The next frontier in patient support: Connected data, AI and personalized engagement (Sponsored)

Introduction to Zayden

0:00In global life sciences, insight isn't the problem. Execution is. Disconnected dashboards, silo teams, and fragmented workflows make it harder to act consistently, compliantly, and at scale. Zayden by ZS closes that gap. Zayden is a life sciences intelligence platform that embeds intelligence into the flow of work based on domain expertise and agentic AI across commercial, medical, patient, and content work. Trusted by more than 150 life sciences organizations

0:32across 100 countries, Zayden is built for global scale, regulatory compliance, and trusted AI. So teams act together across brands, markets, and channels. Zayden, move beyond reporting to intelligent action. Learn more at zayden.ai.

Podcast Introduction

0:52You're listening to a sponsored episode of The Top Line. Welcome, everyone. You're listening to The Top Line, brought to you by Fierce Biotech and Fierce Pharma. I'm your host, Stephanie Butler. Today's episode is sponsored by ZS,

Patient Support Challenges

1:28and our topic of conversation today will be the next frontier in patient support, connected data, AI, and personalized engagement. Despite significant investments in patient support and digital engagement, many patients still experience delayed diagnoses, never start therapy, or struggle to stay on treatment. The challenge isn't a lack of data. It's that the patient data remains fragmented across multiple systems, making it difficult to understand and support the full patient journey. Today, we're joined by Ashish Shukla and Kirti Prakash,

1:59who are both principals at ZS, to discuss how life sciences organizations can build a connected patient data foundation that enables earlier interventions, more personalized support, and better outcomes at scale. Ashish and Kirti, thank you both so much for joining us. And before we begin, I would love for each of you to just give me a brief intro of yourselves and your role at ZS, and maybe we'll start with you, Kirti. Thanks, Stephanie. Hi, everyone. I'm Kirti, principal at ZS, focused on patient strategy and services. I help a lot of life sciences companies enhance how they engage and support patients

2:31by combining strategy, data, analytics, and technology, and help improve both patient experiences and program performance. I am an expert in the patient data management side of things, and I help clients both through our ZS's Aiden platform, that is patient analytics and insights, and also through custom development. Excellent. Ashish. Stephanie, thanks so much for having us. Exciting topic. You can see the passion from Kirti and me coming out to the conversation. Ashish Shukla, principal at ZS. I lead our global patient strategy and services practice area, which prioritizes focus on how the patient can get the best therapy

3:05they can get from the manufacturer, the drugs they are on, and live a healthier life. That's what we prioritize and we focus on. I also lead our ZS in patient engagement platform offering, which basically converts some of the services as a software. So we do bring services as a software platform to our clients, especially when they will benefit from such a platform. Thanks for having us again. Thank you both for that. I agree. This is one of the topics I, too, am also very passionate about and really love talking about. So I'm really excited to kind of get into this today.

Patient Data Sources

3:33So I want to start by saying, look, we know patient services today generate an enormous amount of data, like across the whole ecosystem. Where is all this data actually coming from? And kind of what's the state of it right now? So, Kirti. Yeah, I think over the years, patient services has quietly become one of the most data-rich functions in pharma. You know, be it your patient hubs, enrolling patients into the services, benefits verification, especially pharmacy dispensing or tracking refills, COPE and PAP programs, you know, you name it and the data touchpoint

4:06for that patient is actually captured. So on paper, that should actually mean that we understand the patient better than ever. The reality, however, is the opposite, right? The data sits fragmented across a dozen vendors in inconsistent formats. The hub may have its own view. SP has another. COPE has another. And then call center has some notes that no one else can actually see, which is actually so rich in data, which will tell you how much of, you know, what is the patient actually feeling? Patient can be flagged high risk in one of the systems, and the case manager in another might not have any insights whatsoever, right?

4:40I mean, if you think about it, the data is in abundance. It's really available in abundance. But what we are dealing with is poor decision quality, right? And the cost of that gap and inaction is quite true. If you think about it, there's roughly half or roughly one third of the prescriptions that are never filled in the US. And up to half of those patients who start therapy discontinue in the first year itself. Non-adherence drives an estimated of about $500 billion in avoidable costs. So that's huge, right? So we did a research and we actually found that about 53% of these issues are addressable.

