Latest / StartUp Health NOW Podcast / Gray Oncology Solutions Is Building a New Operating System for Healthcare Workflows

Transcript

Gray Oncology Solutions Is Building a New Operating System for Healthcare Workflows · StartUp Health NOW Podcast

• 0:00
  [on hold music] Welcome back to Startup Health Now, the podcast where we celebrate the entrepreneurs and innovators shaping the future of health. I'm Logan Plaster.
• 0:18
  Before we get into this week's episode, just a word about Startup Health's Moonshot Impact Fund.
• 0:23
  For accredited investors, this fund makes it easy to gain exposure to a diversified portfolio of private health innovation companies.
• 0:30
  You can find all the details, as well as sign up for an upcoming informational webinar, at healthmoonshot.com. Now, on to today's show.
• 0:39
  The idea of health system workflow sounds a bit obtuse, like under-the-hood administrative territory.
• 0:46
  However, we've all experienced the frustration of having to wait for an appointment because a clinic was overbooked or an X-ray lab was backed up.
• 0:54
  Inefficiencies in healthcare cost time and money and frustration all over healthcare. But in oncology, dealing with cancer patients, the workflows are much more complex, and the stakes are even higher.
• 1:07
  That is why my guest today, Andre Diamant, co-founded Gray Oncology Solutions, which joined Startup Health this year.
• 1:14
  Andre and his co-founder are building what they call an operating system for health, one that uses machine learning to line up and map out healthcare resources so patients know what to expect, and they get the care they need on time.
• 1:28
  This interview is excerpted from our Health Moonshot Update series, which appears on YouTube, so you can watch the whole thing at youtube.com/startuphealth. Now, on to the interview.
• 1:38
  [upbeat music] Andre, thanks for joining me today. Thank you. It's a pleasure.
• 1:49
  So, um, I wanna just start by putting it on the table exactly what you're building, so that we can spend our time kind of unpacking, uh, the product, the platform, and the problem you're addressing. So what is Gray?
• 2:01
  What did you build? So we're a healthcare start-- software startup based in Montreal up in Canada, and we're addressing the logistical inefficiencies all across the oncology workflow.
• 2:12
  And, and what this translates to in terms of an actual product is, uh, a platform we call Gray OS.
• 2:17
  We sort of think of this as the operating system for a cancer center, and it automates and optimizes the patient scheduling process across the various silos of, of a patient's cancer treatment, so chemotherapy, radiation therapy, and soon surgery.
• 2:32
  You said logistical inefficiencies. Paint me a picture of what some of those efficiencies look like. I'm sure people who are viewing this may never have been stuck in that quagmire of the hospital workflow. Right.
• 2:44
  So, so oncology treatment, uh, of course, is very complex. There's a number of reasons for that. Um, there's a lot of variation in each patient, so it's not like every patient is gonna take twenty minutes.
• 2:55
  It's not one size fits all. Similarly, every patient may require multiple appointments.
• 3:00
  In a dentist's appointment, you kind of go to a single appointment once every six months, while in cancer you might require three appointments, or you might require thirty appointments, and each of these appointments has varying needs in terms of resources.
• 3:13
  They have varying needs of where they need to fit on the schedule in comparison to other treatments. So let's say you have chemo and surgery at the same time. Maybe you need surgery prior to the chemotherapy, et cetera.
• 3:23
  And what, what this amounts to is a schedule that has many, many dimensions and many, many permutations of possible areas that the patients could fit into the schedule.
• 3:33
  And the number of permutations has so many zeros that it's impossible for a human to, to actually do this in an optimal fashion without tearing their hair out every day.
• 3:43
  So it's sort of-- we think of it l-- you're kind of playing a very, very complicated multi-dimensional game of Tetris at all times. Mm.
• 3:49
  Um, and, and that's just an area that mathematics and artificial intelligence can really help out with.
• 3:55
  Is that why we haven't had this, uh, addressed in the, in the past because we-we've tried to be too manual about it, and you're bringing this, this AI element to it?
• 4:03
  S-so there's a number of hurdles that, that ha-have held us back in the past, certainly now as well. I mean, these are hurdles that we face every day.
• 4:09
  Um, the, the level of mathematics required and operations research is certainly one of them.
• 4:14
  There's also interoperability and making sure that all of the data in a center can actually be used together to, to make a platform such as this and facilitate it.
