Faster Data, Faster Cures with Joe Lengfellner at IgniteData

[00:00:10] Hello and welcome to the Nurse Rosa's INsights. Today joining me is Joe Lengfellner. He is the Chief Product and Technology Officer at IgniteData, where he leads product innovation focused on making hospital-to-sponsor clinical data flow fast, reliable, and research-ready.

[00:00:31] He joined IgniteData after more than 15 years at Memorial Sloan Kettering Cancer Center, most recently as Senior Director of Clinical Research Informatics and Technology, where he led a team of digital health and IT professionals supporting MSK's global clinical research portfolio.

[00:00:52] Before that, he studied psychology at Stony Brook University. So, Joe, I love to see a well-rounded person who is interested in both technology and humanity. So, thank you for joining me today. Yeah, thank you for having me, Rosa. I've had a great time working in the clinical trial space and research for my whole career, and I'm excited to talk about it today.

[00:01:17] Yeah, definitely. And so, when you are looking at clinical trial sites and the sponsors of those trials today, what are some of the biggest technology headaches that they're dealing with? And the part that I really care about, how did those hang-ups trickle down to effective patients?

[00:01:39] Yeah, great question. You know, research sites are managing a lot. Clinical trials, research studies are getting more and more complex. The design of studies, the amount of data that needs to be captured, the regimen that the patients are participating in as part of the clinical trial continues to get more complex. And so, research sites are under quite a bit of pressure to deliver on these clinical trials.

[00:02:07] One of the areas that I saw most acutely when I was at MSK, Memorial Sloan Kettering, was around the data management. MSK is a large oncology research center. There's about 1,800 research studies there. And all these studies require just a massive amount of patient data in order to run.

[00:02:31] And the state of the art up until this point has really been manual data curation. It's research coordinators, research nurses going through health records, going through paper questionnaires, and manually entering this data into research systems at the pharmaceutical sponsors that you can imagine. And we can certainly get into the inefficiencies and quality issues that that really brings up.

[00:03:00] So, you know, I saw that very acutely at MSK. And what really kind of struck me is the research coordinators, the research nurses that were doing this work, you know, they didn't get into research to copy and paste data between systems. They want to work with the patients. They want to find trial opportunities for patients. And so this was just taking them away from that more higher impact work.

[00:03:26] So does Ignite Data really help with identifying people who are potential clinical trial participants? Or is it mostly after they've entered a clinical trial, kind of tracking that data through for their participation? Yeah, it's the latter. So after a patient is enrolled in a clinical trial and the research site starts to collect data for that clinical trial, that's where we come in.

[00:03:54] Ignite Data comes in. And what we're what we're doing is is connecting directly to these clinical source systems, whether they be an electronic health record, an EHR, at a hospital site. There are smaller research sites, community sites that use systems called e-source systems. They're similar to an EHR, but but specifically designed for for research. But again, they're they're capturing this patient clinical data.

[00:04:19] What we do at Ignite is connect to these upstream systems, these source systems and send the data directly to the sponsor. So we don't have humans needing to be the the middleman in moving this. Yeah, exactly. It's it's you know, it's a it's not a rewarding task to copy and paste data to to hand enter data.

[00:04:43] There are much better ways that we can work with these members of the team and not have them just being data clerks. And the program is called Archer, right? Correct. So the the platform is called Archer. What it does is it sits in between these these hospitals, these research sites and the sponsors. And it's connecting to the source systems, extracting the data that is appropriate for the clinical trial, making sure that it's the right data.

[00:05:13] So there there's always a clinical research coordinator putting eyes on the data, making sure, yes, the platform surfaced the correct data. It's ready to move. And then they just at the push of a button, move that data over to the sponsor database, which you can imagine much quicker. We see, you know, 60 to 70 percent increase in efficiency.

