Decoding Sepsis: Rethinking How We Detect Critical Illness with Tim Sweeney, MD, PhD
about the episode
Sepsis remains one of medicine's most difficult diagnostic challenges, where every hour can impact patient outcomes. In this episode of Innovate & Elevate, Sharon Kedar sits down with Dr. Tim Sweeney, Co-Founder and CEO of Inflammatix, to explore how advances in artificial intelligence, host-response diagnostics, and precision medicine could transform the way clinicians detect infection and critical illness. Dr. Sweeney shares the personal experience that launched his two-decade mission to improve sepsis care, explains why traditional diagnostic approaches often fall short, and discusses how Inflammatix's TriVerity™ platform is helping physicians identify the right patients for the right treatment at the right time.
This Episode Is For You If:
- You want to understand why sepsis remains one of healthcare's most challenging diagnoses.
- You're interested in how artificial intelligence is being applied to real-world clinical decision-making.
- You want to learn how innovation can improve outcomes for critically ill patients.
Jessica Federer
About tim sweeney, md, phd
Tim Sweeney, MD, PhD, is co-founder and CEO of Inflammatix, a clinical-stage startup bringing precision medicine to hospital care. He led the company from its Stanford roots through product development, clinical validation, FDA clearance, and commercialization. His background spans clinical medicine (surgery), data science, and translational diagnostics, giving him a practical view of both clinical need and market adoption. Tim is named on more than a dozen patents, has published more than 100 manuscripts and abstracts, and has served as PI on multiple development contracts from DOD/DARPA, BARDA, and NIH. He lives with his wife and 3 boys in Northern California.
Episode Outline
(00:00) The Patient That Changed Everything
(01:57) Why Sepsis Is So Difficult to Solve
(03:10) What Sepsis Really Is
(05:59) Why Early Diagnosis Matters
(08:18) Introducing TriVerity
(12:17) A Life Saved by Earlier Detection
(17:10) The Science of Host Response
(19:07) How AI Interprets the Immune System
(25:22) The Future of Precision Critical Care(31:28) Improving Care Beyond Saving Lives
-
Sharon Kedar CFA: Please welcome Dr. Tim Sweeney to the podcast. Tim Sweeney is co-founder and CEO of Inflammatix and helped invent the company's core technology. He trained as a physician and surgeon and became frustrated with the limitations of existing diagnostic tools for infection and sepsis. Can you just tell us a little bit about when you decided sepsis was something you wanted to take on?
Sharon Kedar, CFA: It's not a small life goal to try to impact sepsis.
Tim Sweeney, MD: I still remember the case. I was in medical school at Duke, and I had been assigned to an ICU rotation.
There was a young man. He'd been in the backwoods. His ATV had rolled over. It crushed his chest, and he was in the ICU on a ventilator, and he had developed a ventilator-associated pneumonia, basically an infection in his lungs, which we actually had treated with antibiotics, but it had kicked off this really bad cycle of inflammation.
And so he was getting sicker and sicker and sicker just from having this really strong inflammatory response in his body. And we pulled out all the stops. He was on a machine to help him with his breathing. He was on medicines to support his blood pressure...
But he kept on getting sicker. And I had gotten to know his wife, and, you know, he was my age. I really cared about him. And I remember going to my attending and said, "You know, this guy's getting sicker. What are we gonna do?" And he said, "Well, there's kind of nothing else beyond supportive care.
We just have to cross our fingers." And I said, " We're one of the best ICUs in the world, and this guy is getting sicker. There has to be something." And he said, "All we have left is prayer." And I really remember that moment of feeling like, wow, there's nothing we have.
Why is that? And, you know, in the end, that young man lived, but it kicked off a program of study, and I've been studying sepsis and critical illness really for the last twenty years. Because I think, in the end, it really struck me that this is a chance to save a young person's life.
This isn't end-of-life care. This isn't one or two more years. If you can get somebody through an ICU stay, they can go on for fifty happy, productive years, and that really seems something worth devoting a career to.
Sharon Kedar, CFA: When you embarked on that, and we'll get into what is sepsis, what did people around you say when you said you were gonna make an impact here? 'Cause it really is a big goal to say like, "I'm gonna try to change this."
Tim Sweeney, MD: Well, they say go after a big, hairy, audacious goal, and sepsis certainly is one of those. It is the number one most expensive diagnosis in medicine, right? It's 50-plus billion dollars a year for Medicare alone. The number one killer of people in hospitals.
It also has been a disease that's resisted any attempts to really modify it. There have been over 100 drug trials in sepsis with zero successes. There have been a number of diagnostics that were launched supposedly for sepsis that failed on the market. It has been one of the most challenging areas, I think, in medicine.
