Best of LinkedIn: Health Tech CW 32/ 33

00:00:00: This episode is provided by Thomas Allgaier and Frennis, based on the most relevant LinkedIn posts about health tech from CW-Thirty Two and Thirty Three.

00:00:08: Frenis equips HealthTech providers with a market intelligence to identify which hospitals target.

00:00:21: So imagine being in a really severe car accident.

00:00:24: Right?

00:00:25: You wake up, you're completely disoriented and realize your smartwatch has already dispatched an ambulance to your exact GPS coordinates.

00:00:32: That's... I mean that is wild to think about.

00:00:34: It is!

00:00:35: Yeah And today we are diving deep into how health tech is shifting away from those clunky dashboard screens becoming this invisible life-saving safety net.

00:00:45: We're looking at the top health tech trends across our sources and really setting the expectations for you listening, The core narrative comes down to maturity.

00:00:54: Yeah we are watching the entire digital health landscape forcefully mature right now like We're moving past AI as just a shiny pilot program.

00:01:02: Exactly it's rapidly becoming the foundational infrastructure For clinical operations.

00:01:07: And were also seeing market aggressively reject those fragmented point solutions.

00:01:12: They want deep product integration Now.

00:01:15: But the real tension, and what we will spend significant time on later in this discussion is that you can build most seamlessly integrated outcome driving tech In The World.

00:01:25: And it won't matter at all if you trip over the hidden compliance hurdles.

00:01:29: Oh absolutely Because those hurdles are secretly killing major enterprise procurement deals

00:01:34: right now.

00:01:34: It's a brutal reality check for anyone building in B-to-B health tech.

00:01:38: Yeah so okay let us unpack this shift from innovation labs to clinical plumbing.

00:01:44: AI is finally running the workflows that nobody actually sees.

00:01:47: Like, a post from Christian Hein highlights a massive strategic move by ICON PoeCown

00:01:52: Right!

00:01:53: The clinical trial platform.

00:01:54: Yes exactly They have integrated Anthropics Clawed into Orbis which their platform And we should be really clear here this not some lightweight administrative chatbot sitting on side of screen.

00:02:05: Clawd is wired directly to execution at scale.

00:02:07: Yeah, it is running state intelligence real-time enrollment risk detection and even protocol optimization.

00:02:15: And the scale is crazy.

00:02:16: It's across roughly forty thousand two hundred trial participants spanning fifty five countries.

00:02:21: wait

00:02:22: forty thousand Two hundred participants.

00:02:24: yeah that is massive.

00:02:26: Let's look at the mechanism there for a second right?

00:02:28: How does Claude actually optimize our protocol?

00:02:31: I mean i have to assume it reads thousands of past failed clinical trials and flags inclusion criteria that are simply too strict.

00:02:39: Like it might look at a draft protocol on say, you know if You require your patients to have a BMI under twenty-five?

00:02:45: You're gonna fail to recruit enough people in the Midwest.

00:02:48: It's predicting human bottlenecks before single patient is even recruited.

00:02:51: right

00:02:52: but And here's my question isn't the real strategic value for icon The vendor lock-in?

00:02:58: Ah, that is the multi-billion dollar question.

00:03:00: Right

00:03:01: because if I'm a pharmaceutical sponsor and every single protocol decision Every site performance signal in every enrollment pattern Is compounding.

00:03:07: inside Orbis icons platform functions exactly like an operating system.

00:03:11: Yes it gets smarter with every single trial i run.

00:03:14: So the deeper that AI embeds into trial planning The more incredibly painful and expensive It becomes for me to ever switch To a different vendor.

00:03:23: You are hitting on the exact core economic driver there, because Alayna Shvets shared data projecting that the market for AI in clinical trials is going to explode.

00:03:32: It's growing ninefold by twenty forty.

00:03:35: Ninefold?

00:03:36: Wow!

00:03:36: Yeah jumping from two point acid on billion dollars to eighteen point six two billion.

00:03:41: and as a reminder For anyone listening who works adjacent to pharma clinical trials currently eat up fifty to seventy percent of time and budget in the entire drug development process.

00:03:50: Which

00:03:50: is notoriously bloated?

00:03:52: Exactly, AI targets that exact gap.

00:03:54: so data gravity icon-is building essentially an unassailable moat.

00:03:58: That makes total sense.

00:04:00: we're seeing a parallel shift from reactive alerts to predictive infrastructure over at home care too Dr Ben Murth up who shared results with Sarah where they are deploying predictive AI spot health risks in patients days before an ambulance ever needed.

00:04:13: The clinical outcomes they logged are staggering.

00:04:16: Like, in just one single borough last year this predictive monitoring prevented two hundred and ninety-six hospital admissions.

00:04:23: Almost

00:04:23: three hundred admission?

00:04:24: Right!

00:04:25: And it almost halved the number of older adults needing to be moved into permanent care homes...

00:04:29: Which is huge for quality of life.

