In this Leadership Insights video interview, Mike Vizard speaks with Shubh Sinha, the co-founder and CEO of Integral, about how to address privacy concerns in the age of digital health care.
Mike Vizard: Hello, and welcome to the latest edition of the Digital CxO Leadership Insights series. I’m your host Mike Vizard. Today we’re with Shubh Sinha, who’s CEO of Integral, and we’re gonna be talking about how to address privacy concerns in the age of digital health care. Shubh, welcome to the show.
Shubh Sinha: Thank you for having me.
Mike Vizard: I think this issue is somewhat overlooked. But I would argue that one of the reasons that health care hasn’t been transforming as much as it might is because there are all these concerns about data privacy, and how we manage all that stuff. So what is the chicken? And what is the egg here in terms of how we’re encountering these issues, and what can be done about them?
Shubh Sinha: For sure. So I think there has always been historically a tension between the business or functional groups of health care companies and their compliance counterparts. And for good reason too; health care information is governed by the United States regulation to the highest severity, I would say, but also just kind of intuitively speaking, like someone’s health conditions gives you the most intimate context into them, and the problems that they might be experiencing. And so treating that information with care has always been the utmost priority for companies. And so that has traditionally led to taking conservative approaches, not leveraging data as effectively; as you know, other industries do – so like e-commerce, for example. And then also just like slow speed. And so the kind of like the chicken and the egg here is that tension presenting itself in the form of excessive care costs, because there’s not enough transparency, because there’s not enough data collaboration, treatments that are not as effective as they could be, because they weren’t developed with the most data driven approach that they should have been, and then a myriad of other use cases. And so that’s why integral exists today to be able to solve that tension. More on that.
Mike Vizard: So is this at the root cause of why everybody goes to the doctor these days, and somebody hands them a tablet, and they fill out what appears to be a digital version of a paper form. And then I come back the next day, and I fill out the exact same form and the same thing over and over again, because we’re overcompensating for the privacy concerns. But as a result, we’re kind of creating a crappy experience.
Shubh Sinha: Yeah, yeah, I think that’s well put. And to drive that home, that point even further, I would say that, you know, in your health care journey, say you watch a particular commercial for a drug ad, or you come across a drug ad online, you go to your doctor to find out about it, they conduct a series of tests, they prescribe you the drug at a pharmacy, you go fill the prescription, your insurance gets involved to make sure you’re approved for everything, each of those points that I mentioned in that journey – so that’s a pretty naïve journey. So five different touch points. It’s all owned by different entities. And so you filling out your form at your doctor continuously. And every time you go to a new doctor, that happens, because there’s no communication between those touch points I just mentioned, which leads to like a very lack of a holistic picture of the patient experience. And so as a result, you read very high friction parts of the patient experience over and over again, aka filling out forms multiple times, even with the same information, because no one is facilitating that information transfer on the back end. Because everybody’s concerned if there’s a drop in information or the compliance is just kind of messy; they would just rather not share the information than worry about the compliance complications.
Mike Vizard: While it’s inconvenient, it’s also a more serious issue, because a lot of patients, especially the older ones, are taking multiple medications. And nobody knows exactly how these things might interact. So people will get a prescription for something and come back three days later and start complaining about this, that and the other. So part of the problem here seems to be that we can’t get out of our own way.
Shubh Sinha: I would definitely say that’s pretty spot on as well. We are a B2B company. We’re a data infrastructure company that helps companies facilitate the safe exchange and analysis of health care data. But oftentimes, what gets missed is that the analysis and exchange of health care data eventually has the downstream effect of letting a doctor or a particular pharmacy, no, hey, this is a double prescription, or this medication is directly going to conflict with this other medication that is present in this data source. And so I think what you mentioned were like these multiple medications and whatnot, the most kind of benign effect of being prescribed a medication, not knowing all the other histories and like all the other parts of someone’s life is that there’s some of the mild annoyance, but there’s so many instances of medications colliding with each other because prescriptions just weren’t known because there wasn’t enough information exchanged such that there was a detrimental effect on a person’s health. And so that’s why the kind of like the end-to-end value chain of, of why we started, what we did, because if you can facilitate the right amount of information transfer, you can also make a very direct impact on the everyday consumer, such as you and me.
