PODCAST: Point-of-Care in 2025: Lateral Flow Grew Up. Where Will the Technology Go Next?

Mitzi Rettinger: I'm Mitzi Rettinger, Chief Revenue Officer at DCN Diagnostics, and this is the Expert Insights podcast. Today we're talking about point of care diagnostics in 2025 using Pat Vaughan's recent piece, "Point-of-Care in 2025: Lateral Flow Grew Up. Where Will the Technology Go Next?" Pat is DCN's Chief Operating Officer and works across assay development, engineering, clinical research, regulatory strategy and manufacturing. Pat, welcome.

Patrick Vaughan: Thank you Mitzi. Great to be here.

Mitzi: You've sat close to both the technical and operational sides of rapid test development. What's your role at DCN and what kinds of decisions do you spend most of your time on?

Pat: Well, my role here, I'm the Chief Operating Officer at DCN. So that entails basically makes me responsible for all, all sorts of development programs through all the stages, right through to manufacturing and also our engineering component as well. Both our design engineering and our LFA reader application work as well. As for decisions, I guess I make them all day. But an interesting question now. You forced me to think about what they are, but the decisions really, it's everything from company strategy, basically some personnel obviously as well. Right. With a great but large team here. And then I suppose maybe I do get involved in some projects at a very top level where direction is needed or where our senior scientists say, well, you know, I'd like your input on this or whatever it is when they have to maybe make a decision as well. And they like the kind of the to create that discussion forum to decide which approach. Or maybe a project is reaching a critical point where you have to make some strategic decision point or input what direction to go involving the client as well. And then, of course, the decisions involving resourcing and costing a program, especially talking to new clients. But and then, you know, the decision on resourcing in general for us being a, you know, a customer service and a contract organization, obviously that's always a very important balance with respect to resourcing and having the capability to execute the projects as they come.

Mitzi: You mentioned when you were talking about resourcing and for, you know, helping out clients if a new client comes in. What is kind of your involvement with client proposals or how are you helping as far as decision making with a client?

Pat: Actually it varies, but it can be quite a lot. Okay. So because, you know, many of our clients, they have a great idea of a product. It could be their first delve into a diagnostic product. They may have therapeutic product. They may be coming towards the diagnostic. There are many decisions to be made and sometimes are a lot at a time. It's not a matter of oh, do I want to develop a product or not? It's more what type? Does it involve? A reader is a quantitative qualitative. There's a whole host of questions that a lot of the time I do need to walk through those with a client. And we do that for many reasons. One is sometimes they don't come from a diagnostic industry, even when they do. These are some very complex questions and important to get right. And rather, we're not going to necessarily make a decision and give the time saying this is a take it or leave that. We never do that. So we like to walk them through the decision process and help them ultimately make the decision, but give them all the maybe the pros and cons or the ups and downs of doing it one way versus the other to help them get to a better place, and ultimately for them to feel comfortable with the path that we've selected to go ahead with.

Mitzi: Yeah. So supporting them in the decision making process. But for you, it's almost like you have to step into their shoes and say, what decisions, what I need to make for this rapid test development and then guiding them I, I like that.

Pat: Yeah. Yeah, that's a very good way of putting it. I do put myself in their shoes basically, and see what would I do in that scenario or what would I like to hear from somebody that is supporting me to help me make that decision?

Mitzi: Early in the article, you open with a reference to the Advanced Lateral Flow Conference 2025 talk titled "LFA is Dead. Long Live LFA." I'm curious what you mean by that phrase, and more generally, by that argument that lateral flow is now an integrated system and not just a strip. And what in your own work led you to that view?