00:04:32: but let's think about how that actually works on the ground.

00:04:34: because its not a wearable detecting fall right?

00:04:38: No no at all.

00:04:39: Predictive AI is looking at subtle, compounding variables in daily data.

00:04:43: A slight change of mobility may be an alteration sleep patterns or just a minor deviation and heart rate.

00:04:50: It aggregates those tiny signals and flags of deterioration long before it becomes the full-blown crisis.

00:04:56: And the financial math behind preventing that crisis makes you stop on your tracks?

00:05:00: A hospital bed costs the taxpayer over five hundred pounds per day.

00:05:04: Wow!

00:05:04: Yeah...and four daily home care visits cost around fifty pounds.

00:05:08: So for the price of a single night on hospital ward, you can support someone at home an entire week.

00:05:13: That proves that free social care is actually financially viable but only if underlying engine driving it.

00:05:21: I'd offer a bit of a reality check, though based on a Deloitte survey of MedTech executives that JB shared.

00:05:27: There is massive gap between these highly successful deployments and broad market readiness.

00:05:33: Yeah like forty-five percent of the executive's surveyed claim they already have agentic AI meaning AI agents who can take independent action running in production.

00:05:43: See i find that forty five percent figure incredibly hard to believe given how risk-averse healthcare is.

00:05:48: Well,

00:05:49: the fine print validates your skepticism.

00:05:51: only seven percent describe those AI agents as fully embedded in audit ready within their actual workflows.

00:05:57: There it is!

00:05:58: Seven percent right.

00:05:59: and even more concerning Only twenty two percent say they're data environment Is actually ready for this.

00:06:04: like there data isn't broadly searchable It's not routinely reused or consistently documented.

00:06:08: so They

00:06:08: were basically rushing to deploy The shiny new tool without building the infrastructure To support it.

00:06:13: yes And what's fascinating here is that this raises a massive, unresolved governance question.

00:06:20: Deciding to deploy an AI agent is high-level portfolio decision but the accountability for that agent has to live exactly where their workflow happens

00:06:29: on the floor

00:06:30: Exactly!

00:06:31: If they AI makes clinical or operational mistake who owns that outcome?

00:06:37: Who decides when patient case falls outside specific parameters?

00:06:41: the AI was trained.

00:06:43: If accountability sits several layers of management above the actual workflow, it's not real accountability.

00:06:49: It is just corporate documentation

00:06:51: Which is why the philosophy behind how we design these tools matters so much.

00:06:55: Samur Siddiqui, he's a physician who just secured a patent for a clinical message prioritization system for live net health.

00:07:01: He posted about this exact tension The

00:07:03: alert fatigue issue right?

00:07:04: Yes exactly clinicians are absolutely drowning in pages.

00:07:08: lab results automated alerts and chart updates.

00:07:10: They all compete for attention which is a finite highly depletable resource.

00:07:14: So his premise is that AI shouldn't try to replace clinical judgment or act autonomously in a vacuum, it should simply act as triage for data.

00:07:24: It needs to cut through the noise and point the clinician's attention to what actually matters right

00:07:28: now.".

00:07:29: But here is the problem with AI triaging all of that data...it is completely useless if underlying systems do not talk to each other!

00:07:36: Right…the silo problem?

00:07:38: Exactly….if AI is the brain it needs a nervous system.

00:07:42: That is why the market is aggressively killing off fragmented point solutions right now and demanding complete product

00:07:47: integration.".

00:07:48: And Teladoc Health... ...is making a massive public bet on exactly that integration.

00:07:54: Kelly Bliss posted an update on their launch of Teladok One.

00:07:57: It's a unified virtual care model that integrates primary care, mental health & chronic condition management

00:08:04: The whole package.

00:08:05: Yeah they're trying to design around the whole person rather than in specific disease state.

00:08:09: But the real story here is their financial commitment.

00:08:12: Teladoc has putting a hundred percent of their program fees at risk.

00:08:15: Wow,

00:08:15: one hundred percent?

00:08:17: Yes they are tying their economics directly to validated clinical outcomes and proven medical cost savings.

00:08:25: That's fundamental structural shift in business model.

00:08:29: moving from fee for service To fully risking your revenue on outcome.

00:08:33: Right but Is this just a desperate marketing play in a deeply saturated post-pandemic telehealth market?

00:08:42: Or are we looking at the new gold

00:08:43: standard?".

00:08:43: That's fair question.

00:08:44: Because if one of the larger players is willing to put all their money where there clinical outcomes, can smaller digital health vendor even survive next procurement cycle...if they aren't willing do same?

00:08:56: I would argue it absolutely.

00:08:57: the new standard and being driven by commoditization base level intelligence.

00:09:03: Rashmi Kumar offered a really good strategic perspective on this.

00:09:06: As AI Foundation models continue to drop in price and become widely available, the competitive advantage in MedTech completely

00:09:13: shifts... Because AI is cheap now!