Mike Vizard: Short of changing the laws, which would take a long time, but it might be a good idea – what can be done today in the context of what we have for regulations? And what are some of the things that health care organizations could do but aren’t doing?
Shubh Sinha: Certainly. So just for a bit of context into how I came to a perspective I’ll voice shortly, I was at LiveRamp, a public ad tech company, specifically in their health care product line, helping large enterprises both in the pharma insurance and also the digital health space, leverage healthcare data amongst a variety of other data sets. So demographics, wearables, etc, leverage this data holistically to have a 360 view of patients, such that more informed ad campaigns could happen, such that better drug development could happen, and so I got a really intimate perspective on the use of data and what exists today – where it can be – and then what happens today to the patient, and how the patient’s lives can be much, much more positively affected. And so what exists today is that, because of COVID, there was just a massive digital push to be leveraging data a lot more. And the government went a little bit lax on some of the regulations just temporarily as kind of like a COVID wartime solution. And so companies like Abbvie, Pfizer, all these like massive, massive organizations, started taking wearables data, demographic data, geographic data, as well as your actual health care data, like your prescriptions, your claims, all that and then combining it to get this 360 view of a patient. Problem is the compliance process today takes months and costs a lot because it’s all done by consultants. And so what’s possible today to answer that question is HIPAA consultants can be brought on, they can be on boarded to a series of projects; it’s a pretty bloated process where they’ll look at the data, they’ll go back and forth with the customer. And then they’ll finally sign off on this long PDF report that says, yes, you can do what you need to do with your data. So analyze it, share it, whatever you need to do. And then companies like Pfizer are free to analyze data just as any other company would. And so that is what’s possible today. The problem is when it takes months, and it costs so much to do it, these companies have to, one, pick and choose which analytics use cases they can tackle, which is a pretty scary thought considering you should be able to tackle every single analytics goal when it comes to somebody’s health profile. And then, two, it just takes so long that if you can’t find out what your data means in nine weeks, that means you have to wait at least nine weeks before some sort of decision can be made, to measure the result of that. Then it goes into effect. And then it finally ends up impacting the consumer, say like half a year later – like that amount of lag time is just too slow for an enterprise, and it’s certainly too slow for a person. And so where we believe it can be is with Integral, where we come in and we automate the health care data compliance required to integrate such sensitive data types, hence Integral, the name. And then we make it such that companies can leverage the data and ours, such that important decisions can be made in the span of days, and they can be measured quickly. And then they can actually be implemented to where the consumer, the everyday consumer, sees the effect in a matter of weeks, which is pretty fast in the health care industry.
Mike Vizard: If I have to wait nine weeks, they’re just going to make me take all the tests over again anyway. But, to your point, if we don’t find some way to automate this process, is health care by definition somewhat going to be uneven? Because right now, the only organizations that can afford to drive this are say, you know, somebody like a mass gen, but, you know, the average folks living out in some rural county are not going to see the same level of data driven analysis to drive their health care.
Shubh Sinha: Certainly, and I think this goes to kind of our long term vision where today only big companies can can truly afford the timelines as well as the costs. And so think of the massive pharma companies, and some of the bigger digital health companies as well. But health care information sharing should be democratized because the doctor you go to or the wearables company that analyzes your data and gives you recommendations for your diet – all these kind of small organizations, they shouldn’t be treated any more or any less differently than any other big company because health care information needs to be democratized so that the patient can actually see the benefit. And so, right now, we’re starting with the big companies because they are the ones that can afford it. But all-in-all, health care will stay uneven. I think, as you mentioned, if this automation can’t scale to the entire industry, and as a result, if information can’t flow freely. And so that’s where our plan is to like scale like wildfire in the industry where we’re easily deployable, we’re easily integrated, such that we can work with the small company X, the big company, and the medium companies to be able to facilitate the safe transfer of information, and thereby becoming that infrastructure to power, like a safe information highway, so to speak. And so until we can do that, or I guess until somebody else can do that, there’s not going to be this even disbursement of information, which leads to disproportionate outcomes, as you mentioned. So rural populations for certain diseases not getting the same attention when it comes to cures as other diseases. And in the worst case, like certain health outcomes being massively, massively improved. But other health outcomes being completely ignored.