• 4:24
  And, uh, it's really been over the past decade or so that this has started to accelerate, where you s-- you're seeing data more and more consolidated, and the ability to actually build something like this is, is becoming more p-possible, which certainly wasn't the case fifteen years ago.
• 4:38
  And, and there is still work to be done in that area. I'm sure. That's why, that's why you're doing the work here. When did you launch, uh, Gray Oncology Solutions?
• 4:47
  So, uh, we really began operations, uh, about a year and a half ago, so it was about January twenty twenty, right before COVID. Mm-hmm.
• 4:53
  Um, and that was when we really dug our feet into this idea of patient scheduling and really dealing with this game of Tetris and how can we increase efficiency of these centers, reduce patient wait times, um, and all the other clinical sort of benefits that come from that.
• 5:10
  Uh, COVID obviously hit right a few months after we, we really dug into this idea. Um, initially we were kind of terrified. Uh, COVID obviously shut down bigger companies than ours.
• 5:21
  We figured, hey, this is gonna, this is gonna shut our doors.
• 5:24
  But instead, the-- we, we found the, the conversations that we were having with hospitals accelerated, and they sort of pulled us closer rather than pushing us away.
• 5:33
  A-and that really just reaffirmed the idea that, hey, this is a problem that needs to be solved, e-especially in the face of the pandemic.
• 5:39
  Uh, besides the idea that there just is a problem that needs to be solved, w-what do you attribute to the fact that they really turned to you in that, in that moment?
• 5:47
  'Cause like you said, there were companies that shut down. There were, um, clients that stopped returning phone calls.
• 5:54
  So one of the big areas, uh, especially in Quebec, the, the province that, that we're from, that was happening, and I've seen this happen all across the world, certainly in America as well, uh, was the concept of, of screenings for cancer were being stopped.And when the screening stops, that doesn't mean the cancer is stopping.
• 6:14
  Right. Um, and you saw a lot of physician concern, especially in Quebec and I-I'm sure elsewhere, where they were worried about what the influx of patients with cancer who weren't being diagnosed was going to look like.
• 6:27
  So at that point, obviously we were optimistic back in March of twenty twenty where we thought, "Hey, maybe in two months we'll be back to normal." That wasn't the case.
• 6:33
  But they were wondering in August, are we gonna have a sudden wave of stage three, stage four breast cancer patients, and what are we gonna do about it? How are we gonna handle this backlog?
• 6:43
  Um, and they had that sort of at the forefront of their minds, so then the conversations we were having with them very quickly slanted in that direction.
• 6:51
  And that was actually kind of one of the first modules of this operating system that we built was a simulator specifically to evaluate this impact that this backlog would have, and then subsequently how can we address it.
• 7:02
  Got it.
• 7:03
  You know, I think of when a lot of people hear, uh, about inefficiencies in the logistics of a health system and, uh, workflow backups, they think about this in terms of, you know, frustrations and, um, uh, inconveniences.
• 7:18
  Uh, talk to me a little bit about sort of the real life and death and health, uh, issues that arise from inefficiencies in this kind of system.
• 7:27
  So there's a, there's a handful of issues both on the provider side and on the patient side.
• 7:31
  Uh, on the most important patient side, um, often delays cause just a significant amount of anxiety, and anxiety is tied directly to clinical outcomes.
• 7:42
  Um, it's very difficult to get better, especially with something as severe as cancer, when you're s- you're already incredibly stressed. You don't need more stress.
• 7:49
  And patients are often in the dark about when a- when are their appointments, are they gonna be delayed, how long are they gonna be delayed.
• 7:57
  Very often you'll see full waiting rooms, um, where i- you know, you have an appointment scheduled at two and it turns out that, hey, maybe it's at two thirty. Who really knows when it is? Yeah.
• 8:06
  And that's just a level of, of stress that we don't need to be putting cancer patients under. On the provider side, the, the capital that's being used to treat a cancer patient is very expensive.
• 8:17
  It's expensive resources. You need highly trained professionals, expensive equipment, and really every, every ten minutes that that equipment isn't being used is-- could be better spent. Interesting.
• 8:30
  We, we could be treating a patient. There are cancer patients out there who need the treatment. Mm-hmm. And
• 8:36
  s- a-at least a part of that is the fact that we can be assigning these resources to patients in a more efficient way. Yeah.
