Host (00:01):
Welcome to Insight Exchange, presented by L.E.K. Consulting, a global strategy consultancy that helps business leaders seize competitive advantage and amplify growth. Insight Exchange is our forum dedicated to the free, open, and unbiased exchange of the insights and ideas that are driving business into the future. We exchange insights with the brightest minds of the day, the most daring innovators, and the doers who are right now rebuilding the world around us.
Sheila Shah (00:40):
Hi everyone. Thanks for joining us to talk about an often discussed yet often misunderstood topic, digital health. Digital health has become one of those terms that means everything and nothing at the same time. However, while opinions are varied on the exact impact and role digital health solutions will have on the healthcare ecosystem going forward, there is solid agreement, which is a rarity in healthcare, that we all need to embrace digital in our clinical and nonclinical lives. How to do this, however, and which companies will be able to do so successfully is still remain to be seen, which is why we're so honored to be joined today by Steve and Sherie from Viz.ai to talk about their company solutions and how they've been able to successfully implement their digital vision, and importantly, why it's working.
Max Lounds (01:27):
Thanks, Sheila. I'm Max, an engagement manager at L.E.K. Consulting and a leader in our digital health practice. I have the honor to introduce the rest of our cast today. You just heard from Sheila Shah, a managing director and partner at L.E.K. Consulting, focused on healthcare. She leads the firm's digital health work. We are joined by two of Viz.ai's best, Steve and Sherie. Steve is VP of business development and strategy at Viz.ai, where he leads that function. Prior to Viz, he held a number of leadership roles at Medtronic and started his career in healthcare consulting. Sherie is a senior manager of commercial strategy at Viz.ai. Prior to Viz, she was a healthcare consultant advising pharma and MedTech companies as well as early stage health tech startups.
Sheila Shah (02:14):
We're so grateful to both of you for joining us. Before we dive into the real focus of our discussion, let's get to know Viz.ai a little bit better for those who may not be as familiar. What problems does Viz.ai aim to solve?
Steve Sweeny (02:26):
Thanks, Sheila. Yeah, well thank you and Max for having us on the podcast. We're obviously very excited to be here with you and talking about what we do and how we help providers and life science clients.
(02:38):
Maybe we just start with the problem. We serve both providers and life science clients in improving patient access to care. What we mean by that is we work with providers to get patients to the right care within the healthcare system. This is actually one of the bigger problems that exists in healthcare that few talk about. We estimate that over 80% of patients that could benefit from a therapy never actually reached the right diagnosis, the right clinician or are on the right care plan such they would be able to benefit from all interventions that exist and have been proven over the years.
(03:12):
These patients that struggle with this are touching the healthcare system, but the problem is they're not properly diagnosed. There are care coordination challenges. The patients are often stuck in the wrong care plan or there are gaps in the workflow that stand in the way. For pharma and medical device companies, this is a massive market development challenge and it's only going to get worse with the focus or the increased focus on rare disorders and more targeted therapies as well as the increasing barriers that exist in terms of accessing HCPs.
(03:43):
So where does Viz come in? How do we help solve this market development challenge? Well, we do so by first helping clinicians identify the right patients. And I'll speak to how we do that in a bit, but it's identifying the right patients and then ensuring that those right patients are seen by the right clinicians, so essentially triaging those patients to the right clinicians. And then empowering those HCPs with the right information and tools to take action.
(04:12):
So how do we do that? We operate within 1,500 hospital systems across the United States. And in real time we analyze data feeds that are generated on our patients. So CT scans, echoes, ECGs, EMR, data, notes, using AI to identify signals of disease or often that the patient has just fallen off of their care plan. So once we identify those patients, we then route them to the right care team within that health system, providing those ACP with the right information at the right time to take the right next steps. We do this through our Viz One platform that's used by over 35,000 ACPs today. And right now across our entire footprint, we see a patient about every 14 seconds.
Sheila Shah (04:55):
Wow. Thanks, Steve. Thanks for sharing that and going through it. As you've mentioned, I think what's really interesting about the problems that you're solving is that it actually solves similar problems that manifest itself in different ways for each of the different healthcare ecosystem players. And so whether you're a hospital, a MedTech or a pharma company, you're sort of solving a similar problem that, again, manifests in different ways. So Sherie, can you talk me through how the different solutions that Viz.ai offers for each different stakeholder are actually quite similar?
Sherie Zhou (05:28):
Sure, Sheila, happy to. As you heard Steve just say, at our core, Viz has two main things. We identify patients at a health system level and then we connect the appropriate physicians so that patients receive timely and accurate follow-up.