Pat: So "LFA is Dead. Long Live LFA." I suppose the premise there is really that LFA has been around a long time, right? For those of us that are familiar with it, it's not news, right? We've been using it and understand it and so forth. But for a huge part of the market out there and the population of potential diagnostic users or developers or manufacturers. There are some preconceived ideas, right? So that it's oh, is it just another pregnancy test or it's just a cheap test? And for some people, just because it's cheap means it can't be that good. For us, cheap means cost effective, right? But high performance. Right? So we know that. So it's more. Well LFA is dead. That's in some people's mind. Well that's just old news. But really it's not because there is and continues to be some huge potential there. And obviously there's enormous capability. And then if you think about it, since the pandemic now everybody knows what lateral flow strip is. Right. So pre 2019 or 2020, you know there was only a small proportion of the world really knew what lateral flow was. Yeah sure. For pregnancy or maybe HIV and malaria and maybe a little bit maybe here in the US at least for other respiratory diseases like flu or RSV, but for the masses around the world, not many people knew what it was, but now they do. So I think people are starting to see and realize, wow, this is something very user friendly and very cost effective that we can use. In our home or in our office or wherever you need to use it. So it's really like it's coming of age. And therefore the tagline then of long live LFA, because there's a lot more that it can do. You know, there's really not a lot of other technologies that you can really take to a point of care with that simplicity. And at that cost point, even from molecular and molecular is going point of care. But it's it comes with a heck of a, a higher price tag on that as well. And, you know, it's, it's a little bit more complicated. It can be prone to, to contamination and so forth as well. So really lateral flow it can rule the point of care market in one sense because of its simplicity. One how user friendly it can be with good design of course, and then how cost effective. And then all of that is now we can digitize everything as well. So it's not just a binary or is yes or no or it's a oh, I saw it and I threw it in the trash bin. So or I think it was positive or I think it was negative. And now you have a fundamental record of that result of or so.

Mitzi: Is that what you mean by it's now an integrated system? When you're saying that you can actually digitize, it's not just yes or no or maybe elaborate a little bit on the integrated system piece.

Pat: Yeah, it can be standalone, of course. And there are many, many uses for where you take it. You read it, you run it, you read it and that's it. But yes, it can also be digitized by use of your phone, a smartphone or dedicated readers where you can record the result, but not just record a yes or no. You can actually quantity at the intensity of the test line as well versus a calibration card. So you can actually get concentrations of an analyte. So in many instances for some health markers you always have them. But it's the important piece of the information is how much of them are in your bloodstream or in your urine sample or so forth. So that's where the quantitation becomes important. And then that it's paramount. Then obviously that you use a reader or smartphone and basically you're digitizing your result.

Mitzi: You describe lateral flow moving from strips and lines to engineered systems. So this is elaborating a little bit on what you were just talking about. If a team is defining a new LFA development program, what belongs in the initial design inputs beyond the chemistry and the strip layout? I mean, what should the team be thinking about?

Pat: I guess what I mean by moving from strips and lines to engineering, part of it is changing the way we think about that and changing the way we think about and talk about LFA. Because if we tell somebody walking down the street, here's an LFA, it's cheap, it's easy to use, but they think it's cheap and oh, it can't be very good. We've already, you know, preconditioned them to think, oh, it can't do much. We don't just stripe a line. We are engineering an assay basically. And I think and I'll delve more into that in a second. But I think maybe the way we proclaim it is very, very important as well comes in because it is it really is high tech. So an LFA, sure. It looks simple. But we put an awful lot of effort into the development of the LFA to make it simple and easy to use for the end user. It's very expensive to make an easily used and cost effective test. Right? And that's because you are really engineering the assay and ultimately then engineering a result as the output from that assay. But also you're also engineering a manufacturing process. And not all tests are the same. Just because you've developed and manufactured one test does not mean you can develop and manufacture a second test. They are completely different. Every test is different. Every analyte, every set of antibodies, they are all different. So what worked for one does not mean that it will work for the second. So that's where it truly is high tech. There's more goes into it than just striping a line, or spraying a conjugate and basically slapping it all together. That's not how it works. In fact, it won't work that way. Mitzi, we often talk to customers as well and said they start by saying, well, we stripe down the antibody and we got nothing or we got everything. We got a big, strong line, but it was there for the positives and negatives. And we're saying we said, yep, that's exactly what happens when you do that. It's more than just, again, it's more than striping a line. It's actually you have to engineer the assay and ultimately engineer the result. And then when you're doing that, obviously as you're setting out to design the test, the team needs to think about who's the end user and how will they use it and how they perceive the test. Those things are going to be critical as you start a program for the team to think about.

Mitzi: So from the very beginning, that's what you're saying. Come with that information. Minimal. Well. When I listened to everything you just described, it's almost felt like it's more about how does the community talk about lateral flow assays in general, and that we need to educate the world on what these can do or the capability that we have beyond something that would be considered simple. So it's more about how we talk about it.