[00:05:35] And of course, the data is accurate because we're these are, you know, computer systems speaking to each other rather than having a human having to to key data in. I love how you're bringing up the human error factor that can often be an issue. Like we oftentimes hear people complaining about, you know, AI hallucinating, which it does. But I mean, if we talk about the error rate for humans, it's actually quite high as well.

[00:06:04] So I love that it's adding another level of checks and balances, but hopefully this would be lightening the load for the mission researchers not adding to it. Is that what you've seen so far with the implementation? Yeah, absolutely. I mean, the this the error rate we see with manual data entry is, you know, almost double digits in many cases, close to, you know, eight, nine, 10 percent.

[00:06:31] And so that that adds a lot of burden to the entire research process, not only at the hospital sites with the research teams, but also on the on the sponsor sides that have to identify these errors, reach back out to the study teams at research sites to correct the errors. There's a lot of back and forth, highly burdensome, inefficient process. So when we take that out of the picture, we do free up the research teams to work on higher impact work.

[00:07:01] They can they can work on identifying patients that would be a good fit for the clinical trial. They can work with their their PIs, their investigators to better the patient experiencing in the research study. You know, we these research coordinators, you know, they they want to be working on the science. They want to be working with with patients. They they don't want to be just moving data around all day. Yes. Yes. Contrary to popular belief.

[00:07:30] I think a lot of people imagine researchers as being like more antisocial. But the more researchers I meet and like from the ones that I've had on this podcast, they they're very human motivated from like their interactions with people really do motivate and form. And like that's why they're doing their research is because they they really care about people.

[00:07:54] And they're not just choosing this career to hide from human interaction. So I think you're right there. And you've already got it rolled out at a majority of clinical trial sites. So when something is being rolled out like that in a new location, is there a steep learning curve? Right. Yeah, we have had really great success of expanding our site network.

[00:08:23] And that that's really what starts to move the needle in terms of accelerating drug development and getting more trial opportunities out to patients is when you can you can have that that network of sites that are all using the same interoperability platform. Right. So we have had some some really strong success there. It's not a steep learning curve.

[00:08:43] You know, the the training that we deliver less than 20 minutes just to show our research, the research team, the platform and show them how it how it works. You know, most there there is a standard training program, but most would would be fine with without it because it's it's working within a workflow that they already understands. They're they're launching the system from their EHR, their health record that they're already working in.

[00:09:11] They're they're seeing the research data that they typically see in the in the EHR, except it's just filtered down to the data that they need for the clinical trial. So it's a cleaner view of the data and then they just move it over to the to the sponsor. So it's yeah, it's pretty, pretty seamless. Yes. It sounds like it would be more of a training of what you don't have to do anymore instead of training of this is the new step you have to learn. Yeah, often.

[00:09:37] I mean, I think with a lot of technologies, right, it's more the change management than the technology itself that that requires the training. And it's these research teams have worked in a certain way for their entire careers, which, you know, could be a year or two or could be at MSK. We saw people staying in this career for five, 10, 10 years. And so it's showing them a new way of doing the work, a much less burdensome way to do their work.

[00:10:04] So once they they fit it into their, you know, their their their day and understands, you know, this is this is this is how I do my work now. They're they're, you know, asking for it. They want to see it on on their studies, which is very cool to see. Be like, how can we apply this level of efficiency to the rest of our there's there's plenty of inefficiencies in clinical trials that we can we can continue to work on? Yes. So as I alluded to, Archer has been adopted in a lot of places.

[00:10:34] So where is it at? As of right now, we're recording this in August of 2026. How widely is it being used at trial sites today? Yeah, so we got our start in oncology. So with with cancer, cancer research. And so that that's where the strongest footprint is today. And so really strong acceptance, especially in the comprehensive cancer center space.

[00:10:58] So nine out of the 10 top cancer centers in the country are using the platform in our live. About half overall of the NCI designated cancer centers have implemented the Archer platform. I think it's about 80 percent or so are somehow engaged with with a team and various aspects of implementation.