(05:40):
If we look at these two activities through a hospital and provider lens, Viz is helping them identify, diagnose and follow up on patients, including ones who likely would've been missed. And then on the flip side, from a pharma MedTech company perspective, Viz is really optimizing that patient journey. And what we mean by optimizing the patient journey is increasing the number of patients that reach treatment and accelerating their time and sort of the pathway to do so. This is really important to us because we know that for many conditions, especially in rare disease in an oncology, the patient journey to getting the appropriate drug or device can take many years if that even happens.
(06:17):
In particular one of the cardiology rare diseases that this is working in now, patients might bounce around the health system for as many as five years before they ever receive an accurate diagnosis. And similarly in cancers, we know that early diagnosis is critical for treatment and successful treatment, but too often patients are often discovered too late despite having interacted with the health system through a number of touchpoint. And we see patients are getting shuttled back and forth between multiple specialties and getting many rounds of tests and potentially many will ultimately leak before they receive a definitive diagnosis or the appropriate treatment.
Max Lounds (06:53):
That's interesting. It sounds like improving the patient journey and addressing care gaps are a common theme to your value proposition across the customer segments that you serve. We're finding that many digital health companies are having a hard time right now. Being able to clearly communicate their value proposition is one of the key challenges that they're facing, and it often inhibits adoption or gaining meaningful traction and usage. Can you share a bit about the secret to Viz.ai's success? What do you find is driving usage of your core offerings?
Sherie Zhou (07:24):
Sure, Max. I'd be happy to. I think at the core, our achievements in our progress to date are really because we're a clinical-first and patient-driven company. It's actually one of the core Viz values and one of the reasons I joined Viz about two years ago. Having that patient-first mentality informs everything we do from what products we develop to the nitty-gritty of the features and the user experience. And before we decide to build a product or go into a specific disease area, we really ask ourselves a question of, "Is there a provider pain point that we can solve?" I think those who are more familiar with technology products might know this as product market fit. The way that we adapt that for a healthcare setting is a lot of HCP interviews and we talk to them from a variety of care settings, whether it's inpatient, outpatient, more suburban geographic markets, urban ones, rural ones, as well as understanding the latest in clinical guidelines and treatment guidelines for these disease areas and we use that as the basis for our product development.
(08:21):
And as we pilot our products and bring them to market, we continue to iterate based on feedback from our physician users. Through our years doing this, we found that it's really essential to map clinical workflows, really understanding who is interacting with who, to surface real versus perceived bottlenecks, and how patients are getting triaged at a site or getting diagnosed or getting scheduled for appointment and follow-ups.
(08:46):
And additionally, when we go into a new site or a new health system, we don't just throw the Viz tool out providers and tell them to figure it out. We have pretty sizable teams in clinical workflow implementation and technical support to name a few to really ensure that health systems get on the right foot with Viz and that they get value out of it. And then post-launch, we have a customer success teams who are the go-to point people post-launch. They also help review on a quarterly basis what the site's usage looks like and the patient outcomes with clinical and hospital administrative leaders to really understand what is the ROI that they're getting out of using a tool such as Viz as well as patient improvements, stars metrics improvements and other things that they might care about.
(09:32):
In a nutshell, providers use our products because they see the impact. They see the impact on patient outcomes, on reduced patient length of stay, on improvements in quality of life, and on life saved.
Max Lounds (09:43):
That's really interesting. I think one theme that I'm noticing in what we've been talking about so far is this idea of integrating and optimizing within current healthcare workflows. There's so much talk in the digital health industry about the promise of digital health, transforming how healthcare is delivered. It sounds like you're suggesting it's important that transformation needs to work within existing systems rather than trying to upend them entirely. Can you tell us more about that?
Steve Sweeny (10:11):
I can talk to that, Max. It's 100% true. And I think one of the more common misconceptions of AI generally, but especially in healthcare, is that you can give someone our team the AI and they'll immediately start using it and get value out of it. When we deploy one of our disease solutions, at a healthcare system, we do a few things to ensure that there's real use and value out of the solution. The first is we understand and really deploy a team into the site to fully understand the current workflow, who's interacting with who, how patients flow through the system today. And with that, we tease out the inefficiencies or bottlenecks or pain points. And as Sherie alluded to this a bit earlier, and then once we have that mapped, we work with the sites to say, "Hey, X, Y and Z isn't working for you today. How can we make this better, easier, faster?"
(11:00):
For example, in one academic site in their heart failure department, we streamline how they see and diagnose a rare cardiac disease for patients from 30 steps down to nine to 10 steps. So you can see just shortening that and changing nothing else saves a ton of time and administrative burden on the different providers involved.