And so to your point, it's not something easy to go after, but, well, if it was easy, it would've been done already.
Sharon Kedar, CFA: You said something that a lot of people do not know. Sepsis is the number one killer of people in hospitals. Now, to be clear, if people are sick, people should absolutely go to the hospital.
Can you talk a little bit about why it's the number one killer, and really just if someone is hearing that for the first time, just share a little context on that, 'cause it's a scary statistic.
Tim Sweeney, MD: Yeah. And you know, this is a good point. It's not that you go to a hospital and get sepsis from the hospital. I think the point more is, when folks die not at home, right? When they, when they die in a hospital, there's only a couple of things that tend to be outside of the hospital's ability to rescue them.
You know, a massive blood clot is one. Sometimes we're not great at spotting those, and they go to your heart and that's that. Severe heart attacks can be another. But sepsis is one of these final common pathways, like an infection in a person with cancer. And the person's already weak, and they go to the ICU maybe for treatment of that infection. And what people don't necessarily understand is actually sepsis is not really the infection itself. Sepsis is the body's over-ramped inflammatory response to an infection. And so even if you've given someone antibiotics, even if you've controlled the infection, that leftover inflammation can be what kills somebody.
The way to think about it is this: everybody when they get an infection has an immune response, right? We get a fever. The body's immune system fights that infection. And then when it's done, that immune system gradually subsides, right? The inflammation stops. It's a feedback cycle. And the problem in sepsis is it becomes a feed-forward cycle, where inflammation starts to cause damage, which causes more inflammation.
And that feed-forward cycle can get out of control. We know in theory that anti-inflammatory medicines should be able to stop that. And yet again, despite over 100 drug trials, none has actually been shown to reduce mortality because the disease is so complex. And so that's one of the things we're working on at Inflammatix is to bring clarity where there's currently a lot of complexity to actually enable new medicines to save lives.
Sharon Kedar, CFA: Okay. So is most sepsis hospital acquired?
Tim Sweeney, MD: About 80% of people who have sepsis in a hospital show up in the ER with an infection. Maybe with sepsis or maybe with an infection that's going to get worse and worse during that stay. So a much, much smaller number might acquire sepsis in the hospital, but perhaps they go in for a major surgery and a few days later get a postoperative infection, and it gets worse.
But the vast majority are people who've gotten infected at home and then shown up at the hospital, and they just get worse while they're there. And so to your point, there's really two aspects of saving lives in sepsis that people focus on. One is, how do you prevent a, "just an infection" from becoming one that's life-threatening?
How do you get antibiotics to the right person, identify that they're really sick, get them prompt treatment to avoid them ever getting critically ill? The other is taking that person who is critically ill, maybe they showed up late in their disease course and there was nothing we could do, or maybe for whatever reason, we missed those early symptoms, and in addition to treating the infection, now we have to find a way to treat that critical inflammation.
And so both are aspects of ongoing efforts to really reduce mortality in sepsis.
Sharon Kedar, CFA: Okay. And I'm sure you know this article from AAMC from 2023. So first of all, it says sepsis is difficult to diagnose, and it says with sepsis, time is of the essence. For every hour of delayed treatment, the risk of death increases between 4 to 9%.
And then it says
making
a diagnosis can be challenging because the symptoms of early sepsis can be similar to other serious conditions, including heart attack or stroke. So can you just speak to that? I mean, as a human and as a layperson, that's a little terrifying.
Tim Sweeney, MD: People talk sometimes about sepsis like it's finding a needle in a haystack, right? And that's true to some degree. The reality is if you've ever treated patients in an emergency room, you're looking at a stack of needles, right? People are sick for some reason, and you're trying to figure out what reason, and in addition, how sick they're going to become, right?
So it's actually not as simple as everybody that comes through we need to figure out is this sepsis or not. And I think actually that focus of, is it sepsis or not, is one of the reasons why prior diagnostics haven't really worked. When I go see somebody and they have a cough and a fever or a rash with a high heart rate or just general symptoms of not feeling well, sepsis is one possibility, but so is a normal mild infection.
Maybe they need antibiotics, but they're safe to go home. Or maybe it's something viral, or maybe it's something non-infectious entirely, right? Maybe this is a mild heart attack that's kicked off a little bit of inflammation, or maybe it's pancreatitis or something non-infectious, right?
All of these things can present very similar ways. And so what we really need is tools not just that say, "Hey, this is possible sepsis," but really instead, what is the cause of illness? What's making this patient sick? Is it an infection? If it's an infection, what kind? But also it could be something non-infectious.