5:14And if we can intervene at the right time to help both the patients and the treating physicians. Problem was never, you know, the shortage of data. It's essentially that the data has never been connected. Ashish, what do you think? Anything that you'd like to add? No, no, absolutely, Kirti. I agree with everything you said. And I think if I step back and think slightly more holistically and look at all the innovation happening in healthcare, the problem, I think the question you asked, Stephanie, is getting compounded. The patient services from manufacturers are getting increasingly personalized. They're becoming very unique to the treatment journey these patients are on.

5:44And what we're seeing is that combine that with provider platforms getting connected. The CMS, for example, in U.S. and similar agencies, the rest of the world, they're almost mandating interoperability of patient data. And one thing that's always a glass half full view for me is that privacy and concern safeguards are increasing. So net-net, all these coming together are just exploding the volume of patients that are becoming available to us, as I think ITI outlined. And of course, when you look back and combine this with the recent research we published in our ZS Future of Health report,

6:15we're also learning that patients will trust AI agent for a medical advice up to 90% confidence. And that's kind of chose us and gives us a pause that it's a clear signal that patients are willing to share their personal health data in return of a valuable service, especially if they manage their disease with their providers and the physician and so forth. So assertion remains that the data evidence has never been a problem. I think the skepticism around decision quality, as Kirti flagged up, has been a question mark. And what we're observing is that it'll only get amplified

6:46if you don't act on it. Yeah, it's a great point. So what do you think is fundamentally changing right now, right? In the way that data is being generated, consumed, and as you mentioned, trusted, and in the way pharma engages with patients? Ashish, want to follow up on that? Absolutely. And another fascinating question, Stephanie. So the way I think about, and I think collectively in the US, how we think about this is like, what is shifting? What's changing in industry? And we've summarized in four categories, four shifts. One shift is what we call healthcare mindset shift.

7:17If you look at historically, healthcare has always been about illness management. I think we intervene when some issue has already happened. What we're observing is that this shift is happening from a positive and intervention, reaction-oriented, to proactive and predictive wellness care. Almost the patients are demanding it. And if you look at the, we are living longer and the problem is just getting expounded. We do want to be proactive in managing the healthcare of our population. The impact is profound across healthcare, but staying focused on data.

7:49What is shifting is that the relationship we had with patient data from the past. Historically, what we have done is we process these patient data on a batch, kind of asynchronous model, like a pharmaceutical will take data from their specialty pharmacies and the hubs, and they will process it like once a week or once a day. But the demand of this mindset shift almost is pushing that I need the pulse on the intelligence to drive better decision, to intervene in a timely manner. And that's forcing the data to become more real-time integrated in the ecosystem,

8:21almost to a level where we are streaming data from things like variables in other places. So, submission is the mindset shift is changing and the data implication is that we shift from a once in a periodic relationship with the data to an ongoing relationship with patient data and figure out how do we enable it. So, that's one big shift. Second shift, which I kind of mentioned the first one, what's driving it is the timely intervention. One of the key aspects is patients and the provider, if you ask them, they need help when they face a barrier and potentially if you can predict and intervene

8:54before they face a barrier. Barriers such as like getting a pre-authorization for a script gun or benefit verification eligibility of a particular patient on a therapy option they're considering. And these are real barriers they face day in, day out. And these are barriers which kind of come in the way of managing the disease for the patient and providers are frustrated. And many times these agents off late are supported by what we are calling as AI agents. It's becoming almost table stake off late. And if you think in those terms that humans and AI agents are working together to support the assembly intervention, the implication on data is that you cannot afford to only create data

9:29which is post-event, reflective, insight-oriented, but you have to pivot towards creating in-the-moment data to the agents so they can actually intervene using the information. So that's a big shift. It's a big one for us, I think, across the board. And our thesis in this particular shift is that the data probably, which has been historically built for humans and BI tools from a dashboard and reporting perspective, it needs to evolve into execution-ready data sets, actionable-oriented data sets. And that's a very different structural-wise and so forth, which we observed in the past.

10:02These two are the main driver, I think, from a data perspective. There are two other ones which are impacting our thought process and where, for them, what's changing. Third one is, I think, the consumption side. We all live in a new generation, the new generation of people getting, showing up as the patients and all that. And they consume information in sight very differently than historically what we've consumed. And these are human behavior. They are consumerization of healthcare, you can call it. In technology realm, for example, you'll hear this a lot as headless, headless engagement with data. Nobody cares about logging into an app or getting onto a particular portal.