(11:20):
And then lastly, we make sure the right people have access to the application to the modules that we're deploying. It's not that everyone in the hospital needs it. And in a lot of these cases, it's only a select number of people. So when we're dealing with patient identified information, we're looking to identify the right specialists, the cardiologists, neurologists, but also members of the patient's care team, the nurses, the APPs, office managers, to ensure that those folks have access to the application that's specific to the particular use, that we train them on them on the application, and that they're actually able to fully use it in the way that it's meant to be used.
(11:57):
And then once we have that all in place, then it really starts working. And obviously as Sherie mentioned, on our ongoing basis, we have customer success teams and support teams that help ensure the tools and solutions are working as best as they can within NHC of the different health systems.
(12:16):
So that's generally how we think about working within the existing health systems. But one other common misconception I think folks have around AI is that AI is going to replace humans. We absolutely disagree with that completely. We don't even think it's the right goal to strive for. Where we think AI adds value is really helping augment and streamline clinicians and administrators tasks day-to-day, whether this is in the form of AI serving as an additional reader, improving communication in terms of friction points between specialties or internal external parties or other use cases. Our intention with AI is to enable clinicians to focus on clinical care, really the reason they got in the profession and allow AI and really the digital solutions to act as a bridge and take out some of the friction that exists in the system today.
(13:07):
I think one thing that we've particularly seen during COVID and even subsequent to COVID is a significant number of staffing challenges, which has led to a backlog in patients along appointment wait times. It takes forever to get an echo these days. Enabling AI or even workflow solutions to tackle some of these friction points really allows patients to get in faster, physicians to focus on clinical care, and then ultimately addresses the challenges that exist in terms of developing a market if you're a farmer or a MedTech company.
Max Lounds (13:40):
Got it. So the not so secret secret to success is ensuring that you truly understand your customer's pain points and needs and staying hyper-focused on solving those. That seems so simple, but I'm sure in reality is hard to really know when you're hitting the nail on the head. Can you share some of the feedback you hear from your customers that has helped you to know that you are meeting that goal and improving their day-to-day?
Steve Sweeny (14:06):
Yeah, absolutely. And it does seem simple. It's certainly not simple, but the goal is to make it as simple for the providers and the folks that are on the application as possible. So some highlights of how we're working to knock down some of these market development challenges that exist for our pharma and device clients. I'll start with where we started with Viz is in ischemic stroke, which is before Viz came into existence, a high percentage of stroke patients faced delays in getting care and weren't being treated as quickly as they could be treated such that the number of mechanism thrombectomies was a lot lower than it could otherwise have been. So we started there.
(14:49):
And how we operated and what we built for that particular use case was really a tool that used CT scans and AI on CT scans to identify patients in need of care by an interventional neurologist or suspected of having an LVO and that triaged those patients to those right stakeholders. And because of that, we were able to significantly increase the number of mechanical thrombectomies happening at the sites that we deployed, and as a corollary showed a significant reduction in patient length of stay among other patient outcomes. This is in a condition where the standard of care was already fairly good. So just by identifying these patients early off the CT scan and then mobilizing the care team through our workflow solution, we're able to not only improve patient outcomes but significantly grow the market for mechanical thrombectomies. So that's one example.
(15:40):
Another example I can give is a use case that actually doesn't really even use AI at all. It's just a workflow solution, and that's in cryptogenic stroke. So we created with a life science partner a solution that looked to address the post-stroke care coordination challenge. And just to give you a sense of what that looked like, today almost about 40% of patients that have a stroke are deemed to be cryptogenic, or in other words, the cause of the stroke is unclear. A significant portion of these patients are at high risk for secondary stroke often due to atrial fibrillation. And a high percentage of these patients also don't actually receive follow-up by the cardiology team, which is a critical step in the workup to not only fully diagnose those patients and address the underlying cause of their stroke such that you're preventing a subsequent stroke, but a lot of these patients don't actually receive the follow-up.
(16:34):
So what we did was we actually created a solution that allowed the neurology team very simply to triage these patients post-stroke to the right cardiologist while still at the hospital. Building it wasn't that simple, but as far as the users go, it's a very simple solution that allows the neurology team to essentially triage the patient to the cardiology team. What happened was we were able to show a 2X increase in the number of patients receiving a cardiology consult. We moved those consults and the intervention from those consults from weeks to months out to days. And so these patients were going home with the right care to prevent subsequent strokes. And for the life science partner, we saw a significant increase in the number of patients they were able to access their therapy.