Pat: Yeah, absolutely. If we talk it down, they're going to perceive it as not very good. So we need to talk it up because we are telling the truth. It is high tech and it has a lot of precision and performance criteria that are critical as we move to decentralized medicine for, for example. So that is important. And we need to educate people as to how effective these assays can be.

Mitzi: I'm kind of putting some of these pieces together that you were just talking about, and you just threw in a decentralized setting as well. So let's make this concrete. Let's imagine a seed stage team comes to DCN with a working benchtop lateral flow prototype that the founders developed, say, as part of a graduate research program at a university. They want a reader assisted test for a decentralized setting. So without getting into any client specifics, how do you walk them through defining this program? And what are some of those first decisions that have to be locked down?

Pat: Okay. Well, first off, you know, we always and then obviously these are the questions we're going to ask that university team is what are the design specifications? What are the original inputs that you used to get the assay as far as you're showing us today? Okay. So did you do all that before starting to develop the test? Right. That gives us a target as to what the test needs to do or needs to be. It needs to achieve. It obviously may not be there yet. All right. Obviously they've come to us for some sort of guidance and help us to get it to stage further. But at least we know the endpoint. So as I said, that gives us the target. We take a look at the obviously the assay itself. And the one thing obviously you said, okay, so they want a reader. The first thing we look at is the background right. Does it have a clear background in one sense right. A reader is like your eye. It's kind of binary. Do you see something or not? And then it'll obviously can measure the intensity of what you're seeing. Obviously background is very, very important. If you or I look at a test and it's got kind of a, you know, a pink or a dirty background. We're looking at it with our eye, but our brain is computing what the test line is, kind of as with respect to oh, that's a strong test line. That's a weak one. Irrespective. It's kind of like, well, I'll discount the background. You put that into a reader. It's going to see what's the difference in signal between the background and the test line intensity. So it's making it binary in that way. And I'm talking about it very simplistically obviously with, you know, you can create a lot of tools with AI and with respect to how you understand that a little bit more. But here I'm trying to explain it to a group of students. To put it in very simple terms, no matter what you're doing, whether you're using a reader or not, you should have a clear background, right? You know, you go to these shows, trade shows and stuff. People are showing your tests and stuff, and you look at some of the data, you look at some of the pictures and you see that the background is horrible and some of those to me, that's not a viable test. So at DCN and if we're developing a test or giving people guidance on a test is, you know, the objective is a clear background with a clearly distinct test line or test lines, basically. So that's priority. So then if you start taking a look at the reader, then we have to assess, well if it hasn't gone too far, well do you need a fluorescent reader or a visual reader and why. You know, I suppose. Why do you need a reader first? Right. If you're using fluorescence, obviously you have to use a reader. If you're using visual, you maybe you don't. But if you want a quantitative, then you do need to use a reader or, you know, a smartphone with a specific app that can calculate and versus a calibration curve. Going back to the first conversation, we get them to lock down basically the specifications. What sensitivity are you looking for? What specificity does it need to have a certain precision at a certain concentration? Those are locked down. You can't deviate that. That's the target that I spoke about. And then the other thing that's probably from a university in a research mode and saying, oh, let's get our hands on all these antibodies and antigens and start creating a test. What we tell our clients, any client, whether they're from an academic background or commercial, is what's your regulatory strategy? Where are you going with this test? Who's going to use it? How are they going to use it, and how are you going to get it regulated to be to be allowed to put it on the market? And I think that's a critical teaching point as well, because it all comes together. If you have your inputs, your specifications, and you have a regulatory strategy, you have the blueprint to move the project forward. And apart from whether antibodies are available or they work properly or whatever it is. You're setting the foundation for a successful project. That's the difference between a university project for fun versus are you truly developing a product that needs to be commercial?

Mitzi: So kind of coming with, you know, if I think of a requirements like these, you know, I'm going to hand you my product requirements and then you're going to tell me, is this feasible? Possible, and what is it going to take to do it? If I hear you, you know, we need the product specifications. We need to know who's going to use it, where they're going to use it, and if they have a regulatory strategy or some path forward to market early in to help you in developing the test moving forward.

Pat: Yeah, absolutely. It gives you the structure to know what to do basically. So it's not just about conjugating an antibody or striping it on a membrane. And let's see what happens. There needs to be a purpose towards it. Otherwise you can you can stripe and spray for all your work. And it's not going to get you anywhere. You need to have a direction and you need to have that target that I spoke about.