10:36I think the engagement, the new consumers are showing up, and they're showing up using LLMs or co-pilots or multi-system agents to interact with information. They don't really care into logging into some system to do it. And if you think about it, most of these are in sight. I think we flagged up the research about 91% will go to an AI chatbot for medical advice. Guess what's driving these medical advice? It's the underlying data. It's the underlying thing and what we observed. And we do this routinely that most of these agents in the real world, they're failing and they're failing to live up to the expectation of the user, both patients and provider.

11:09And one of the foundational reasons is underlying data. They just lack the semantic context to make the response relevant to the patient, to the provider, to their situation versus like a generic response with 1 million patients are getting it as well. And the fundamental backbone of that is lack of semantic context. That's the third shift we are seeing that a lot more data need to be structurally ready so that the new consumption patterns actually find it valuable. And the last I will close out with is that if you combine and layer in these three things

11:40about how the supply is happening or the demand is getting generated, one thing will never go away is the trust because we're talking about healthcare, we're talking about patients. And trust is almost like a gating factor in our opinion, the shift wise. And we know that if we add value, patients and providers will engage. We know that the technologies and the AI agentic world is evolving where you can actually put it to very good use. But if in this new world, data isn't governed, well-defined, compliant, these AI agents actually

12:14will compound the compliance risk for everyone across the board. So there's a third area where how do we design the foundation where trust is like a first principle thinking around the data? Like you have constant management, preference management, big into the way you manage the life cycle of patient data. And that historically has not been the case, but we believe that shift is real and I think almost as critical for us as we move forward. Yeah. I love that point about trust too, because that is always going to be underlying fundamentally. But if you lose that, it starts to compound exponentially and then it becomes an even bigger

12:48problem, right? Agreed. And all the improvement we've done in the past, everything goes off the window because nobody trusts this platform, nobody trusts the system and all that, which is kind of moving a step back. Exactly. I mean, that's a lot of change and very quickly, I think.

Organizational Response

13:03So given all of that change, how should organizations respond? What does getting ready actually look like? Kirti, I'd love to hear kind of your thoughts on that. Yeah, if I were to look across healthcare and respond to this question, Stephanie, I think we'll just need two or three more episodes of this podcast. What's a point? Yeah, but I think let me limit my response to how pharma should respond. And I think some of it Ashish already touched upon, right? It's an easy answer, but poorly executed. Comes down to primarily two connected imperatives, right? And both have compliance baked in.

13:34We were just talking about trust and how the data needs to be managed compliantly. I mean, that is absolutely the essence of everything that we do in the patient world, right? So I think the first one I would talk about is a strong connected data foundation and managed compliantly. What does that mean? That essentially means that, you know, you are linking all of the patient level data into one single longitudinal view. Your interactions at the patient support program, hub, specialty pharmacy, whether or not the patient is taking copay. What does the medical record for the patient indicate?

14:06So this is the data part of it. Along with that comes the capabilities. So are you tokenizing the data? What sort of patient mastering are you doing? How do you de-identify the data correctly so that, you know, again, all of the data is handled in a compliant manner? And how are you storing this data in a HIPAA and Hydrast aligned environment? So everything has to be very governed. So, you know, governance is not just a band running alongside the architecture. It is the precondition for every action that is happening in the patient world. Even things like consent, right?

14:37So consent actually dictates a lot of that last mile interaction that you can actually do for the patient. So what is that consent language, consent scope? What sort of PHI masking rules will you apply on the data? And all of that has to be encoded in some sort of a... Ashish was talking about runtime, right? So how do you make sure that some of these things are encoded into the machine operable policy? And that is enforced at the query time. So I think that is the first one that I would talk about. And the second imperative is the last mile action itself, right? So not just the data sitting underneath it, but how are you reaching out to the patients?

15:11That has to be equally compliant. It's just not enough to have a clean data core. What is the case manager interacting with the patient? How is the case manager interacting with the patient? What does that FRM intervention look like? And all of that has to happen within the consent boundaries. So the same governance essentially has to travel all the way to the point of the engagement, right? So I think nowadays, technology is sort of the easy part. You know, technology is not the challenging aspect. The harder shifts are the ways of working. Think about it, right? Like someone has to own each of the nudges.