[00:11:21] So we've had really strong success with that that that that space here in the US and starting to build the global footprints as well. The company was actually founded in the UK and then came over to the United States. So they have a very we have a very strong footprint in the UK as well. Germany, Australia, Australia does a lot more clinical trials. And then I realized when I first started working at Ignite.

[00:11:49] And so we have a strong site network in Australia moving into South America. So, yeah, the footprint is growing rapidly. Awesome. So let's think about this in its impact for the patients. How does moving that trial data faster actually translate into new medicines reaching patients sooner?

[00:12:13] Yeah, I love this question because it is very clear how this this impacts the overall drug development pipeline. A big so you probably know the stats, right? It can take up to 10 years for a new medicine to to receive FDA approval of far, far billions of dollars. It's far too long, far too expensive to get a new drug to to market.

[00:12:37] A big piece of that pipeline and, you know, both from a time perspective, but also a cost perspective is this data management problem that I'm describing. There's various stats out there, but, you know, up to 20 percent of the cost of developing a new therapy is this data management.

[00:12:59] So if we can get better at how we manage data, make it more efficient, make it less time consuming, which the platform does as well. We can expand the capacity of our drug development pipelines. You know, really, I saw this at MSK when I was there.

[00:13:19] You know, 1800 studies sounds like a lot of studies for a hospital to run, but we were we were not serving every patient out there that needed a clinical trial. We were really at a ceiling for the number of studies that we could do just given the sort of administrative burden that comes along with running clinical trials. And so humans could fill out those. Right. Exactly.

[00:13:44] How many humans could fill out the forms, how many humans, you know, how many people we could hire to to do the job. And so when we are able to reduce that burden and run more clinical trials, that's more clinical trial opportunities that that we can present to patients. And we're going to, you know, get to an answer on these clinical trials sooner. Does a new therapy work? Yes or no.

[00:14:08] And either, you know, go go full steam on on that on that that one that works or cut our losses on the ones that that that that don't. And so the work we're doing at Ignite very directly impacts our ability to deliver new new drugs, new treatments to to our patients. And as a nurse, patient privacy is sacred to me.

[00:14:32] And so I know that our listeners feel like that as well and feel like privacy is something that's becoming more and more scarce every minute. So when you're touching this, yeah, how do you keep it protected? Do you have your is that what your guard dog is for that I hear? Sorry about that. Yes, he gets very excited. Yes. So this is one of the things I really enjoy talking about because it's something we take so, so seriously.

[00:15:00] And from from the start, the the Archer platform was built with privacy in mind. You know, one of the things that that makes my job a little easier as on the sort of CTO side of my my role is that we're not persisting any data. So we're not storing any patient data within our infrastructure and within our data centers.

[00:15:23] We are simply extracting and pulling the data from the health record and those those APIs, those those endpoints that exist at the hospital, presenting them in a a view for our research teams to look at. And they are validating that and then pushing it off to the sponsor. Once they push the data, the data is no longer retained in memory. We're not storing any of the data. We're not becoming a data aggregator.

[00:15:50] Some of the other vendors that are doing similar work to us are have, you know, other other initiatives around data aggregation and, you know, commercialization of data. That that's not what what we do. And so from a privacy perspective, we have a very strong stance because we're not we're not persisting any data. You're just transmitting it. Yeah, exactly. It's it's the same. We're transmitting the same data that a research coordinator would be transmitting manually.

[00:16:19] We're just doing it automatically with the the archer platform. Without a fax machine. I know it's a surprise. It's it's and you know this well, right? It's amazing how many fax machines you still see in the in the the workflows. You cannot tell me that things are more secure when they go through a fax machine. You know, this paper is sitting on top of a fax machine. Sitting sitting on somebody's desk.

[00:16:46] If you're telling me that it's going via that fax machine into directly into the computer and it's not going to be sitting on somebody's fax machine as paper. Like, I don't think you've been in some offices. That is happening. I can definitely back you up on on that. And yes, we're certainly held to a higher standard when we implement a new technology. But it's it's all for for for good, good reasons. Oh, yeah, for sure. I'm just mocking the old tech and how it's not necessarily more secure. That's not quite analog.