Mitzi: The team throws in they want are considering biodegradable or biocompostable or plastic free. How does that impact all of this?

Pat: That's a question that's being asked more and more. The answer is. It could have some impact, but maybe it has none. And what I mean by that is it's the same principle in designing the assay. You may have some restrictions in what you have to select, so you may be forced. So let's say I have of a certain type of material I have ten options out there on the shelf, but I can only use five of them because the other five have some plastic component in them or some adhesive or something like that. So but your principles are the same. One thing I would point out is that the biodegradable or biocompostable materials can have different heating and shrink rates. So you actually when you design the mold for them, you actually need to do it for that material, not for artwork. For a plastic like a polystyrene or whatever you're using. You may have to alter your mold to make sure you get the correct sized part out the other end. So that is a very important decision that you will have to make up front. But other than that, most of these materials that are now becoming available are as effective and perform in the same way. Obviously, you have to be very careful and do all your usual studies especially. I would recommend doing paying close attention to stability studies because all materials, whether they have plastic or not, kind of gas off. And all of those things can impact your reagents and can impact the shelf life of those and the performance of those as well. So those are always very important. And then I suppose one thing that I always call out when people start talking about biodegradable is I prefer to think about having biocompostable material. Right. Because the difference comes in that some material can be bio degraded, but it degrades it down to microplastics. Right. That's not plastic free then. So that's why it must be I prefer the target to be biocompatible, because that then truly is sustainable.

Mitzi: Thank you for that. That was insightful. So now I have a two part question. Part one what factors drive your choice and reader technically and from a regulatory or deployment standpoint? And secondly, when does that decision need to be made to incorporate the reader to avoid any further rework in the development pipeline? Is there like a decision point?

Pat: Yeah. Okay. So I might I might kind of mix the answers a little bit because they're kind of intertwined as well. So. So thinking about the idea of thinking about a reader later, you got to still make sure that you've already designed a robust and clean test. Remember I spoke earlier about having a clear background that's paramount no matter what you're doing, if you have noisy tests with background, that's only going to lead to a lot of other problems. Whether you're looking at it by eye or you're using a reader. Okay, so that's fundamental. A reader won't let you overcome a poorly developed test. It's kind of garbage in, garbage out. Basically. Right. So but then with the choice of reader, I mean, there are obviously some the fundamental decision to be made is, is going to be reading visual output or absorbance, reflectance, or is it looking at a fluorescent output. And that'll be driven by the assay. Right. Primarily dependent on the analyte concentration. The lower the analyte concentration you may be moving more towards needing extra sensitivity. So therefore you have to go the fluorescent route. Whereas if your analyte is present in higher concentrations, you may be absolutely fine using a visual label and therefore a visual reader, which also then means you could potentially use a smartphone, right? So if it's fluorescence, you have to use a dedicated reader, at least for now, until you know there's some way of putting a, you know, a very usable, you know, attachment onto your phone down the line, which I'm sure will come in due course as well. But obviously with a visual label, you can potentially use your smartphone. Now, there's a whole science behind that as well. It's not just a matter of clicking a picture and hey presto, it's done. You need a validated device, but you need a consistent and reproducible way of capturing that image and maybe analyzing it as well. So there's a little more to it than that. Actually, one point that I want to make as well is that I know Mitzi, you and I have been approached by this as well, is that people have this idea that, oh, fluorescence is going to be way more expensive. And that is simply not the case. I would say if anybody is telling you it's going to be more expensive to develop, it's more expensive to make and manufacture and all of this. It's nonsense basically. Right. You know, there's practically no difference between the cost of a visual label versus the fluorescent label. The instrument is going to be the reader is going to be the same, more or less. It's just different lenses or LEDs in there. And it's the same effort to develop it anyway. So there shouldn't be any. So that should not be holding you up. And that should not be a deciding factor with respect to the cost. And then obviously with readers as well is be very conscious. Just like for the assay. What's the endpoint. What's the regulatory strategy for your system. Both the assay and reader. And if anybody is following the latest guidance documents and everything that's coming through, cybersecurity is a huge commitment to make. Now, you know, obviously for your product, it's a huge part of what goes into that development. So don't underestimate what that will take and evolve. And then if you say you have a reader, oh, I'll add it later, but I'll go to the market first. Remember, if you're adding a reader later, assuming that it's completely compatible with your with your test, you're going to have to do your both your analytical and clinical validations all over again because it's de facto a different test. It's now a system. So. Think carefully. Do you want to add it now and do it once or do it later? Maybe seed the market with the visual test that you're looking at or whatever you're doing, but just think about it with respect to can you co-develop the reader along with the assay and get it all out there once and do one set of analytical and one set of clinical validations?