And then of course, how sick is this patient going to become? We can look right now at a patient and see who is critically ill right now. That is obvious, right?
What we're very bad at is determining who is on a trajectory of being really sick tomorrow. And so the big risk is that you see somebody, you give them treatment, maybe they go home, but at home they get much, much sicker. And now they've come back and you sort of missed your window for early intervention.
And so both are a difficult part of the current standard of care, and that's one of the reasons we developed our product, TriVerity, was to separately address both of those issues.
Sharon Kedar, CFA: Can you talk about what is TriVerity? How have you approached sepsis diagnostics?
Tim Sweeney, MD: For a long time, people have viewed diagnostics for sepsis as a problem of infectious disease. Let's see if we can find a pathogen. And in particular, many of the audience may be familiar with the term blood poisoning, and a lot of people feel like they've heard that sepsis is an infection of the blood.
And that's sometimes true. Maybe about half of cases of sepsis have an infection that has actually penetrated the bloodstream, and we can catch it there. But for about half of cases of sepsis, the body has walled off an infection, and it hasn't penetrated the bloodstream. And more to the point, if you looked at people that aren't all the way into sepsis yet, just somebody who's in the emergency department who has a bacterial infection, the rate of bloodstream infections in that group is maybe about one in 10.
They have an infection, it needs antibiotics, but it's limited to the lung or to the skin and soft tissues or the belly. And so the traditional approach of we're going to look for pathogens isn't that useful. And it's not to say it's not helpful in the long run if you happen to find a bug, but to a physician seeing a patient right now in the emergency department, I have to make a decision about whether or not to put you on antibiotics, whether or not you come into the hospital, and all those pathogen tests, that's all a later question.
So I need help right now seeing you, determining what's making you sick and how sick you're going to be. I was at Stanford in surgical residency, seeing patients, and started doing some work looking at the body's inflammatory response using what we call transcriptomics, which is a study of how the different cells in the body turn on and off their genes, basically what they're reacting to, what networks of activation are going on.
And we found that by looking at the patterns of activation of different genes in white blood cells, we could determine not only what the white blood cell was reacting to: is it infectious or not? If it's infectious, is it bacterial or viral? But it turns out we also could predict, in this inflammatory response, who was going to get really, really sick in the coming days.
And so TriVerity is a product that we have been developing for several years. We got an FDA breakthrough device designation for the product in 2023, and now it's actually on the market with FDA clearance. This is a 30-minute blood test that uses AI to interpret patterns of inflammation by looking at patterns of gene expression from a patient's blood.
So basically, we can figure out if an infection is present. If so, what kind is it, and how sick this patient is likely to be, all in 30 minutes, pretty close to the point of care. And so it's really about changing the way we initially diagnose and treat patients with infections. And I'm happy to tell you some stories, but it's been really rewarding now to actually have that out on the market and see the effect it has in saving lives.
It really is making a big difference.
Sharon Kedar, CFA: This is astounding in a positive way because most times, these ideas fail. And one thing about you, Tim, is that knowing you since 2018, your persistence, your hyperfocus, it just is inspiring to other people who are going to try to take on what previously couldn't be taken on.
I always say that for innovation, there's a lot of naysayers until it really happens. But now what's cool about where you are and where Inflammatix is, is we're starting to see lives saved. Can you talk about, if someone shows up at the ER, what does it look like if a hospital has TriVerity versus not?
The reality of the toolkit that doctors have today without TriVerity is if you show up sick at the ER to know whether you have a bacterial or a viral infection, and therefore whether you need antibiotics or not, it's a clinical assessment. While doctors are great, they're human. So can you just talk about what the present state is and what it looks like with this thirty-minute blood test with definitive answers?
Tim Sweeney, MD: I 100% agree that there, in the end, is a human story in the emergency department, right? Both on the side of the patient and the patient's caregivers and the physician. Let me give you a couple of patient stories that I think highlight the value of getting these calls right.
Sharon Kedar, CFA: Great.
Tim Sweeney, MD: About a month ago at one of our sites, we had a gentleman. He had come into the emergency department twice, both times complaining of what we call malaise. Basically, he just didn't feel well.
He said, "I really don't feel well. Something's wrong." And both times the emergency department assessed him, but pretty normal vital signs. They ran some basic laboratory tests. He had a normal white blood cell count. He had a normal lactate. There was really nothing indicating that he was particularly sick.
And so twice, they told him, "You know, this is probably something that'll pass. Maybe it's viral. Go home and wait and let us know if it gets worse." Well, he came back a third time and he said, "I really don't feel well." And they said, "Well, maybe we should try this new test, TriVerity, and see if it can help."