(17:19):
Now I'll leave you, Max, with one final example. And this is an example I think Sherie highlighted earlier with a rare cardiac disease that we're doing work in. So we're currently sponsored with a large life science client with whom we created an AI-based solution to detect and triage patients suspected of having hypertrophic cardiomyopathy. For that disease, it's estimated about 20% of patients are currently diagnosed, that means 80% of the patients with disease are out there undiagnosed. And for those that are diagnosed, it takes years to diagnose. So this is a huge problem not only for the patients, the providers, but also this life science sponsor who has a drug that's proven to work for this disease.
(17:59):
So we created a solution to address this and recently received FDA clearance for that. The solution uses AI leveraging ECGs to detect these patients and then triage them to the right specialists. So it's still early in our launch. We just recently launched. But one recent case example I can give you is of a 74-year-old woman who came into a cardiology clinic for a post-op visit for another condition. Her cardiologist ordered a routine ECG. Fortunately we were to flag that patient as having suspected ACM. We were then able to triage her to the ACM specialist and she got the appropriate workup, which ultimately confirmed that diagnosis. This case was especially gratifying because she was known to the health system for many years. Shed had many touch points with the cardiology department but had gone undetected until obviously we flagged her.
(18:48):
And so, broadly around this solution, we've been able to show even just in the short timeframe that we've launched a significant decrease from years to months in terms of diagnosis. And so these are just some examples of the work that we're doing to knock down barriers to grow the market for life science clients and then provide better access for these patients to the care they need.
Sheila Shah (19:11):
Thanks, Steve. That's super helpful and really great concrete examples. So on more of a personal note, I have to ask, there's so many opportunities in digital health right now. There's so many great companies that are looking for talent like yourselves who are both extremely successful and intelligent. Why did you join Viz.ai? What about the organization and the mission really spoke to you?
Sherie Zhou (19:34):
Sure. I'm happy to take that one. I think two main things attracted me to Viz, it would be its people and its purpose. When I joined, this was in 2021, Viz had already accomplished a number of really impressive improvements for stroke patients. I think I was just exceedingly excited about the opportunity to be a part of that growth into other disease areas and into other customer types. I think equally important, just the people that I met, everyone throughout the multi-step interview process was incredibly sharp, incredibly kind, and just everyone has been united in this very patient-centric mission.
Sheila Shah (20:09):
Thanks, Sherie. That's so great to hear and thank you for sharing that. I think it echoes a point that I think is often heard across the digital health industry that the purpose the work provides is a huge motivator.
(20:21):
As many in the industry have experienced and as we've discussed some already, achieving the purpose though is quite challenging. One of the obstacles that can stand in the way of progress is that buying and implementing digital health is fundamentally different from what most healthcare organizations are used to compared to the pre-digital health era. And so, while some customers are sure might be familiar with purchasing from digital health companies and might understand the SaaS software models, that really hasn't been the case and that's not necessarily pervasive. And so, novel tech-based solutions and business models borrowed from the tech industry can be foreign as we've discussed. How has this AI navigated this dynamic?
Sherie Zhou (21:03):
Sure. Yeah, you bring up a really good point. When we think of the main industries that lag in tech adoption, healthcare and financial services are the two main ones. It's honestly very understandable because there's a high level of security and privacy involved in both of them. I would say it's been an easier journey now than when we got started in 2016, 2017 and AI and health tech tools are much more a part of the dialogue. Our customers, whether it's health system executives, MedTech commercial leaders, pharma commercial leaders, everyone's now aware of AI. And so they're asking questions about it. And for us it's been a mix of educating on what we're able to do with AI-enabled workflow tools, but also listening to and ideating with our customers. Some of the things that we've had at brainstorming sessions with them are around market access or mandatory disease monitoring. Ones that we might not have thought of as quickly on our own.
(21:56):
I think the second part is we always get a ton of questions related to security and privacy standards and adhering to HIPAA, whether or not we're FDA cleared. We are. And we've also passed all of our FDA audits with zero findings, so I want to give a shout-out to our regulatory and QA team there. And for anyone who's curious, we have all of our certifications and best practices listed on a website, the Viz Trust Center, and ready to share that with any stakeholder who wants to learn more.
Sheila Shah (22:23):
Thanks, Sherie. Have you observed changes in the receptivity, the decision-making of customers in the past three years as the market becomes a little bit more mature and as they navigate competing priorities and low budgets? Has your go-to-market approach changed as a result of it?
Steve Sweeny (22:39):
Thanks, Sheila. I can take that one. We have seen increasing interest, particularly as we've been able to deliver on a number of these programs. Growing the market is one of the biggest levers you can pull if you're a MedTech or a pharma company. And so it's a huge need. Now what we're providing is a bit novel, so not all of our pharma and MedTech clients have full budgets allocated to this type of work. However, we've seen increased maturity in the space and companies are beginning to make long-term investments in budget allocations to this type of work. So the interest is definitely growing.