Mitzi: I was recently having a conversation with Dan and Emily about this topic, and we were talking from the regulatory side and some of the regulatory changes with the addition of the PCCPs and how really incorporating that. The earlier you do that, the better off you are in your submission process. So I think what you're saying here really aligns with that. It'll allow you to be able to make modifications to your device without having to resubmit early in.

Pat: Yeah, those types of decisions are not new to point of care and lateral flow, even outside of determining whether it's a reader or not. We make those decisions all the time, you know, do you scale up to a high throughput system or do you leave it at a batch level? Do you do it now? Do you do it later? Do it later. You have to redo some validations as well. So obviously there's a timing and cost factor that's significant to all of those. But that's the discussion that we have with our clients. And we lay it out one way or the other, whether it's scaling up to a 5,000 unit batch or a 50,000 unit batch, or whether you're adding a reader or not. We walk our clients through all of that to help them make the best decision.

Mitzi: You know, in a lot of the conversations I've had with clients, they tend to underestimate the effort when they move from their early prototype to a regulated product. How do you explain that process to them and what is on your. This will take longer than you think. Checklist.

Pat: Well, one thing is when you're developing a product, you also have to develop the manufacturing process for that product. Okay. So just because you've done one test and you moved it across. Yes. And you went from a benchtop deposition to reel to reel deposition just because you've done it once. You still have to think through that process for and that manufacturing process for that new product. Because does that is it a quantitative test? Does it have a calibration curve? When am I going to develop the calibration curve? Where in that process is that done? How do I integrate it later as well? So there's a lot of things that you have to think about. You're better off thinking about it up front and plan, you know. But maybe you don't implement things until later. But you know, you need to have had that discussion or conversation up front and have a plan saying, okay, I'm definitively not going to do that now. This is my plan, and I'll do it later. At least you're aware of what's coming and you know you've made the best decision for your situation, be it from a timeline or cost perspective.

Mitzi: So define for me the difference in that prototype and product. You know, a lot of times, you know, hey, I've done this in the lab. I've done it a couple times, I've got this, you know, let's manufacture it. And then they're like, whoa, why does that? You just told me we've got like another 7 to 9 months before I'm going to have this product. How do you explain that to them? Why is that so important that timeline in between?

Pat: Well, you must have to listen to me a lot, obviously, because you use the two words that I always describe the process as is prototype and product. And there's a huge difference between it. So prototype is I've got it to work. The test seems to work. I got a test line. It's reacting. There's a dose response to the analyte or I found antibodies or something commercially. And I'm kind of fairly close to the sensitivity level that I want. For instance. Great. It works kind of once it's a few weeks or a few months work and hey presto, that's fine. But remember, a product is you need to make it over and over again in thousands or hundreds of thousands of devices over and over, with different lots of raw materials coming in, different people moving in and out of the manufacturing process and stuff. So it's very different than a dedicated scientist giving their whole attention to this. I'm going to say 24 over seven, but, you know, at least 8 or 9 hours a day. And probably when they go home, they're thinking about it as well. Right. As to making sure when they do that experiment that it works. Right. And the experiment in that instance could be just running the test. Remember, the end user isn't going to be running an experiment. Or the manufacturing people are not going to be running an experiment. They want to take the SOPs and say, okay, every time I execute this, I'm going to place the orders for all the reagents and materials, get them in, make sure they meet specification, but they are different lot numbers from the manufacturers. They need to be confident when they put all those together. Every time that the proper product with the meeting, the correct specifications pops out the other end. And that's the difference, right? So after feasibility or early prototyping, it's a prototype. You make another one. It might be slightly different because you haven't made it robust enough to overcome all the tolerances, you know, the differences in the materials and one batch of antibody from another. You're tweaking it basically, right? You're still in experimentation mode. Whereas at the end of your optimization development, when you're validating it, you're validating now a product because it truly is. It's something that you can make reproducible. It's going to be consistent and it's actually manufacturable But it doesn't mean that you get a device out. It means that you can give it to your manufacturing group, and they can make it consistently time after time after time. And there's a big difference. And they're the words I use. There's a there's a big gulf between prototype and product. They mean two completely different things. You might test a person with a prototype and a product and get the same result. Yes, but if you do it many times, you may not get it with the prototype, you will get it with the product. That's why there's a significant effort goes into what we call phase two or phase three of a program where you move from feasibility, concept and feasibility into true optimization and development. As I said earlier as well, you're developing that product, but you're also developing the process to make that product.

