there was a restart, a frustration factor where she had to explain everything. But because she had her electronic health record, imagine a LinkedIn of your health profile. She had that. She had that. Okay. Most people don't even have that.

Her big brother was in the industry. So, of course, Tanya, you. used glimpse, which is a former company, Apple acquired it, and in a transaction, and some of the health records that we see them releasing it to the world is some of that technology. And so she had the ability to share her profile, her portable health record with all her meds, all her labs, all her genomic information. What we had done is we had built a depth of health record that could operationalize her cancer care. This is in contrast to others who think that a mile wide and an inch deep in data is the way to go. We actually think in order to

operationalize health data, you have to find self-incented people, and that is cancer patients, lupus, HIV, autoimmune disease. We found that we needed to add imaging and genomic, which glimpse didn't have. We want to democratize that across all seven billion people. It shouldn't just be my little sister. So one of the promises I made to my little sister is I would go on and continue this work. And so we are starting with cancer, moving towards autoimmune. All chronic diseases, actually. Correct.

That's what they all have in common. It's actually the most frequent touch points into the health care system. And you don't need to convince them of the value of their data. They know it. Yes. Anyone of the chronic condition knows, I know this. Yeah. But the three things that we have found that we absolutely need to solve for patients is lower the

friction of three touch points. So three long poles in the tent. One is how does someone actually recall? their records to be released. And it should be, eventually, something as easy as making a payment at Whole Foods with Apple Pay. Just as an example, someone should send you an invoice and you should be able to adjudicate the financial transaction by touching your thumb and boom, it's paid. We'd like to help health records released that easily.

So that's one of the long polls. The request point is one. The release of information or request point is one. That's where your HIPAA right of access gives you all the power of the federal government and the regs to say, I can get this from you. Patients can do what hospitals and insurers and everyone else can't. Yes, absolutely. And they can help other physicians by being part of the referral network because patients are in the middle. The second thing is most of the information that's released

is not in any coded form. It's documents and PDFs and XML files. This is the dark data part of it. This really is. It's hidden from machine readability. And a lot of the cancer information is in the pathology report, which doesn't come out of APIs yet. We've got a fingers crossed. So we need a long pole in the tent about a data refinery that takes this crude oil of documents and converts it to data. It's like turning a word document into Excel. In Excel, we know how to apply a feature or a function and operate on it.

You're basically taking unstructured to structured, but you're actually adding a step even before the instructor, which is making that data machine readable in the first place, especially in the case of a PDF. Correct. Then the final question is, so what? Now I have my structured data. What can I do with it? Well, you can certainly share it like you share a LinkedIn profile with your physician.

That physician, he or she might say, thank you, God, for bringing all this because I never get to see this sort of back. But beyond sharing it with people, you can also share it through an API, as Suzanne has said, to other app developers who can then say, look, citizen, you've done a great job producing this fuel, this computable. From your data refinery? Yes. But we'll take it from here and we'll run a calculation that might be a cardiac calculator, or we'll, we'll aggregate populations together and do a population survey. Or ideally, because we can automate clinical trial inclusion, exclusion, you could have a series of these algorithms running in the cloud all the time.

And every time a patient becomes a citizen in our parlance, a trial match can be detected on their behalf. That's the way we reduce friction, as Vijay said. So, you know, one of the things that I think gets sort of covered almost too fast is this unstructured to structured because structure could mean lots of different things. Really, like, if you can go all the way deep with ontologies and a true semantic structure, you really go from just words to understanding. And understanding is like the holy grail for machine learning and AI right now. It's a hard thing to do. And when you can finally use ontology and other things that people have driven, now the data actually really becomes useful.

And I think Excel is a very natural one. And the key thing is that it's not just even in a spreadsheet. It's in a spreadsheet where a computer actually knows what each of these things mean. It's right. And that's going to be key because a doctor does not have time to sort of go through pages of things to understand. He or she wants to be able to go very rapidly, just get a sense of the lay of land. And from that, say, you know, where we are and what we need to do. And I think, you know, we talk about friction.

There's friction in each of these levels. Friction at the patient getting the data. Friction at sort of what the data is and friction at the doctor side. I think the ideal is to reduce all three. So take me up a level then beyond the individual patient experience and talk about this at a structural health care system systemic level. What does this mean for insurers, for hospitals, for researchers, for clinics, drug testing. We're not to boil the ocean here, but how does this play into that? There's so many different ways that data can inform us.