• 8:44
  Uh, this is probably an in-inadequate metaphor, but it makes me think about the way that, uh, Uber and Lyft changed the way that we thought about waiting for a cab.
• 8:53
  If you know that it's five blocks away, uh, and you can see it tracking towards you, your, your heart rate goes down and you're like, "Okay, I'm going to get picked up today. Things are under control."
• 9:04
  Just like the knowledge of kinda how the system is moving and where your place is in the middle of it obviously is a- Exactly.
• 9:10
  And i-i-it is-- it's a stretch of a metaphor, but it does apply and in some ways it's an apt comparison because we are intending to do something similar eventually. That hasn't been developed yet, but the- Okay...
• 9:22
  this idea of being able to show you what is the traffic like in the clinic at that given day. If, if you have your appointment booked at two PM and we already know by noon that you're gonna be backed up.
• 9:35
  We're backed up, whatever. A nurse had to go home, a machine broke down, who knows?
• 9:39
  We can tell you that right away, so you can sort of see and have the peace of mind that, yes, your treatment will be delayed about thirty minutes, but you can stay at home, you can stay at work, you can go for a walk, whatever.
• 9:48
  You don't have to go sit in a waiting room staring at a wall and hoping your name is called. Um, so, so that's an area that we're looking to get into.
• 9:57
  We haven't yet, but it's, it's a big part of the vision of ultimately what we want to bring to, to sort of the healthcare ecosystem.
• 10:05
  Uh, you said that hospital systems, instead of stopping returning your calls, kind of drew you closer.
• 10:10
  Talk to me about kind of where your traction is now, kind of what you're excited about for the summer and the fall of twenty twenty-one. Right. So this was mostly, uh, in Canada.
• 10:22
  Uh, initially all of our conversations were in Canada, of course, coming from Canada and Quebec. Um, the majority of the connections we had and the folks that we knew in hospitals, um, were within the city of Montreal.
• 10:33
  So that's where our, our initial hospital centers have been.
• 10:36
  Um, we're currently deployed in one center there, deploying in a second, and then have a third on the way, and hopefully a fourth out in On- in Ontario, sort of as we start to break out of Quebec.
• 10:46
  Um, in terms of the, the coming year especially, we're, we're really looking to break into other provinces in the States as well in terms of securing a, a pilot that, that is concerned about these same operational efficiency issues and wants to work with us to really provide this platform and, and be on the cutting edge of how we can use operations research and AI to improve patient care.
• 11:09
  [lips smack] That's great. Uh, it's my understanding that through Startup Health you met the folks behind a company called Careteam, a patient engagement, uh, software company. I think they're in British Columbia.
• 11:21
  Um, and I wonder if you could speak to-- you don't have to sort of, uh, you know, go into that, um, relationship or, uh, you know, if even if there is one, but just I wonder if you could speak to the importance of, um, just, just collaboration between health innovation companies given the complexity of the system, sharing knowledge and just, uh, taking this, this understanding that, look, we need to do this together.
• 11:46
  None of us are sort of big, big enough or smart enough to do the whole thing. Absolutely.
• 11:51
  So, so that i-is sort of a nice segue from our previous, uh, the previous metaphor you gave about Uber and then the description I gave about this, this idea of putting this operational information directly into the patient's hand.
• 12:02
  Uh, Careteam has a great patient engagement platform. There are many other companies with great patient engagement platforms, and frankly, the, the last thing a-- the last thing patients need is yet another app.
• 12:11
  So relationships such as with Careteam or with other patient engagement platforms can beTremendously efficient, really, in terms of getting this value into the patient's hands. So
• 12:25
  in essence, we have this operational tool, and we can transmit that information to a patient engagement platform that maybe their patients are already using. Got it. And then it's kind of just a win-win for everybody.
• 12:35
  It's easy for the patient. They don't have to install another app, but they're getting that extra information that we can provide. So building a patient engagement platform is incredibly expensive as well.
• 12:46
  [chuckles] Yeah. So it just doesn't make sense for us to, uh, put a ton of capital, a ton of time and energy into building this when something already exists out there, and we-- collaboration can accelerate that process.
• 12:56
  Got it. So the Gray operating system layers over patient engagement software. It layers over, uh, an EHR, like an Epic or a Cerner, right? Ex-exactly.
• 13:06
  And the EHR, uh, point is, is a very good one, where we aren't an EHR. We're not intending to be an EHR, and we're not intending ever to be an EHR. We explicitly layer over it.