Max Lounds (23:17):
With our remaining time, we want to shift gears to discuss a particular development that's on everyone's mind these days, artificial intelligence. We've already heard it mentioned a few times in our discussion today. And it's in your organization's name, so we would be remiss to not get your thoughts on it while we have you. AI assisted analysis of diagnostic imaging to support disease detection is probably among the more well-established and discussed applications of AI in medicine. But where else does Viz use artificial intelligence or see room for artificial intelligence to impact healthcare in the future?
Sherie Zhou (23:56):
I can take that one. We see a number of use cases throughout the patient journey. So if we think about detection, which you touched on just now, that's sort of the bread and butter of what we've been doing recently, alerting HCPs to patients that would be missed or lost to follow up.
(24:12):
And then if we move to the next step in the patient journey and really in care coordination and figuring out the next best clinical action, that's where the intelligent workflows come in so that the right specialists are communicating with each other on a timely basis and that the right patient information is getting shared and so that patients ultimately get worked up, diagnosed and treated in efficient and in a timely way.
(24:34):
I think where we see the next biggest area for AI impact is in patient monitoring and adherence, particularly three main things that we're exploring right now. One is many of the novel innovative therapies coming to market that Steve highlighted earlier. They have pretty heavy monitoring requirements. So whether it's regular MRIs every three to six months or getting regular lab tests to monitor patient response to therapy, this presents a pretty big increase in number in the amount of medical data that specialists, including radiologists, have to look at. And we also know that radiologists are already time constrained, so there's huge opportunity for AI to work in parallel here. I think just echoing what Steve said earlier, we don't think as a company or personally that AI is here to replace a doctor or physician or a nurse at any point. It's really a complimentary tool for them.
(25:25):
I think second is around patient adherence. We know that there are a number of factors that affect whether a patient will stick to therapy aside from how well they tolerate it or the efficacy of it, but there's a whole host of social determinants of health that impact whether a patient can make it to their next infusion appointment or get the laboratory tests they need before their prescription is refilled and many, many more. There's a huge potential for AI to identify patients who are potentially at risk of lapsing and flagging them to the system so that they can get the additional support that they need.
(25:58):
And then lastly, on the remote patient monitoring side, we see an opportunity for AI to risk stratify or potentially predict patients who are going to need a little bit more follow-up post procedure than the average patient or just patients who are more likely to have a complication and fogging them to the health team so that that potentially serious event can be avoided. Just a couple of things we see in terms of use cases along the patient journey.
Sheila Shah (26:23):
Thanks so much, Sherie, for going through that. And Steve and Sherie, thank you so much for taking some time chat with us today. As we wrap up here, what are some of the things that you really want to make sure that listeners take away from this conversation?
Steve Sweeny (26:36):
Yeah, I can share some final thoughts, Sheila. And again, thank you to Max and yourself for having us on the podcast today. What I'd leave everybody with is the commercial models in life science for both MedTech and pharma are already challenged. This is honestly related to a lot of the challenges providers face, the staffing challenges, the explosion of information that they have to track, whether it be in the clinic or outside the clinic. These are all very real. What that means is if the providers are facing challenges of being able to care for all these patients and the explosion of new information, then moving a patient along their journey is going to become even more tasking and challenging.
(27:24):
And so AI has a place in addressing these gaps in these challenges. However, AI by itself is not going to be sufficient. The key is having the clinical workflow layer, really how do you translate that AI into action and then the ability to deploy it across an entire network. So how do you actually make it scalable? I know this is an area that a lot of life science companies are focused on, and how can you use AI to address some of these commercial challenges? And I think the keys are really how do you make it actionable so that clinical layer, particularly if you're going to be operating within the health system setting, and then how do you make it scalable.
Sheila Shah (28:02):
Thanks, Steve. I just want to say on behalf of Max and myself as well as Steve and Sherie from Viz.ai, thank you to the listeners for taking the time to really engage in this conversation with us. It's been really exciting learning more about what Viz.ai is doing, and we're really impassioned about where the future of healthcare is going. So thank you again for being part of the conversation and listening in. If you have any questions about digital health and where it's going and really how you can get involved in it, please give me and Max a shout. We'd love to have a conversation with you.
Host (28:33):
Thank you, our listeners, for joining us today at the Insight Exchange presented by L.E.K. Consulting. Links to resources mentioned in this podcast can be found in the show notes. Please subscribe or follow for future episodes wherever you listen to your podcasts. Also, we encourage you to submit your suggestions for future insights online at lek.com.