[00:17:16] So, OK. And you can't have a health care conversation these days without talking about AI. So it sounds like it is AI part of Ignite Data's technology and what role does it play? Yeah, we are, as as many people are, you know, putting a lot of our R&D efforts this year. And this will be on our roadmap for next year as well.

[00:17:40] Looking at how how we responsibly use AI in these these pipelines for a lot of what we do and a lot of what I've described. So far, you actually don't need AI and you don't want AI in it because it's just an added complexity. There's there's a lot of this data that is sitting within health records that is already highly structured data that doesn't require a whole complex data pipeline in order to work with.

[00:18:07] A good example or lab data, vitals data, medication data. It's all very structured data within the in the health record. And that's the data we're moving today at scale where we are starting to utilize AI in various flavors of AI, which we could we could get into is with with highly unstructured data there. And Rosa, you know this well.

[00:18:33] The health record contains lots of clinical notes and pathology reports and radiology reports. And, you know, a lot of hospitals are using ambient technology. Right. And capturing, you know, transcripts and, you know, other summary level data that's just highly unstructured, but also highly valuable for clinical trials.

[00:18:55] And so how do we turn that unstructured data, that conversational data into structured data that we need for research downstream? And that's really where we start to apply AI and AI being a very general term. We can we can use sort of natural language processing, machine learning, some some techniques that have been around for a long time to extract some of this data. And when that's the case, that's what we do, because it's very efficient.

[00:19:23] It's very fast. It's it doesn't cost tokens. And, you know, there's how good it. Yes, exactly. So that that's where we apply a lot of this this tech. Sometimes you do need the more advanced techniques, the small language models, large language models, et cetera. And where that's required, we apply those.

[00:19:41] What's, I think, really exciting about our approach and and unique to Ignite Data as well is we are not using one of the large, you know, third party vendors to send off patient data, write a fancy prompt and have it extract data. We're building very specific pipelines for each data domain that we want to work with.

[00:20:05] We have a an AI science team that is working on this data directly, working with our collaborators at different hospitals to to build to validate these models on real data. And we're able to build these purpose built pipelines for each data domain. And that's a very efficient way to leverage AI. It's also going back to privacy.

[00:20:30] We know, you know, all this all this AI, all these all these pipelines are running on our infrastructure. We know where the data is going. It's not a black box. We know the results, you know, why a model suggests a certain value. We we have all that within our control. And that's a very strong position to be in. Yeah, it sounds a lot safer than it's just floating out there in the Internet.

[00:20:56] So for a research hospital that doesn't happen to have Archer yet and they want to switch to something like that, what questions should they be asking before signing on with a service like that? Yeah, I think a few things. I think you want to look at the technology lift that's required to to implement a platform.

[00:21:19] You know, this our research teams are already stretched thin enough without having to to manage another brand new technology system that they're they're not that they're not familiar with. So something that is more of a configuration out of your EHR or e-source solution rather than a standalone platform, I think, is something to to to look at. Understand what that that tech investment would be.

[00:21:45] I also think platforms that are agnostic to both the kind of source side and the the sponsor side are important. And what do I mean by that? I mean, we I don't think you want a technology that I'm going to use this technology for sponsor a, you know, pharma company a I'm going to use a different one for pharma company B.

[00:22:10] Yeah. And that's that's what we see some of that fragmentation sometimes in this this space where this this middle piece that's that's doing the data movement is is only working with a specific, you know, research database type. I think, you know, I'm certainly biased in this opinion, but a platform like Ignite that is working across all different EHRs, across all different EDC systems.

[00:22:36] That's the sponsor database allows your study teams to work with one solution rather than multiple solutions, depending on the clinical trial, because all these study teams are working across multiple studies and you don't want them jumping between different solutions. I tell you what, I hate jumping from platform to platform like I was at a conference. I can't even remember which conference, but I was passing some vendor booth and they're like, oh, you're a nurse.