15:42You need a governance and consent model. How are you actually measuring the impact is also equally important, right? So historically, if you see these were being operated as a one-off project. Like there's some use case, you want to be able to enable it. Okay, you do it as a one-off project. But teams have now started to see the value of building this as a capability than actually doing a one-off project. So I talked about like ZSS, Zaden, patient analytics and insights platform. So that is one of the platforms that actually helping us accelerate that journey for pharma and many of our other clients.

16:12And yeah, I think also one thing when it comes to the mindset shift, right? So our experiences of working across clients have become much, much better at tackling some of these compliance questions. If you think about it, like a few years back, everybody was very risk averse. Then we used to talk about patient data or PHI data and whatnot, right? Nobody was willing to make those investments to connect the data. But now that people have started seeing there is so much value in it, right? We are actually seeing a significant hope. I mean, we are actually hoping that almost everyone wants to help patients manage their

16:46disease better. And our experience of partnering with some of these compliance teams has also shown us that, right? Obviously, they are concerned about security and compliance. So we just have to navigate the questions and the genuine concerns that they raise. But I think good news is that the combination of the platform, process, policies, et cetera, they are able to, through that, we are able to address most, if not all of the objections that we've encountered. Yeah, I completely agree, right? Like I have seen a shift in, you know, understanding that having the patient data, even though once

17:17you deal with the compliance challenges, once you have the data, the end goal is to help the patient. And if this is one of the ways we can really do that, that should be one of the priorities, right? Right. So, you know, we're talking a lot about our systems, but I want to kind of get a little bit practical here, right? So if you're a pharma's head of patient services or the head of technology supporting, you know, patient data, and you're listening today, what are those kind of no-regret moves, the things that they can start with now, regardless of where the future goes? Because sometimes that's a little unpredictable.

17:48So, Kirti, I'd love to kind of follow up and hear your insights on that. Yeah, no, that's true. And maybe I'll start, and Ashish, you can add as well, right? But, I mean, I'll bend upon my last response itself, Stephanie, right? Like, the absolutely start now categories is structuring your data as data products. So if you think about it, again, earlier, the idea was that you feed the end use cases through a big, large table. But we don't want to do that anymore. Don't try to serve everything from one giant table. You know, how do you essentially structure your data into data products?

18:18So things like foundational data products, you know, which is integrated transaction level activity across Scope, Hub, SP, et cetera. How does that go into functional data products, which is essentially, you know, enterprise-like standard definitions of KPIs, business rules, et cetera, so that there is no rework at the end. So patient journey, access and affordability, trade channel. Then there is a concept of fit for use case data products, right? Which is essentially used for specific decisions and use cases. And then now, increasingly, in the last few months, I would say, we're also hearing a

18:49lot of the fit for AI layer, which is essentially all of that I've talked about, right? How do agents essentially read from that AI layer? How do you enrich the semantic context around it? And so on and so forth. So the layered data products essentially balance scale, trust and actionability. And trying to force everything from that one giant foundational table, it just creates a lot of duplication and rework, I would say, right? So that's one. And I think the other one is just knowing the personas that you're activating. So being very explicit about who is consuming that final information.

19:22Is it patient care analytics? Is it field reimbursement, patient marketing, and so on and so forth. And because if you think about it, the data foundation and the last mile design, both of that will be designed based on these personas, right? So it's very important to know who are those last mile actors. And yeah, my advice would be just start now. Start unifying your data now. Don't really wait for a perfect data warehouse, data lake, because some of these things you typically tend to depend, right? Unless and until you have a good data architecture, like I would say just go on that journey already.

19:55Ashish, anything you want to add? No, you covered very well. And I think just building upon Steph, your question. One thing is very clear that if conviction, if for example, head of patient services, and I think technology leads, they are not in business of just doing technology for technology's sake. They truly believe that they should serve the patient, as you flagged up Stephanie earlier. So from a strategic perspective, there are two things I'll flag up. Technology evolves. Yes, you can always wait for the best technology to become available to you. Or for example, as Kinti flagged up, all data available to you can solve the world hunger.