So just one off the top of my head is that if you're a pharmaceutical company, you want to be able to understand how things are going for patients, maybe in a clinical trial, maybe in a baseline. And you need to be able to get a large number of patients that are the right. ones. So this is not necessarily 100 million people. This is maybe thousands of the right ones that can give you the days you need. And what I think we'll start to see is, especially as new statistical methods come online, that real world evidence will be very useful in some ways will be more useful than what you can do in a clinical trial, both due to the power and due to the fact of real life is different than a clinical trial. Right. It's like an in situ experience, in vivo experience,

but the real world is a laboratory instead of the actual laboratory. Yeah, definitely. And this is useful for pharma, but frankly, it's also useful for payers. in a world where things have gone past clinical trials, but really now the new barrier is not getting past the FDA, the new barrier is reimbursement. The payer will want to know with real-world evidence. Like, is this really helping. And this is something that you obviously can't answer without data

and you can't answer without the right type of data structure in the right way. Data is fuel for an ecosystem. And so when we talk about the structure of the health care system as it stands now, we talk about pharma, we talk about the payers, We talk about hospitals. But a big part of this are the patients who don't appear on anybody's organizational chart. Oh, that's such a good point. And yet they have so much insight to share.

And we need to make sure that we are pushing the power out to the edges of the network. That's where expertise lives that we don't even know about. We don't yet know what will really happen when we free the data and allow people to create the dashboards that they really need. We don't yet know how different patient groups are going to create something really useful, how an entrepreneur is going to look at this opportunity and say, I could create something that really helps people. And by the way, it could be a small group, but have a significant impact. Actually, the analogy that comes to mind from me when I hear, and that is I'm a big historian of the history of computing, tech, internet, and one of my

favorite themes is the idea of permissionless innovation. And what I love about this is what you're describing, because if you think about it, what happened with the internet, and permissionless innovation allowed people to build on top of the platform that is the internet, if you think of data as a platform and what people can build, you cannot predict the use case. There's second order and third order effects that nobody, the designer of a system can never predict up front. So what I love about this is this is permissionless innovation in a permissioned way where the permission is actually coming from the patient. Because essentially the patient is saying, you have my permission to move this data around, this portability. And then to your point,

who knows what that can unleash? And that is a really exciting thing. And it's also, it's really American. Like we're so. Yes, I mean. Permissionless American. It's totally true. We're such a country of rugged individualists.

For good or for ill, right? So a lot of health care depends on whether you have the wherewithal or whether somebody in your family has the wherewithal. And we as a country win when we make it easier for everyone to participate. So speaking of something very American in a sad way, because I do agree with you that permissionless innovation is incredibly American. It's why I'm. a capitalist. But a sad reality of American life today, especially something that we talk about a lot in health care, is the opioid crisis. And I, for one, would never, ever say something that can be solved through technology alone, because it's a social cultural problem. But in this context,

how would something like this play a role in a public health crisis like the opioid crisis? The opioid crisis is one place where data is the canary in the coal mine. It is a early warning system where if you digitize the data, you could literally have an Excel function. And I'm, you know, I'm trivializing it that is scanning for populations and where it sees a concentration of certain things happening. And population health folks have termed this syndromic surveillance. The CDC calls this the ability to look over large populations and see trends and patterns because of data. We started seeing signs of the opioid crisis in. death certificates. And it was actually public health researchers who started looking at data and

seeing, wow, that we're really seeing an uptick in this sort of death among young people. Death by addiction to opiates? Well, the problem was that it was unstructured data. So people started looking at the death certificates and started understanding what was happening. What I'm passionate about is learning lessons from the past so that in the future we can take the temperature of the country more accurately and more quickly to solve those problems. So how might we create a dashboard for the country so that we see something like the opioid crisis happening? Everybody played a role in the opioid crisis. The pharmaceutical companies played a role. Public health agencies played a role. Payers played a role. Why did payers play a role really quick? Oh, because of the way people were

being reimbursed or not reimbursed for pain management. All across the world, pain is managed using therapies other than drugs. So there's all kinds of ways. Here in the United States, we encourage the development of drugs and through a complicated history, which I won't get into, it became more the norm in the fashion to reimburse for drugs, to prescribe drugs for pain. Instead of alternative therapies like tens and various other things. Exactly. And so how does data enter into this? Data can tell us when a crisis is happening. It can show you also where we're actually gaining ground. We're seeing that we're gaining ground in Ohio against the opioid crisis. Data is telling that story.