[00:23:04] Let me ask you this. If you were creating a new dashboard, what would you want to be on it? And I was like, I would want no new dashboards. I would want you to not have to take me away from wherever I'm already at, because it's hard enough to get everything in. If I change like what dashboard I'm looking at, I'm probably going to forget what I was doing. Because like, have you heard that thing where if you go through a door frame and you forget what you were going in the other room for? Have you heard that?

[00:23:33] I am not, but I get the sense of it. It's neuroscience, though. It was proven. And then if you want to remember, you probably have to walk back through the door frame in order to remember. You should try sometime. Personal experiment. Do your own clinical trial. But for me, I feel like dashboards are like that, too. Like, you know, you have all these different windows open. And when you go from one to the other, you're like, wait, you get distracted by notifications there. And you're like, what did I even come here for?

[00:24:02] So I really appreciate when things are integrated and you don't have to hop from one to the other. Yes, I completely agree. And, you know, this is always a hot topic in the clinical trial space because there have been many studies done or, you know, sort of investigations on the number of platform systems, technologies that a research team needs to use throughout their day. And, you know, the numbers are sort of all over the place.

[00:24:28] But it's at least 8, 10, 12 different systems that they're logging into, all with different, you know, usernames and passwords, of course. Jumping between systems. And so, yeah, I think that that's important. And that's really why our company exists is it's trying to connect those systems together so that we're not giving another platform, but we're pulling the data together from those multiple systems. So we're not adding to the problem. Exactly.

[00:24:56] And I mean, I definitely have personal experience with that, entering in like our clinical data into like the American Heart Association's database. And we have multiple hospitals that we work with. So I would have had to log out of one and log into another hospital, even in the same system, to enter in the data for that different hospital. And they finally updated it where you can just toggle back and forth between facilities without logging out and back in.

[00:25:23] And it was such a relief to finally be able to do that. So I'm really glad. A hundred percent. And, you know, what you're describing is a great parallel to what we're doing just in the clinical trial space. Imagine if you didn't have to enter data into that, those systems, but it was just pulling from the patient's record. That's really another level of efficiency that I can't wait for. Yeah, exactly. And that is what we do. We do see this use case quite often with hospitals.

[00:25:53] Sort of, I would lump them under this idea of a registry, these registry studies where you're following patients, right, for a long period of time. And there's just, there's so much data that these registries require. And it's often not the primary responsibility of whomever is responsible for populating it. And so the data exists, right? And so we shouldn't be able to just push it to where it needs to go. Yes.

[00:26:20] If you want to work with the American Heart Association on streamlining that as one of your efforts, that would be great, too. I would love to learn more about what data is required. Yeah, the program is Get With The Guidelines. And so it uses like heart attack as well as stroke data is a lot of what we look at. So anyway, I wanted to ask you, because there's a lot of buzz around the FDA pushing towards like real-time clinical trials,

[00:26:49] which you kind of may have alluded to earlier when you were talking about how long it takes for drugs to come to market. But for listeners who haven't heard about the idea of real-time clinical trials, what does that mean? And where does your technology fit into that? Yeah. This has been a hot topic. I remember the day where the FDA made this announcement, you know, just all the pings that were going across the company, because this is exactly what we want to see.

[00:27:16] We want to see the FDA standing behind this type of technology. Make it a priority. Yeah, exactly. And making it a priority, bringing people's attention to it. And it, yes, very much aligns with what we do at Ignite Data. It's all about accessing the source data and getting it to where it needs to go in a more timely manner, you know, as close to real-time as possible.

[00:27:44] Where we play a role here is really being that initial ingestion point for the data. So we have access to this EHR data that the FDA is wanting in these real-time clinical trials. What they're really looking at are safety events. So what we would call adverse events in a clinical trial or serious adverse events.