• 13:16
  I think of the EHR as a hard drive, and we're the operating system. We're managing resources with that hard drive, but we're not actually storing the data. Got it.
• 13:25
  You know, just to shift gears for a minute here, um, I believe you came straight out of grad school, started this company. Is-- This is your first company, correct? This is, yes. Okay.
• 13:35
  And there actually was a bit of an overlap, uh- [laughs]... between ending grad school and starting the company, which I would not recommend. Yeah.
• 13:40
  So if I take a second PhD and start a second company, I'll make sure to spread them out. Little bit of a gap, plus you have small children at home. Uh, I, I know a new baby, uh, there at the house.
• 13:52
  And so I just-- I wonder if you could speak to the experience of, of being a new CEO. Um, what surprised you about that experience? Um, I'm sure COVID kind of shaped that in some way, but has it been what you expected?
• 14:08
  Um, maybe tell me what you've learned in the last year of, of starting a business. I, I think, I think one of the biggest ones was I probably underestimated just how much emotional bandwidth it takes up at all times.
• 14:22
  Yeah. And I don't mean that in a negative way, but I mean in the sense of it's not the sort of thing to, to, to take it to the other extreme. It's not the type of thing that you can come home at four or five- Yeah...
• 14:32
  and it's cut off. And, you know, you go into work the next day, you start thinking about work. You come home, you're not thinking about work. It's really just always on your mind, very similar to a child.
• 14:40
  [chuckles] So, so in that sense, I kind of have three children. Yep. Um, and, and they're all kind of competing for space in the head, uh, and hopefully I'm doing a good job at it.
• 14:48
  [laughs] Are, are there ways- I take it a day at a time. I know. Uh, you know, I have two children, and so there's days when you do better than other days.
• 14:55
  I mean, are there, are there strategies that you've found to help you that could be, um, instructive to another CEO in a similar position? I, I think, um, yeah, a few.
• 15:06
  Maybe, maybe first is the concept of taking things a day, a day at a time. Don't try to think too far ahead. It's, it's just gonna overwhelm you.
• 15:13
  Um, second, I do always try to explicitly make time for, for the actual family. Mm. Uh, no disrespect to the company as a child, but it's not a real child.
• 15:22
  Um, so making sure to have some protected time, especially on weekends, where, where I am just spending it with them. Um, and then finally, routine.
• 15:30
  Even, even before children, I've always been someone who's very, very focused on routine, and it really helps me get through the day.
• 15:36
  Um, and, and I was just thinking back is how now my wife and I and my children, we have a very specific routine.
• 15:41
  It's like in terms of when I'm dropping the older one off at daycare, when I'm doing this, when I'm doing that, when I'm preparing dinner, it's always kind of the same. [chuckles] Mm.
• 15:50
  And obviously, you know, the plans sometimes change, but at least having sort of this equilibrium state I find really helps me. You're try-- You're reducing your decision load, kinda like- Exactly. That's, that's-...
• 16:00
  the way Steve Jobs would always wear the same thing to work, right? [chuckles] That's probably the literal, the, the literal way of putting it if you were to speak to a, a professional in these matters.
• 16:08
  [laughs] Um, awesome. Well, uh, that's about the time that we have. Uh, last question is, um, what's an ideal partner? Um, who are you looking to collaborate with? Big, small, Canada, US, global.
• 16:24
  Um, who would you like to partner with? So definitely on the American side, um, because of, of course, the ecosystem there is very, very different and very, very similar at the same time.
• 16:34
  It's still similar machines that are treating similar patients. We're all human, after all.
• 16:39
  Um, but of course, I have to acknowledge just how different the healthcare system is there in terms of the provider-payer relationship, et cetera.
• 16:46
  Um, so that's an area that we-we're really looking to, to validate, both in terms of clinical impact, but also of operational impact directly to the healthcare providers. So, so that ideal partner really is a, a
• 17:00
  relatively large-- When I say relatively large, I mean not necessarily a single radiation therapy bunker, but rather a healthcare system, something like a hospital that has chemotherapy units, so an infusion center, radiation therapy machines, and surgery.
• 17:16
  And the fact that there are multiple patients who will be treated at all three of those kind of as a one-stop shop rather than bouncing around multiple centers.