Mitzi: In DCN's Basic Lateral Flow Training course. I've heard some of the attendees comment at the end, when they talk about what surprised them. In learning at a high level what you've just described, what goes into that entire timeline of going from feasibility all the way to manufacturing? And I think one of the biggest thing that surprised me was they had no idea all the variables and the different things that are moved through that process and what they have to consider, that's even beyond just the strip with the lines and the cassette, but even the inside of the cassette and the sampling and the all of those different components beyond what they and their mind had already thought about and how it can impact things over time. So very insightful. You cite Rapid Molecular Point of Care as a clear example of what same visit can look like for lateral flow teams. How should that influence performance targets and the user experience expectations?

Pat: I guess what my comment is really talking about here is that in recent times, in more recent times, the turnaround time for executing a molecular tests, say at a doctor's office or in a clinic or whatever it is, has shortened considerably. That's a major step forward. But we need to remember is guess what? LFA has been doing that for many, many years. LFA is already there, right? But what I mean then is the molecular diagnostics. When you're looking at a target that has nucleic acid, right. Let's face it, it has to have nucleic acid. It is very sensitive and can be more sensitive than LFA. And I guess what I'm saying is that, well, that's upping the bar and can we push LFA closer? So it's an encouragement for developers to try and get that sensitivity up there, to get it closer to the molecular result. Again, anybody that's been developing respiratory tests in the last in the last decade, really, because the FDA changed its guidance, many before the pandemic, actually for flu as well, is where it started to compare your immunoassay LFA to a molecular test as the predicate device. So it upped the bar. So I'm more saying let's look at the bar has shifted up. Let's get LFA better okay. It's already meeting the turnaround time because it is very rapid and it's still very easy to use. So the other thing as well is just because a molecular test can be more sensitive in many instances you don't need that extra sensitivity potentially. Right. So there's more than enough sensitivity in an LFA as well. So always remember what's the clinical outcome that you're looking for. And what's the clinically relevant concentration. Is that the most sensitive to test can be or is that the most sensitive that the test needs to be. And maybe then there's a perfectly adequate and high performing LFA is the solution that has your quick turnaround already, but at a much better price point as well, a lot more cost effective solution. So it's really kind of a I was kind of almost playing with the, with the words there as well as that. It's upping the bar to keep improving it. But in one sense, LFA is already at that bar of a fast turnaround at the point of care or at the point of need.

Mitzi: In the section Connected Diagnostics. In plain terms, you argue that results should be born digital and move as structured data. What should teams decide early so connectivity does not become a problem?

Pat: Okay, so apart from the quality of the test itself and how it's manufactured, let's assume we all got that locked down and it's all good. Now you have to think about how is your reader set up. In some instances you have to think how does our little reader on the bench going to talk to the hospital infrastructure, for instance? So how is the data going to go out? How does it receive input or information or instruction basically as well. What's that crosstalk between the two systems. And obviously the security between that is your instrument or reader set up to talk in a secure fashion in addition to all its safety certifications. It's not going to go on fire or it's not going to interfere with some other Bluetooth or other device in the hospital lab, for instance. And then obviously with just security. But like the cybersecurity, as I said, I mentioned that's a fairly onerous but very important aspect of that connectivity today as well. So all of that requires a lot of planning. This is not accomplished by a late night in the lab or a late night coming up with some software code, and hey presto, in the morning is ready to go. It takes months of coding, testing, iterating and planning side by side with the assay being developed. And, you know, walking through, how is this all going to work? How is what's the workflows and stuff like that as well? This is not a kind of a something that's imagined in if you are or is it? It takes a lot of planning and then obviously the simple thing as well is that. Is your reader talking the same language that is going to be used by the, you know, the hospital limb system or and so forth as well. So a lot of considerations. It's not it really isn't kind of a plug and play. You plug it in and something magical will happen. That takes a lot of effort and a lot of planning.