[00:28:09] And so, you know, the FDA can be notified if there is some sort of safety concern with a therapy, right? But also, like, other types of signals like outcomes. Like, is the drug working? Is the clinical trial making a difference in the patient's lives? So when we're able to deliver that data directly to the FDA through, typically through, still through a pharmaceutical sponsor, decisions can be made more quickly.

[00:28:38] Whether, you know, you stop a clinical trial or, you know, maybe you move it up in the approval process more quickly. Because you see, you know, from site A, it's working really well. Site B is working well. So you can bring together this signal more efficiently than you can in the current paradigm. Yeah, exactly. Well, this brings me to my favorite question that I get to ask every guest on the Nurse Rosa's Insights podcast,

[00:29:04] which is, if you were given a grant, Joe, for $1 billion to meet a need you see in healthcare, how would you like to see it used strategically to have the most sustainable impact? Yeah. I mean, my, I thought a lot about that. I knew this question was coming. I've heard your other episodes. I'm like, oh, what's going to be my answer? And, you know, I don't think I'm going to have a surprising answer given what I do.

[00:29:32] But I think it is about interoperability, right? Other industries have figured out how to leverage their data across the world much more effectively than healthcare has, right? Look at the, you know, finance industry. You know, data is shared, you know, financial data is shared across the world in milliseconds. As long as you can log into Plaid, right? Well, right. Well, and they figured it out, right? This is not sponsored by Plaid, but it's not with us to sponsor this episode.

[00:30:02] So that's fine. Well, maybe we need a Plaid for healthcare data because it's such a challenge to get these systems to communicate. So, you know, I wouldn't spend a billion dollars to generate more data or anything. I think we have the data. We have a lot of data. We're getting more and more data every day.

[00:30:20] What we need to do is invest in the capacity to implement these shared kind of platforms to move data, to utilize data across, you know, barriers today. And there's probably a lot of regulatory work to do as well so that people, that we do it in the right way, that people that are comfortable with sharing data in this way.

[00:30:42] But if we invest in that type of infrastructure, you know, then we'll really see the power of the data that we already have today.

[00:31:20] Yeah. It's a political team. Right. Which sounds like it should be so intuitive, but it's so hard. It is so hard. And it's cliche to bring up this example. But when you go to a new doctor and you're filling out the paper forms on a clipboard, that still happens, right? And it's the same data you've already provided to all your other providers in the past. And so it's just, you know, we shouldn't have to work in this way. All this data just needs to. Yes, exactly. Too much redundancies.

[00:31:50] Yes. Well, I definitely agree with you there. And I mean, how about the novelty of this idea is let's make it not cliche anymore. So it's not an endemic problem. Yeah, definitely. Right. It's the fact that we're still joking about fax machines and it's not even a joke, right? It's just a symptom of the world we're living in. Right.

[00:32:18] Because it's like sometimes I feel like I'm living in Star Trek and then other times I feel like I'm living in the episode where they go back to the 1980s. It's like, which one is it? We're just back in the middle. Right. I mean, the clinical trial space is a perfect example of that in health care, right? We're working with these investigators, these doctors, these scientists that are curing disease, curing cancer. Yet we're still using fax machines.

[00:32:44] We're still having research coordinators hand type data into databases that already exist in other systems. So, yes, we can do better. Well, I'm glad you are on the case, Joe. And if someone wants to reach out to you, is LinkedIn the best way to do that? Yes. I'm very active on LinkedIn. So please reach out. I always love to meet new people and especially people interested in this space.

[00:33:11] So, yeah, looking forward to meeting some new people. Fantastic. And I'll have the link to your LinkedIn as well as Ignite Data in the show notes below. For our listeners, if you've enjoyed this episode of the Nurse Rose with Insights podcast, please be sure to like and subscribe wherever you're listening or if you're watching on YouTube. And leave me a comment and let me know what you thought.

[00:33:36] And if you have another answer to the billion dollar question of how we can build a better health care, feel free to shoot it my way because that's what this is. We're looking for answers. And I will see you in the next episode.