It's called TriVerity because it produces three separate scores. So it's actually not a test that says: sepsis versus non, or something like that. It's really meant to provide a holistic view of the patient. So the three scores of TriVerity answer, is this likely to be a bacterial infection or not?
And separately, is this likely to be a viral infection or not? Because of course, some people can have a bacterial viral co-infection, and lots of people have neither. There's something non-infectious going on. And then you have a third important score, which is, from any cause, is this patient likely to end up needing ICU-level care in the next seven days?
So those are the three scores. So you come back to this patient. Twice had shown up, twice normal labs, ran the TriVerity test. TriVerity test said there's a very high likelihood that this patient has a bacterial infection, and he's actually at high risk of ending up in the ICU in the next few days. And the physicians there thought, wow, that's, that's very different.
So they admitted him, drew some blood cultures, began resuscitation. Well, over the next few days, his blood grew out a bacteria called listeria. And in fact, he had listeria meningitis. He had an infection that had descended into his brain. And he had a good outcome. They started him on antibiotics promptly and he survived.
But listeria meningitis carries about a ninety percent mortality if left untreated. Again, every other test said nothing's obviously wrong. But by looking at the very earliest patterns of what the immune system was beginning to react to, we could spot, hey, there is a bacterial infection underlying this.
And by the way, this patient's immune response is not a healthy one where he's gonna get better soon. This is someone who's really getting sicker.
Sharon Kedar, CFA: That is such a huge deal because when we talked about this in the early days, the whole idea was saving lives. And now you've got multiple stories of lives being saved. And just for the lay audience, knowing if you have something bacterial versus viral is really important because if it's viral, that is not typically where antibiotics are indicated. But if it's bacterial, and especially if it's bacterial sepsis, the sooner you get on the right antibiotics, which I know that's a whole thing, the more likelihood of saving a patient's life.
So can you just touch on that for a moment, just so that folks understand?
Tim Sweeney, MD: It's actually even a little bit broader than that. So to your point, yes, bacterial or viral, only the bacterial needs antibiotics. And antibiotics can be life-saving. They can both prevent progression to sepsis or in somebody with this really severe infection can save a life. TriVerity adds an extra dimension, though, right?
It's not just bacterial, viral, it's also what if this was something that was neither of those? What if it was something non-infectious? And then adds this risk stratification piece. We had another story that sort of illustrates the point, and this was very early.
It was one of the very first patients where TriVerity was used. We were at one of our very first sites, and a gentleman came in. He was in his 60s. He had a little bit of shortness of breath. He had a mild fever, and he had told the physicians that he had some sick contacts at home. So they were really thinking, "Oh, this kinda sounds like a pneumonia."
Ordered a basic white blood cell count. They ordered TriVerity, to your point, because they thought, "Gosh, we'd really love to know, is this bacterial or viral?" But the TriVerity test came back saying this is non-bacterial and it's non-viral, but the patient is at very high risk of ending up in the ICU.
Non-bacterial and non-viral. What could it be? So they broadened their workup, and they started to look for a heart attack. They sent an EKG. They drew a troponin. Turned out he was having a mild heart attack, and the doctor at the time said, "Thank God for your test. I'd have been treating his heart attack with antibiotics."
And so really the idea is people show up, and it's not a needle in a haystack. Our goal is not just to identify sepsis. It's to make sense of a haystack of needles and try to figure out which of many diagnoses could this patient have and get them very quickly onto the right path. And again, in that case, that gentleman was sent right to the coronary care unit, the heart attack unit in the hospital, and he had a good outcome.
But if he'd been sitting on the wards with extra fluids and antibiotics and everything else, he actually could have had a much worse outcome over the coming days
Sharon Kedar, CFA: Tim, can you talk about the science of host response? So what exactly underlies the TriVerity test in terms of, how are you reading the immune system, and then also how are we overlaying AI?
Tim Sweeney, MD: When our body first has an insult, white blood cells go to the location of that insult. And that's true whether you've, fallen and scraped your arm or you've got a bacterial or viral infection, right? There's an insult there, white blood cells go to the site. And that early immune response, we call it the innate immune response, and it happens over a couple of days.
That's how we get an early fever. We establish a response to something. Now, many of you are probably familiar with the later immune response. After you've had a virus for a few days, your body begins to make antibodies, fights the virus, and you get better. But that happens over a matter of, you know, 5-7 days.
Early on, the way we're able to fight bacteria and fight viruses is very broad tools. These tools have evolved in our bodies over hundreds of millions of years. So for instance, when you first get a virus, your body can't tell, is it a COVID virus or a flu virus?