Mitzi: I feel like we can't have a conversation about a current and future state of diagnostics without talking about AI. In your article, you write that AI enabled interpretation must remain a regulated medical device, not an ongoing science experiment. What does that require operationally?

Pat: What I'm saying is, like AI in one sense, right, is iterative and it's constantly maybe improving something. Imagine you have and I'm not saying you should. I'm just saying let's say you had AI active and this is a non computer nerd talking here right. So let's say it's active in your reader. It's learning and it's starting to change things. Well the test that you validated in the first place. But this AI brain in there is computing and changing it. And maybe after weeks of work and weeks of data being imported and analyzed and coming through, it's suddenly changing. How do you know you have the same test, right? That, to me, is an experiment. Okay unless it's done in a very controlled way. Right now, I think the most valuable application of AI is in the researcher's hands. Or, you know, the scientist with the software engineer side by side, using it as a tool to develop better algorithms so that in the end result, with the test and the reader or the smartphone, it's fixed. Right? So, you know, it's not an experiment. You know exactly what it's going to do when it receives something. But you've used all the power of AI in that image analysis and computing and coming up with very clever algorithms that give you the best possible result and then take that, lock that down and put it into a reader into a validated system, basically. Now you can think about it as well. Well, that's in the background while we're still giving the, you know, computing it in the same way, let's capture all that data and see if we can improve the algorithms in the future, but do it in a stepwise. So take that data and bring it back into the lab, so to speak, and walk up and say, how can I improve this test even more? Revalidate it and put it back out onto the onto your install base as well. So do it in stepwise controlled ways rather than you know, it's kind of like letting the reader run amok on the marketplace and make its own decisions, decisions, kind of thing like that. So maybe I'm personalizing the reader a little bit more than I should, but keeping it structured, but really using its power in the hands of the experts, basically. And, you know, we're always told, well, think of AI as an intern and do that, use it as an intern to help you develop and create the tools that you're going to use and then validate it, put it out there and then iterate on that multiple times.

Mitzi: Last year. In your 2025 predictions article, you called out digital integration, sustainability and regulatory trends. Was there one prediction that surprised you in how fast or how slow it's moved and why?

Pat: I thought you might ask me something like that. And I think no matter how I'll answer it, I'll start contradicting myself halfway, halfway through the answer. Right. So it's kind of yes or no. I think those are the big movements that happened for sure. So as you said, digital integration, sustainability and regulatory. What moves the fastest? Probably digital integration. I think you see more and more of it. Can it move faster? Maybe. Ultimately, yes. Now, surprisingly, there's still not many, if any, smartphone applications approved, let's say with the FDA. There were a few. I think there's 1 or 2 at the moment. There's there were several more or a handful more during the EUA or, you know, emergency use authorization during COVID, but they've been removed from the market. So I think we'll probably see more of that. So that's maybe a little bit slower, but maybe not surprising because it's very difficult because, you know, so many things change with a smartphone, right? So the carrier updates the software, the manufacturer updates the software and stuff like that. So how do you how do you ensure you always got a validated device. And, you know, Mitzi, you and I talk a lot of customers about ideas for products. So not on the market yet. So we can we have a kind of a bird's eye view into what the future holds. And we know that there's a lot of that coming. But on the other side and maybe what has been slower, maybe the whole there was a lot of talk about sustainability, but maybe it hasn't moved as much. I think maybe there was a time a few years ago where these sustainable materials were way more expensive, and I think people might still think maybe most of them are. Some of them are, but I think there are a lot more cost effective options. So I think maybe we can see a little bit of picking up the pace on that in the future as well kind of thing that's so it's kind of there's a bit of I was surprised and not surprised by several things.

Mitzi: Pat, thanks for walking through this. For listeners who want more detail. Pat's article is titled "Point-of-Care in 2025: Lateral Flow Grew Up. Where Will the Technology Go Next?" And it's available on DCNDx.com. Pat thanks again.

Pat: Thank you Mitzi. It's a pleasure as always, and I'm excited for 2026. We're living in very exciting times that are rich in change, but obviously rich in opportunity as well. So let's see where 2026 takes us.

Mitzi: If you're developing a rapid test and you want to talk through any of what we've discussed today, you can reach out to us at DCNDx.com/contact. Thanks for joining us for this episode of Expert Insights.