So that tells me how it informs public health and people thinking about this. But how can something like what we're talking about where this dark data unveiling actually solve? I'm not, again, trying to advocate for a solutionistic view. It's a larger, bigger problem beyond technology. But how can it help? Yeah, I mean, there's different issues with the opioid crisis. One is that often the first intervention is opioids while you wait to see the back surgeon. or a musculoskeletal surgeon.

That's an accessibility issue. And then once that starts, then the problem is that the opioids are more accessible than the other solutions. And so now you have people sort of doing a doctor or opioid arbitrage between places. And so hopefully you can understand first why this is happening and then why it starts to sprout

and then what's facilitating it. And having the records in one place would do it, I think it's a little challenging because if it's driven by the patient, that patient's going to want to have to... Obscure. Yeah, to get past this. Yeah. So all of the records,

of these diseases we've been talking about so far, cancer, lupus, autoimmune disorders. I have a chronic condition, nothing to worry about, not to scare my listeners, but in the context of I have to see a doctor regularly, et cetera, these are all cases where you have multiple touchpoints in the system, and often longitudinal data helps. How would the longitudinal data and having a patient at the center now, what does it do? So longitudinal data is going to become ever more important as compared to episodic data, and that's because we're moving from acute care to chronic care. If we had a pillar or procedure, we could do something in a hospital. So that's an acute episode.

The fact that it's chronic, the condition is chronic, means we don't have a pillar procedure. It's going to take place over a long period of time. And people are increasingly mobile. So their health data portability problem is exacerbated as we go forward unless we address it now. There's been really some classic examples recently where the beauty of having times used data is that you're comparing you against your previous. yourself. Unless you have that times used data, all you can do is compare you against the population. And people are just so different in such high variances and overlapping distribution. The well-known example recently is Ben Stiller had prostate cancer. But his PSA level actually

never went high from a population standpoint. It just went high from his own baseline. That's actually a much stronger signal. Much, much stronger signal. And so that's why his doctors actually could tell he had prostate cancer, even though if he just didn't want off, there would be no reason to think that. Right. It's actually a lot like new moms and the pediatricians, is telling them, don't worry about the normed curve, track the kids' curve, because you just care about them growing and gaining weight. But otherwise, a new mom's lose sleep when their kid is like in the 25th percentile and so the 75 percentile, it goes to that same type of thinking. You just watch the deltas.

Right, exactly. The other thing that people can use their longitude in health history for is to participate in clinical trials. Most clinical trials upwards of 90 percent or worse don't get filled. And it's because there's no frictionless way for a clinical trial. Inclusion and Exclusion Criteria to detect a candidate patient. But if you have on the one side a digital health summary in cyberspace, on the other side you have a inclusion exclusion, let's say match made in heaven, and we think we can really move the needle for both pharma who wants to detect patients

and patients who want to be matched up with targeted clinical trials. Yeah, and the thing about the current state of coming up with new therapeutics, new drugs, the cost of clinical trials is really high, and too many things fail, and there's different reasons for failing. One reason for failing is actually not getting the right patient cohort and not designing the trial to run such that you would have the successful outcome at the end, and that's really a data problem. And the opportunity is actually, with that, is that more drugs could get through, and even certain things in principle could even be rescued. That would actually radically shape how these therapies get to market. I'm going to go even more radical and say that when patients all have

access to their data, they could form a coalition and ask for a certain clinical trial. That's beautiful. And actually, in a sense, there's sort of centralized versions of that with like, you know, the cystic fibrosis foundation with vertex and so on. But you're talking about it more like pop-up crowdsource. Precision. Yeah, something where you don't need the army there that you could ban together. The other thing with chronic care is that because it's such a longitudinal time horizon, an 80-year-old person, if they had their entire health record. There's no single institution that would keep the record on file for that long. So it turns out that patient portals and APIs give you a smaller window than an 80-year-old

history. Some of the information available is aged out over a couple of years. So you can't rely on a single interface to keep all your data. one because you're getting treatments every other place, and two, because they're not responsible for keeping it. And so it becomes incumbent on the patient to manage at least the longitudinal aggregation if not the interpretation of the data. So what I'm hearing overall is a theme here is that when you get the horizontal data that connects all the players in the healthcare system that can now communicate to each other through the patient at the center. And then there's a vertical piece, which is the history of the patient in the past, moving forward, et cetera.