It reacts to the fact that both of those viruses happen to have a form of molecule called double-stranded RNA. Double-stranded RNA only exists in viruses. It's not in a single mammalian cell. And so if it senses any double-stranded RNA, it says, " That's a virus," and starts this antiviral program.
There's very, very similar responses to the types of proteins we see in bacteria or the types of damage we see in certain kinds of cell tissue. And each one of those kicks off a different response in the white blood cells. They start turning on different pathways: that's an antiviral pathway or an antibacterial pathway.
And functionally, what we're doing is we took a few of those genes that are involved in those various pathways, and we measure them. And by looking at that pattern of activation, we can use pattern recognition to say, "This is the pattern that corresponds to viruses. This is the pattern that corresponds to bacteria," et cetera.
And so that's the AI piece. We look at actually 29 different markers direct from blood in 30 minutes, and we apply this AI to say both what is the cause of this immune response and also then, of course, how sick is this patient likely to be? Is this immune response likely to lead to critical illness?
And that's taken a really long time. We've gathered thousands and thousands of samples from dozens of clinical studies from patients literally around the globe to really make sure that we had a broad representation of this patient mix. Because people that come into the ER, they're old, they're young, they have symptoms at home for one day, they have symptoms at home for 10 days.
They come in critically ill, they come in with the sniffles. It's just the broadest cross-section of humanity you can imagine. And so we need a test that works very broadly in that group. And so that was the intent behind how to build something like this and why the use of AI is so useful, is to create something that generalizes across all of those kinds of presentations to provide trustworthy information.
Sharon Kedar, CFA: But when you initially set out here, was it 29 markers or did that evolve?
Tim Sweeney, MD: When we were still at Stanford, we actually published a series of three peer-reviewed manuscripts. We were the first in the world to show that small sets of genes could still be diagnostic in, what we call, validation cohorts. So we'd learn a signature in one set of patients, and then we'd go in a new set of patients from a totally different setting and test it and show, "Oh, look, it still works."
And there were 11 genes that were part of infection versus not. There were seven genes that were part of bacterial versus viral, and there were a different 11 genes that were part of severity prediction, and that was the 29. Over time, some of the genes changed to make better targets, and of course, with thousands and thousands of more patients.
And we want the best set of genes that works together, and that's where the hard computation comes in. It's not just that you take the genes that are most associated with viral infection because the reality is many of them actually are telling you the exact same thing.
What you want is to create a network that across multiple patients works together to boost your diagnostic signal, and that becomes what machine learning scientists call a non-convex problem. There's no guaranteed solution. So we have to use a lot of computing power and a lot of special tricks to come up with the best signature that's the most stable across the most patients.
Sharon Kedar, CFA: So then it's not an AI overlay. It's that AI is an integral part of the TriVerity engine. It's basically how we built this.
Tim Sweeney, MD: Inflammatix... actually even the word Inflammatix was a combination of inflammation informatics. We are a data science company at the core, and the product is the algorithm, and we happened to need to build the world's highest multiplex, fastest point-of-care test system to deliver that algorithm.
But in the end, yes, it is AI native at its core.
Sharon Kedar, CFA: Tim, I don't think I've ever asked a question like this, but do we know how many lives we've saved already?
Tim Sweeney, MD: We have been on the market for a little over a year with the test actually being used. And what I can say is that at every site we've gone to, when we talk to the physicians involved, all of them have anecdotes about: here's a life that you saved last week. My sense is that there are a number of people that are walking around today, even within this first year of launch, that without the test probably would have been missed.
And the interesting thing actually is that even at the sites that do have the test, literally this morning we talked to a site where the CMO said, "Recently somebody came through our emergency department. We missed whatever was going on, and that patient was dead three days later. And I bet that if we had used your test, we would have spotted that patient and he'd still be alive."
And so we continue to drive more usage at these sites to make sure it's appropriate. You don't want to over test people, but also that we want to catch these folks earlier in their course of disease and get them onto the right treatment pathway.
Sharon Kedar, CFA: This story has such great sort of first 10 years, next 10 saving lives. It really has, for a long time, reminded me of a company that Mike and I invested in called Intuitive Surgical.
So at the time, I was on the phone with a couple doctors, and the doctors actually said, from the expert networks, "We would never use a robot. How can a robot ever be better than an actual surgeon?
The Da Vinci robotic surgery instrument's got to the point where if you were a hospital, patients were basically saying, "Do you have the robot?" And you were perceived not to be a with-it hospital if you didn't have the robot. And so there was obviously the impact on care, but then there was this huge patient pull, a little bit like Intel Inside, right?