• 17:25
  Of course, we have a strong academic background, um, coming. All of us come from PhDs, and then we have a handful of founding advisors from, from universities across Quebec.
• 17:34
  And because of that, a university-affiliated center i-is a draw to us, one that is interested in, in not just evaluating this from a purely operational and commercial standpoint, but also evaluating this from, from a research perspective and, and looking into how this is affecting survival outcomes and, and
• 17:53
  similar metrics like that, which, which really require academic expertise as well. That makes sense. I lied. This is the actual, the actual last question. No problem, no problem.
• 18:02
  Uh, which is, um, you're focusing in on oncology. You know, you said before that cancer is a uniquely complicated workflow, uh, and so it needs, it needs this added level of intelligence.
• 18:16
  But I can't help but think about a whole health system.
• 18:20
  Uh, you know, a health system might have, you know, twenty-five hospitals, five hundred points of care, and the need for an incredibly intelligent system guiding it all.
• 18:30
  So when you think out longer term, you think of, you know, kind of more health moonshot sort of vision, um, is that where you think theWell, I guess question number one is, is that where you want Gray to go more broadly, um, or do you see the healthcare system going that direction?
• 18:49
  Are you optimistic that we're sort of moving in that direction? Certainly. I'm-- I, I am optimistic that we are moving that direction.
• 18:55
  And, and absolutely in terms of where Gray is going, breaking out of oncology as well, sort of vertically within a hospital and looking at this as an operating system, not just for a cancer center, but for healthcare itself.
• 19:08
  And the, the, the idea of we think of, okay, cancer treatment is complicated, and that's why we started there, and it is. But a patient's trajectory of their health itself is also very complicated.
• 19:19
  And that is an area that if you start really breaking down the walls across the various treatments that they might be receiving for any ailment they have, you can really start to have this holistic view and optimize each piece together.
• 19:32
  And, and then, I mean, the sky's the limit when you talk, you mentioned multiple hospitals within one health network.
• 19:38
  The idea of having one sort of larger body that this, of course, in Canada with socialized healthcare would mean a province that is able to look at all of its different resources across various sites or cities and determine, okay, maybe the hospital on the island is being really flooded right now in terms of, um, just how many patients are there.
• 19:58
  We can relocate some of the patients over to this other center. And you can start bringing geography into the picture of, oh, great, I mean, this patient lives right beside this hospital anyway.
• 20:07
  Why are they going all the way over there when we can move them here?
• 20:10
  And, and there's just a whole lot of efficiency gains and, and gains for the patient as well in terms of their satisfaction and just their quality of life that you can really start to bring in to healthcare once, once you start looking at
• 20:23
  this whole ecosystem as kind of one big piece as opposed to individual silos. Uh, listening to you just makes me think of the just-in-time logistics that we seem to have mastered in, like, delivery, like Amazon. Yeah.
• 20:37
  Uh, you know, getting the right place, you know, the right product to the right place at the right time, and yet, uh, we have this big gap in healthcare where we can't seem to optimize the resources.
• 20:49
  So there, there, there's a vision for the future there. There is. And there's a, there's a handful of very good reasons as to why healthcare isn't there yet. Yeah. I don't think it's unachievable, obviously.
• 20:59
  Otherwise, we wouldn't be doing this. Yeah. But, um, I mean, just the, the most obvious ones are, are one, data, which we already touched on.
• 21:05
  Um, data security is paramount, and it-it's obviously a challenge in itself to, to overcome those hurdles and how can we, how can we bring this efficiency into the world while still respecting patients' privacy.
• 21:17
  And then second, just the unpredictability of patients. Uh, packages are, are [laughs] a whole lot easier to- Yeah...
• 21:22
  to when you understand how traffic is gonna operate, to know that it'll take twenty-two and a half minutes to drive from here to there, that sort of thing. Yeah. While patients are much, much more complicated.
• 21:32
  But- Yeah... that doesn't mean it's not achievable. Yeah. It's awesome. Andre, thank you for taking the time with me today. Uh, I love what you're building.
• 21:39
  It's really a great vision for the future, and I also love that you're so steeped in the research, you know, you and your partner both coming out of PhDs, really caring about proving out what you're building and that it's useful to, to hospitals and to patients.
• 21:56
  So, uh, thanks for all the work you're putting in and for taking the time with me today. Thank you very much. It's a lot of fun. All right. Be well. You too. Bye. [outro music]