Amazon’s Vin Gupta talks about the retail giant's entry into health care

Vin Gupta, MD, MPA, a pulmonologist and Air Force veteran, is the chief medical officer for Amazon Pharmacy. He was hired by Amazon in January of 2020, and is helping lead the company’s efforts to grow its presence in health care. Medical Economics sat down with Gupta at the HIMSS23 Conference in Chicago to discuss the latest trends in health care and what Amazon’s growing presence in the industry means for physicians.

Dr. Vin Gupta, ©Amazon

Medical Economics: Where is all the consolidation in health care going? Will we end up with a few big players dominating the market in the same way that happens in retail?

Vin Gupta, MD: I feel like there's a ton of opportunity in the market. I'm stepping outside of my Amazon role for a moment, but it's estimated for there to be $6 trillion in spending by 2028 in the US health care system, but access still is a struggle, we have one in three people that are underinsured, and there's a lot of opportunity. And that's where I feel very lucky to be a part of Amazon, because I think we're just one of many players, but we're really focused on access and engagement and convenience. And so, for me, it's not about consolidation. I think it's about solving problems that are rampant right now, because without newer entrants like ourselves and others, I think those problems are going to continue to exist.

ME: So you see Amazon as bringing a fresh perspective to health care?

VG: Absolutely. It really is just engagement through convenience. We think we can make it a lot easier to get your medications to help you stay at home. And if you're somebody who has chronic conditions and is on multiple medications, the simplicity of free at-home delivery and price transparency at the point of prescription, auto refills and pill packs and other features can help. The ways in which we experience health care today are to accommodate us as providers, not to engage patients. That's where I think we have an opportunity to add a fresh perspective to build engaging experiences, and we know how important that is to improving outcomes.

ME: That sounds great for the patient, but what does it mean for the doctor?

VG: Let’s take the example of coupons for pharmacy. I'm very sanguine about what that means for my fellow providers, that it makes the point of prescription far more useful for patient, but also for the provider. Now we know if there are concerns with the information, there is an automatic application of a coupon -- you don't have to search for it, there's no administrative burden. We can have that discussion at the point of prescription while I'm educating the patient on pros and cons of the medication about what they can expect to actually pay out of pocket if there was an out-of-pocket expense. Previously, that is not the standard experience for providers. Usually it's, “Hey, doc, at the retail pharmacy down the street, it's too expensive, and I didn't know that coupon existed.” A study in 2018 showed that 85% of those individuals never took advantage of the coupons available, so simplifying basic experiences, I think not only helps the patient, it dramatically helps the provider reduce administrative burden. It keeps patients happier, and hopefully it keeps them out of the clinic and at home healthy, and I think that also reduces workflow.

ME: A big theme at HIMSS this year is AI. What does all this big data and AI mean for the average physician? Where does the physician fit into all this?

VG: I am incredibly optimistic about some of the especially large language models that are being built that can help facilitate a differential diagnosis. We're seeing a lot of these applications now at the fingertips of clinicians of trainees, particularly for medical education, and enabling a young physician or young nurse who's going through training to have better resources at home, or when they're seeing patients for the first time. Right now, the way we test and train our trainees is you're right or you're wrong. On licensing exams on continuing medical education, we're now entering a world where we can actually help you understand decision-making in real time. Where was our failure point in the clinical decision-making process? Because of AI, we can move to get at the root cause. Oh, they got eight steps right in this diagnostic framework, they just missed step nine, so we'll just go deeper there. Right now, it's binary: you're right or you're wrong. So our ability to provide very targeted support is going to make better clinicians across the board. It's also going to help where it's nice to have an adjunct to say, did you think about this? Did you think about that? Ultimately, I hope it helps enable better care in the longer term.

ME: Burnout is a big issue for doctors everywhere. Do these technology solutions help? Doctors were promised that EHRs were going to help and they became a top contributor to burnout.

VG: I think it's too early to say, but I really think technology can help maximize potential. As a clinical adjunct, you're seeing the announcement of ChatGPT being embedded within Epic. I love that type of innovation, because it helps with transcription, it helps with workflows. I think thoughtful applications of AI, either to make the life of a clinician easier like that, or for decision-making support, whether it's for medical education, or for a stressed care provider in the inner city. There are ways in which this will directly tackle burnout and make the lives of clinicians easier and enable them to practice at the top of their license, which is vital. We have less MDS and DOs to care for a greater population, and by the end of this decade, one in five physicians is going to be over 65. And we will have less, in relative terms, care providers to care for them. So we need the top-of-license practice of medicine ASAP. I think AI can help enable that. But again, if we don't partner and listen to providers directly, it won't be as effective. We're doing a lot of listening and partnerships at Amazon pharmacy; it's part of why we have experienced double digit growth in the last four months, driven by partnerships. I love how open we are to those types of dialogues. And I think those at the forefront of AI have to do this as well.

ME: Administrative burden plays into burnout as well. It sounds like AI can help with the paperwork monster that doctors have to deal with on a daily basis.

VG: I think you're already seeing some signs that AI can help with streamlining prior authorization, which is one of the worst experience experiences for clinicians and ultimately then downstream for patients. Across the board, I know multiple stakeholders are thinking about that problem, just like we are. And so over the course of time, I suspect that we're going to see durable solutions where AI can solve these issues. AI is part of the solution. But frankly, I think just going back to the basics is important, and what we do really well at Amazon pharmacy today. I think what we're trying to consolidate and scale over the course of 2023 are offerings like coupons, price transparency, affordability, value selection, delivering your stuff when you expect it to the home, doing the basics that we know well from our retail roots, and applying it to pharmacy. If we can get that right at scale, I'm excited about what that means for patients and what that will do to reduce administrative burden. I think just what we’re doing with coupons for pharmacy takes out the administrative hassle from what should be a really simple experience. And then in parallel, these innovations in AI will be helpful. But we haven't even gotten the basics right at scale. And I think that’s where we can derive a unique value from our tech-forward approach coupled with our core strengths and logistics.

ME: You mentioned prior authorizations. Do you think we can get to a point where we're getting real-time approvals?

VG: Sort of stepping into my role as a clinician, and from what I've seen from the broader ecosystem on this is totally what we're doing as a pharmacy today, I think that's possible. I do think that we're getting to the point where we can be smart enough from a technological software standpoint where it takes ICD 10 codes, it takes some understanding what's been trialed in the past, and combining that information in an algorithm and saying this person does actually qualify for the next step-up in therapy. It's not difficult to imagine that getting automated; it's something that we're thinking about, and we know what a lot of our peers are thinking about. I suspect over time, that's going to be something that gets addressed.

ME: What do technology companies need to do better in the health care sector?

VG: I think what we're building is going to build trust with our patients because we can deliver on things that they expect. And yet there's a wave of digital innovation for other health services that's meeting this moment where messaging and health information is, and there isn't a lot of trust because of what we've seen over the last three years. And so just the conversions of that to me, not from an Amazonian standpoint but more from just being a clinician, I always wonder how patients are going to think about all these amazing tools that they have at a moment in time where I think trust and health messaging has been injured. How do we ensure trust in things that work? We have to recognize that we're starting from a place that maybe we weren't at back in 2019. How do we make sure we don't forget that we have to build and we have to earn back a lot of patients’ trust, even with these great tools that we have.