So how do we get to that point?
Tim Sweeney, MD: Well, I think there's two kind of trends that Intuitive followed that we hope to do the same. One, to your point, is just the fact of robotic surgery in especially prostatectomies, the evidence became overwhelming for that initial kind of beachhead where this was something that was necessary.
And the other, of course, was more innovation over time to expand the product suite and to do more offerings across more folks. So on the one hand, we are sitting on a lot of data from sites who've used the test who either have completed studies or are going to publish soon. And I think those studies showing the beneficial outcomes will continue to create pressure and I'd expect to have thousands of patients in our registry, for instance, soon.
And that kind of data makes a big difference. Anecdotes are really important. For hospitals, they want to see the clinical and economic benefit in numbers. And I'll just say we have those data and they are looking good. And as hospitals continue to sign up, that's one of the reasons they're signing up is not just for the anecdotes of saving lives, but for, "Oh, wow, we can free up bed days and we can save money and we can enable stewardship programs and all those things."
The other thing I'd say is around the kind of the new product offering. Our intent is to go up the chain of severity with the patient, right? Right now, TriVerity's very focused in the emergency department. How do we find these patients that may be overlooked today that in fact really need critical care and get them the treatment that they need?
We talked a little bit at the start, though. Some people show up critically ill. Or even if you know they're going to be critically ill and you start prompt treatment, they still progress through that disease. And we kind of come back to that first patient at Duke, the one where no medicines were available.
Our next goal is to begin to focus on bringing medicines to the ICU, and we're doing that following almost like a playbook that kind of was established in cancer. So think about this, right? In the 1970s, a woman with breast cancer had cancer of the breast. That's what we knew. And every breast cancer was treated the same.
They got a modified radical mastectomy, and they got really damaging chemotherapy. They got platins [platinum-based chemotherapy]. And survival was pretty bad. And over time, what changed was a molecular understanding of the disease. We understood eventually PR and ER receptor status, and that brought tamoxifen to bear. And a couple decades later, we understood HER2/neu status, and that brought Herceptin.
And in the 2000s , there was the advent of what we call precision oncology with MammaPrint and with Oncotype, and they brought with them first through estimating severity, but eventually identifying subclasses of breast cancer. And now if you go to an oncologist and you have breast cancer, it's immediately sent for molecular fingerprinting, and we select one of ten or more new treatments.
And mortality has fallen substantially because we really understand breast cancer is not one disease, it's several diseases that all happen to show up in the breast. ICU treatment is like cancer in the 1970s. We talk about you like you're in the ICU because of an infection or because of a car accident or because of major surgery.
But what keeps people in the ICU is this concept of critical inflammation. And it turns out we've shown, we published a major study in Nature Medicine last year, there are actually different kinds of critical inflammation, and we can identify them. And we've shown that those kinds of inflammation respond differently to different immunotherapies.
And so in the next phase of the company, it's really about now coming into the ICU and saying, "Well, if treating all patients with sepsis with an anti-inflammatory doesn't work, maybe that's because there are different subtypes of sepsis, and each of them requires a different drug." And that's the journey of the next decade.
Sharon Kedar, CFA: Are you saying that the drugs are not in the ICU right now?
Tim Sweeney, MD: 20 years ago when there was no drug to treat that young man who'd been crushed by an ATV, the ICU 20 years on is the same. We still don't have any drugs. If that gentleman came into an ICU today, his care would be the same. He would get on the ventilator, drugs to support his blood pressure, maybe replace his kidney function, and then we would cross our fingers and pray.
And in 2026, that's not enough. More money is spent in ICU care, $250 billion a year in the US, than in all of oncology, and yet we haven't had a single new medicine for those ICU patients. So the goal really is to drive precision medicine into the hospitals, to the emergency department first with TriVerity, and then moving continually up so that we can actually save the life of someone who comes in and is in this cycle.
And the amazing thing is, it really does seem to work.
Sharon Kedar, CFA: Yeah.
Tim Sweeney, MD: It's not a small task, but neither was changing the way we diagnose infections in the first place.
Sharon Kedar, CFA: Do you foresee a day where patients can go to inflammatix.com and see what hospitals have the TriVerity test? 'Cause it's an interesting time where, because of social media, because of people using ChatGPT and Perplexity Health, patients, whether the AI is hallucinating or not, they're bringing information, factual or not, to the table.
Tim Sweeney, MD: It never hurts for patients to walk into their ER doc and say, "Oh gosh, do you have this test?" I do think that it's pretty clear that the workup in an ER will be different in 10 years than it is today.