20:28I would add that any of our pharma clients, they're managing a portfolio of products. Some of them, of course, have one asset and all that. And there is an investment required, almost like a fit-purpose approach to where the assets sit in life cycle. And the fundamental belief is your barriers, the patients and the provider and the payers face in adopting the therapy which we are promoting in the marketplace, depends upon where the product ends in life cycle and, of course, what the competitive scenario

20:58and things are. For example, a pre-launch of a rare disease asset is a very different than an established brand. And our submission is that adopt fit-for-purpose compliant patient data management. Your data availability, sometime a third of actually creating data which doesn't exist, especially if you're launching into a new therapy areas, becomes critical. And that has to be part of the consideration set. Another element of just building on strategically is that start consuming this as services, as a software, because every time you face a situation like this where you're launching a new asset in the marketplace and you're thinking about building it from scratch, you

21:32just don't have the time or the ability to absorb all the innovation happening elsewhere. So consider the services as a software, which is the AI agents are just accelerating that concept very quickly. And the last thing I would say is that discrete and portfolio-specific focus, that has yielded results like you can deliver two to three impact at a much lower price point because the question always comes back as, well, I don't have the budget, I don't want to increase my budget to support these functions. But the idea is what you have, you could actually be very differential about the value you generate

22:02for the asset you're doing. So that's one area where I've become very focused on the portfolio lifecycle and the product assets we are supporting indifferentially. And second, I'll mention in patient services unique in this regard, Stephanie, where it's always a mix of services we outsource, the pharma typically outsource to the operators and others and rely on nurses and borrowed contract sales, borrowed field reimbursement officers and things like that. So there's always this mix of what we call services, which we already relied on third-party partners to provide us within the patient services mix and what we have retained because

22:36we thought that was a tragic value. But irrespective of the strategy they're following in this area, you should definitely make provisions for owning the data. Because if you own the data, you fundamentally own the experience, irrespective of what kind of a model have you used in getting the services up and running. And that's a critical aspect. A lot of people think it's outsourced and they forget about the value that outsource service is creating. And going back to the point Kitty made earlier that we can write a contract where they are still providing services on your behalf, but they are feeding the data back into it.

23:09So those will be no regret moves, tragically. I think if you start behaving this way, we feel you can never go wrong. Fundamentally goes back to the purpose, I think Stephanie, you asked, is all about the patient. We care about patient who can benefit from our therapy, should start therapy as soon as they can. So we should do work towards reducing the barriers. We should prevent any kind of a non-clinical reasons why a patient is dropping on therapy, discontinuing on therapy for reasons which we can support by providing patient assistant program or exposing them to some services we as a manufacturer provide.

23:40And we have done multiple math around it. Steph, one thing I would submit is, yes, we cannot talk in terms of return of investment because this is all about patients and the focus is on them getting the best therapy we can. But doing this right thing actually is a very good business model, both sustainable economic model, but also the right thing to do for the patients and the providers. No, that's a great point that you made at the end because it can be both, right? And I think, you know, great insights from both of you and great guidance that people can kind of take back and start working on now as we move forward.

24:14I know we can't predict the future, but I do think we can look at trends. So I'd love to close out with looking at, you know, two to three years out. What's one thing that you think will be fundamentally different and one capability that nobody can afford to skip? Maybe start with you, Ashish. Perfect. Now, Steph, I love this blue sky thinking question, but to be very candid, this is what excites us. And I think this is the future worth building. My belief is patient support, patient services genuinely becomes anticipatory.

24:45We have mechanisms where AI can sense a risk. They can predict a risk. It can trigger a right intervention before patients and the physician who are treating these patients, they hit the barrier. And that's the best service we can provide to the both patients and the provider in long run, because fundamentally they manage the disease best for better. So that's one hope, promise of future. What remains non-negotiable, I think I keep underlining this throughout this podcast, you will see a theme, is that stay compliant, stay connected, make the data ready for consumption

25:19with AI and intervention we are putting in the marketplace with the acute eye to the trust we established into these services. That's a future worth building. Excellent. Kirti? Yeah, no, I absolutely agree. I think something that was already covered was just the patient relationship and the data behind it, right? That will essentially be owned by the patients. And, you know, manufacturers will have a lot more control. They'll be great custodians of the data with built-in consent and, you know, making sure that the data is not scattered across different vendors. And it's equally important to be able to measure and prove impact, right?

25:53So be it through causer, test control, so on and so forth. But unless and until you measure the impact, I feel that it'll be difficult to defend some of the investments and, you know, keep scaling. So that is essentially one more guidance that I would think would be necessary. Very exciting. There's so much to look forward to in the future, and I think so much that can be done. So I really appreciate both of you for sharing your insights today. Thank you to Zias and to Ashish and to Kirti for such an excellent conversation. And thank you to all of you for joining us on this episode of The Top Line.

26:23I am your host, Stephanie Beller, and that's The Bottom Line from The Top Line.