Now, this gets us to the idea of big data, because now you have a lot of data to work with. We've been talking about the canary in the cold mine. We've been talking about all this stuff that people can do on top of this data. Honestly, it's a buzzword I hear all the time, like big data in health care. What's the big picture here on that front? So here's the dirty little secret about big data in health care. There isn't any. And the solved to big data in health care is small data.

That means what Susanna said earlier. The small data is at the edges. That's where patients live. So there's a moral, ethical, and technical imperative to work with and through the patient, and really for the patient. If you took the top few vendors for electronic health records and you reduce the entire U.S. population down to those vendors, those vendors in aggregate will hold about 6% of all data being generated digitally on you today.

Where's the other 94%? It's in the imaging systems. Of course, images are large, so there's a lot more data. It's in the microbiome. It's in the genome, whether it's the full genome or just a portion of it. It turns out that the electronic health record systems weren't built to handle all of this other explosion of data. And even if there's a consolidation of vendors in the marketplace, which we predict there will be. And I think that's going to be a good thing. The consolidation is going to be outstripped by, the runaway fragmentation in digital data. The only person who is legally, ethically, morally, incentive to pull it all together is the patient. When we think about the patient at the center of the circle and how all the big players are sharing data around the circle, and for them, the most useful share might be to a community of fellow patients. It might be to create a longitudinal record that they share with,

other people who have lupus, other people who have cystic fibrosis. Yes, this actually reminds me of a web satirium that I used to be obsessed with patients like me. I loved it because it's essentially like the long tail of the internet to find like many people suffering with whether it's dysfunctional uterine bleeding, which is a weird category or, you know, some thyroid, there's like a million things that you don't know. Eustachian tubes collapse. There's a million specific things. I love that aspect of people being able to center and create community around these things. And patients like me is a great example. It's like a time machine that you can travel backward and forward in your own record and in other people's record. And what's essential is that people are creating this small

data for themselves. And what is the opportunity is to create an industrial strength version of that. Yes, I agree. It makes me think about the Google doctor problem, the Googling problem where patients think they're doctors because they're informing each other and Googling things. And it actually creates more problems for a lot of doctors in that space. So when you say industrial production side, it to me means taking that data and putting it in a more rigorous system than one that's just so informal, foxy. I wouldn't say that patients like me is fulksonomy. I would say that they are a serious taxonomy of people's own tracking. And when I think about industrial strength health data, I think about the data that is currently being held by the clinical system that most people don't

have access to. And when people get access to that, where are they going to direct it and the choices that they have? For many people, the joke is they go to Dr. Google. What does 300 cholesterol mean, you know, and something like that? And I think the opportunity here is something much grander. If you finally have the data in a portable way, you can actually just ship this off and then get information from a real doctor who now has everything all in one place. I'm going to disagree. Because I'm really passionate. about looking at the expertise that patients really can have. So there are expert patients out there who really know their disease.

They really know their condition and can bring that expertise. And what I want to see is that everybody operates at the top of their license. And by the way, I think patients can operate at a pretty high level. And so, yes, there's a danger of Dr. Google. Yes, there's a danger of amateur pathology. There's a lot of amateur dermatology. But there's also the possibility of everybody being educated, everybody raising their game in health care, including patients and caregivers. What I would like to see is an ecosystem flourish around the possibility of access to industrial strength health data so that we can see we have no idea what's going to happen.

We have no idea what kind of engagement can happen around health data because we don't yet have access to it. again, we're leaving half the team on the bench by not giving patients access to their own data, much less access to each other, which I think is really going to unleash well-being. I love this idea because there's going to be two buckets. There's going to be the very acute, highly clinical bucket of information. And you want to license a professional to weigh in a quarterback. But I, as a migraine suffer, I have shared a recipe I have with whoever will listen to me in the moment. I do the same thing for allergies and things.

Exactly. I'd love to put that out there. Yeah, me too. Because it's a solve for me. And I have the protocol that I follow. And it's a chronic condition and the clinical establishment has said, you don't have a brain tumor, go home, you're fine. But I still haven't solved my migraines. So that's where peer-to-peer health care, which Suzanne has been promulgating for a long time.

It really, that's the outer circle that we don't know how that will be shaped. That's an ecosystem play that we're very excited to help. power. I'm actually hearing all three of you say the same thing in different ways, which is it's about empowerment. And if you put the P in power, and that should be the P and HIPAA, actually, not just portability and patients, but that's what we're talking about, is empowering the patient. Thank you guys for joining the A6 and C podcast. Yeah, thank you. Thank you.