It's going to be. There are enough new technologies on the market that someone will come in, they will get a host response profile, one where we've looked at the immune response. We can say, "Is this an infection? How sick is this patient?" That will engender next generation follow-up testing. If we say, "By the way, there's a bacterial infection," then you should go on to get cultures.
We should try to find that bacterial; figure out if it's something drug-resistant, what antibiotics help. And for those that we identify as being critically ill, that they should go ahead into that next stage of precision medicine of getting that immune response profiled, picking a different drug moving forward.
In the '70s, you didn't need a tumor board because there was only one treatment available. Right now, you wouldn't go to an oncologist who wasn't part of the tumor board to select from among all the best therapies to personalize care for you. And I would imagine someday it'll be the same to go to an ICU, that there's an intensive care board, and there are a number of different modalities and treatments, and we can really understand what's going on with this patient.
And I think it'll bend the mortality curve in the same way we've seen with cancer.
Sharon Kedar, CFA: As we sit here in 2026, it's a really interesting time because we're talking about saving lives, and what we're talking about is the humanity in the ER. I used to think it's like, oh, the ER is this place that's just gonna solve a lot of stuff, and it's a wonderful place for a lot of things.
But it's basically this massive triage of, we don't want the patients to die. But there is so much room for error that, you know... making a clinical assessment that could be right or wrong in terms of human life.
I mean, obviously that's their job, and thank you to all the doctors who do that. But you must get a lot of gratitude from them.
Tim Sweeney, MD: We've had a couple of webinars recently where some folks have been able to share their experiences. And to your point, everyone is grateful for better care. The reason we go through so many years of training and do a hard job in medicine is to help people. That's in the end the root of it.
And the ER is a really tough place to practice. There are a lot of things that come in quickly, and our goal in the ER is to stabilize, make a best next management decision, and then move on to the next patient. And on top of that, as you've probably heard, emergency departments right now are really in crisis nationally.
Extremely high levels of boarding, where a patient has to sit in the ER for hours and hours or sometimes days and days. Cuts to certain kinds of insurance have made more people seek primary care in the ER. So there's a lot to wade through there.
And there's also a lot of data that shows that practice across emergency physicians is very difficult to standardize. The rate of who gets admitted to the hospital is extremely different. Almost twofold different between kind of the top admitter and the bottom admitter with no real difference in mortality.
Sharon Kedar, CFA: Wow, really?
Tim Sweeney, MD: It's a hard job, yeah. And
Saving lives is important. Literally, one of the very first patients that ever got the test was in one of our very first studies. It was even pre-clearance, but they were allowing folks to treat. And there's a 25-year-old young woman, and she was brought in by ambulance because she felt she was so sick.
She was shaking with chills at home, and she came in, her heart rate was super high. It was really obvious that she had a bad UTI. And each hospital has a sepsis alert, so the sepsis alert had gone off, and she obviously needed antibiotics, and she got them.
But they ran a TriVerity test because after they had given her a little bit of resuscitation, some fluid, she looked better, and she really didn't want to come into the hospital, and she wanted to go home. She doesn't want to stay in the hospital for two days. It's expensive. It takes her away from her family.
And in fact, the TriVerity test said, this is clearly a bacterial infection, but actually the severity score is low. This is a patient who's unlikely to get worse. She's young. And so they talked to the patient, and working with her, were able to say, "You know, look, here's some oral antibiotics. You can take them by mouth.
You can go home. Give us a call if you get worse." And she did fine. And sometimes it's not just about saving a life, it's about just doing right by the people that are there. You don't want to have someone come into the hospital unnecessarily. You can't sleep. You get poked with things, et cetera, et cetera.
So I think it really is just about improving the appropriateness of care for people across the spectrum, and it's one of the things that I think really drives home the value of the kind of test in the ER.
Sharon Kedar, CFA: Yeah. I mean, a UTI, if she had to stay for a couple days, I'm sure she wouldn't be thrilled about that. That's a really hot button topic right now in midlife. So I think it's such a great example. What has surprised you most on the journey?
Tim Sweeney, MD: I know this is gonna sound maybe a little odd, but a founder's role is to speak a vision into the world. I think that there's this opportunity to really change the way we do this. And what has surprised me is how many amazing people have joined in that journey. There are employees and investors and clinical partners and now hospitals implementing the test and say, "Yes, I agree." I'm so grateful for all of those partnerships because that's who made the test, not me.
Sharon Kedar, CFA: I love that answer because it's, to me, most great innovation always starts with that. It's like the naysayers, you gotta sort of scrape, beg, and plead to get the time, and they're like, "Well, I'm not sure. I'm diagnosing pretty well." And then it just starts to become this snowball or flywheel.