Mike Vizard: Do you think also, we might be able to find the root cause of a lot of illnesses? Because we hear things, you know, there are clusters of these types of cases in specific geographies are among specific types of people. But today, we don’t seem to be able to get at that kind of analysis.
Shubh Sinha: Yeah, yeah. And I talked to a couple of companies who are in this space where they want to do predictive analytics, because if you can predict the next flu season, if you can predict, even like, to an extreme degree, if you can predict, like, predict, like the next COVID variant – that lets you be more proactive in protecting a population. It lets you be more proactive in understanding of population. And so the challenge seems to be right now, one, like kind of taking all the data types, and then making sure you can actually do the thing you want to do. So the compliance piece of it. And then the fact that that takes so long, the actual implementation of a decision, like once you find out some sort of insight, the next logical step is to make some sort of business decision. And that is just not measurable right now, because once you make a decision, you gather more data. And so then you bring it back to the original data set that you were analyzing, and it’s another set of weeks and compliance or months and compliance. And so the problem here with compliance – it’s not this one time, kind of expensive cost. It’s continuous, expensive cost that needs to be, frankly, doesn’t need to be there because patient privacy should be the utmost priority. But it doesn’t need to be synonymous with slow or blocking progress. And so yeah, today, you have like, a bunch of kind of cycles of iteration and measurement. They’re just all slowed down by how slow compliance is today. And so if we can speed that up, we can enable data scientists or like software engineers to do what they do best, which is iterate very quickly using technology at their disposal.
Mike Vizard: Of course, you cannot walk down the street these days without somebody leaping out to tell you about their great new AI thing. So we have all this healthcare data. So how will AI be applied to this data in a way that might be interesting and compelling? And hopefully not as scary as people might imagine?
Shubh Sinha: Yeah, certainly, I mean, I’m just as excited about AI as anyone else, and I find its ability to supplement human interaction or the human brain, so to speak, to be extremely fascinating. And so in health care, however, I do think the integration or the application, however you want to think about that, that is the part that needs to be careful. And specifically what I mean by that: AI is generally – its fundamental value comes from the datasets it’s trained on. And the health care industry is not sharing their very valuable, very highly sensitive data assets with just some random AI model, right? And so what happens, at least in the conversations I’ve been having is, there’s some amount of conservatism, deservedly so, to sharing data with like an open AI model or something like that. And so they would rather just not do it than spend years trying to see if they can come up with something or internally contract a specific firm to build something specifically for them. And so this leads to lag time, expenses, etc. Where we see this going, and we just tackled a use case like this actually, in us being the data infrastructure, we would rather plug in with the AI such that when data goes through integrals platform, it comes out pristine, compliant and ready to share. Instead of sharing it with another company, we would share it with a model, or we would share it with whatever the customer wanted in terms of like the AI model they want to use. And so where we see it going is – use the progress of the AI industry. Don’t recreate the wheel by any means; like leverage that by making the data scrubbed and safe, rather than like making the AI scrubbed and safe, so to speak. And so that’s why we’re particularly excited because we’ve already seen use cases that once we clean the data and make it shareable, the health care companies are not worried about releasing a sensitive data asset because it’s sufficiently masked, and they feel comfortable releasing it, and they get all the benefits that an e-commerce company would by sharing their data assets with an AI. They get the insights, they get the quick creativity, all that and more.
Mike Vizard: Sure, thank you for sharing your insights. That was pretty great. I’m gonna remember the phrase random AI model for a while and see how that goes. But hopefully we’ll talk to you again soon.
Shubh Sinha: Yeah, that’s great. Thank you.
Mike Vizard: All right. And thank you for all watching the latest edition of the Digital CxO Leadership Insights series. You can find this episode and others on the digitalcxo.com website. We’ll also have the transcript there and until then, we’ll see you all next time.