Okay, and here's my last one. If you were not a medical doctor or CEO and co-founder of Inflammatix, just like totally different profession, what would it be?
Tim Sweeney, MD: It's a good one. Well, so two things. One is, I would love to just be a dad right now. My boys are 4, 6 and 9, and we have a great time together. But if I had had it all to do over again, I'm sure I would've gone into something in the hard sciences. One that always fascinated me was this brain computer interface that's taking over now. I am constantly amazed by the progress that humanity makes and being on that technological frontier; I think that's what I really enjoy.
Sharon Kedar, CFA: What do the boys think you do?
Tim Sweeney, MD: So a long time ago, we actually needed to do some testing of the test at altitude, and I live up in the High Sierras [Sierra Nevada]. So they've seen the device, they have a dim sense that we build tools for doctors to use. I think they have a sense that what I try to do is make care better for folks everywhere.
Sharon Kedar, CFA: That's awesome. Kids always have the best responses to stuff. Thank you, Tim. Appreciate you joining today.
Tim Sweeney, MD: Thank you so much. This has been an absolute delight. I really appreciate the opportunity to be a guest on your podcast.
Additional Resources
Inflammatix: https://inflammatix.com
About Your Host
Sharon Kedar, CFA, is Co-Founder of Northpond Ventures. Northpond is a multi-billion-dollar science-driven venture capital firm with a portfolio of 60+ companies, along with key academic partnerships at Harvard’s Wyss Institute, MIT’s School of Engineering, and Stanford School of Medicine. Prior to Northpond, Sharon spent 15 years at Sands Capital, where she became their first Chief Financial Officer. Assets under management grew from $1.5 billion to $50 billion over her tenure, achieving more than 30x growth. Sharon is the co-author of two personal finance books for women. Sharon has an MBA from Harvard Business School, a B.A. in Economics from Rice University, and is a CFA charterholder. She lives in the Washington, DC area with her husband, Greg, and their three kids.
CONNECT WITH SHARON
Connect with Sharon on LinkedIn: Sharon Kedar
Follow with Sharon on Instagram: @sharonkedarcfa
Learn more about Innovate and Elevate innovateandelevatepodcast.com
Join the newsletter to receive the latest episodes in your inbox: Innovate and Elevate Newsletter
Rate, Review, & Follow on Apple Podcasts
Did you love this episode? Let us know by rating and reviewing the show on Apple Podcasts. It’s easy - click this link, scroll to the bottom of the page, and select “Write a Review.” Let us know what you liked best about the episode. While you’re at it, consider following Innovate and Elevate. Follow now!
This podcast is produced by Brave Moon Podcasts.
Please note: The content shared in this episode is for informational purposes only and does not constitute medical, financial, or investment advice. Please seek guidance from your own qualified professionals before making decisions.
FAQs
What is sepsis and why is early diagnosis so important?
Sepsis is the body's overwhelming inflammatory response to an infection. Early diagnosis is critical because delayed treatment can significantly increase the risk of severe complications, organ failure, and death. This Innovate & Elevate episode with Dr. Tim Sweeney explains why sepsis is difficult to identify and how new diagnostic technologies may help physicians intervene sooner.
How does artificial intelligence help diagnose infections and critical illness?
Dr. Tim Sweeney explains how Inflammatix uses AI to analyze patterns within the body's immune response. By evaluating multiple biomarkers simultaneously, AI can help identify whether a patient is experiencing a bacterial infection, viral infection, or another cause of illness while also predicting the risk of clinical deterioration.
What is host-response testing in healthcare?
Host-response testing measures how a patient's immune system reacts to illness rather than looking directly for a pathogen. In this conversation, Dr. Tim Sweeney discusses how host-response diagnostics may provide clinicians with earlier and more actionable information when evaluating patients with suspected infection.
What is TriVerity and how does it work?
TriVerity is an FDA-cleared diagnostic platform developed by Inflammatix. The test analyzes gene-expression patterns from a blood sample and provides insights into infection status and illness severity in approximately 30 minutes, helping clinicians make faster treatment decisions.
Which Innovate & Elevate podcast episode explores the future of sepsis diagnosis and precision medicine?
The Innovate & Elevate episode "Decoding Sepsis: Rethinking How We Detect Critical Illness with Tim Sweeney, MD, PhD" explores how AI, host-response testing, and precision medicine are reshaping infection diagnosis, emergency medicine, and critical care. The conversation also examines how these innovations could help save lives and improve patient outcomes.