00:09:15: Exactly if everyone has access to world-class large language model via an API simply having an AI feature isn't differentiator anymore.

00:09:24: So where does value accrue then?

00:09:26: The winners in the next decade will be companies that can seamlessly integrate their intelligence into physical hardware and complex, high-stakes workflows.

00:09:35: Oh like robotics & imaging?

00:09:37: Yes!

00:09:38: It is about merging AI with advanced imaging, real time surgical navigation AND Robotics.

00:09:44: it's about delivering intelligence to environments where millimeters and milliseconds matter and regulators require intense validation.

00:09:52: That makes a lot of sense.

00:09:53: By the way, if you are listening to this and want to keep tracking how these hardware software integrations are reshaping the market make sure you subscribe so you catch our future deep dives.

00:10:02: we monitor these strategic shifts closely

00:10:05: Definitely.

00:10:05: And we can already see the major corporate players executing on this exact integration strategy.

00:10:11: Byrne Montag reported that Q-three results for Siemens Healthineers and they hit a book to bill ratio of one point two seven.

00:10:17: Okay,

00:10:18: let's translate that metric for second.

00:10:20: Siemens hitting at One Point Two Seven.

00:10:21: Book To Bill Ratio isn't just dry finance stat.

00:10:24: No it is not.

00:10:25: It means every one euro equipment they actually build deliver.

00:10:29: They are receiving one point twenty seven euros in new orders.

00:10:33: Hospitals are frantically stockpiling integrated hardware.

00:10:36: It signals a massive market panic to upgrade before their facilities get left behind.

00:10:41: it

00:10:41: absolutely does and Siemens is liking in that demand through major new strategic partnerships with the Cleveland Clinic and Vanderbilt University Medical Center,

00:10:51: right?

00:10:52: They are focusing heavily on advancing diagnostic imaging.

00:10:56: there are nostics an AI in radiology To automate those high volume workflows.

00:11:01: They aren't just selling machines, they are locking in long-term ecosystem partnerships based on an integrated portfolio.

00:11:07: We actually got a great look at how this advanced hardware integration is changing frontline care.

00:11:12: Wiebke Planker shared insights from her West Coast tour visiting the University of Washington Medical Center and UCLA.

00:11:18: She highlighted how Photon Counting CT Is rapidly moving From an academic novelty to true clinical standard

00:11:25: Or the Photon counting tech.

00:11:26: The resolution capabilities that technology Are completely changing what clinicians can see.

00:11:30: Yeah, it's incredible.

00:11:32: For those who haven't seen this transition in action traditional CT detectors work by converting x-rays into light and then converting that light to a signal.

00:11:41: every conversion step loses resolution.

00:11:43: Right!

00:11:43: It gets blurry.

00:11:44: Exactly.

00:11:45: Photon counting CT measures individual X-ray photons directly is like upgrading from an old blurry tube television directly to a four K OLED screen.

00:11:56: And because of the integration Seattle Children s using portable ct tech directly in the pediatric ICU for post-surgical followups.

00:12:03: That's amazing!

00:12:04: It

00:12:04: is, they are completely eliminating the need to transport highly vulnerable kids down the hall

00:12:09: and at UCLA.

00:12:10: They're using advanced photon counting CT to visualize individual coronary arteries and plaques with unprecedented detail.

00:12:18: it proves The point.

00:12:19: the intelligence And the tech are only as valuable as their integration into the physical realities of frontline care.

00:12:25: So if we pull this all together you need AI infrastructure You need deep hardware integration and you need to prove economic outcomes.

00:12:32: That is the baseline.

00:12:33: Yep, that's table stakes now

00:12:35: but here is the harsh reality for anyone building in this space.

00:12:38: You can build the most incredible outcome driving health tech on the planet And none of it matters if you get rejected at the door by a hospital procurement team.

00:12:48: Oh man, the compliance systems are unforgiving right now.

00:12:52: Alex Ferraro highlighted a massive blind spot that B-to-B health tech vendors are slamming into digital accessibility compliance.

00:13:01: Here's where it gets really interesting.

00:13:03: Imagine you're sales director in final stretch of closing huge enterprise deal.

00:13:09: You have been working this hospital system for nine months Exactly.

00:13:14: You were at the finish line.

00:13:16: suddenly, The procurement team asks for a VPAT of voluntary product accessibility template.

00:13:21: If you don't have it that deal dies right there before you even get to negotiate the price.

00:13:24: It's like this is like a bouncer in a club Right?

00:13:27: Your tech could be life-of-the-party But if you don' t have the right ID ,you aren't getting inside.

00:13:31: That s perfect analogy And regulatory mechanics driving us are strict.

00:13:35: Hospitals accept Medicare or Medicaid funding fall under section five oh four.

00:13:40: The Department of Health and Human Services has explicitly extended that obligation to all digital properties.

00:13:46: So it's not just physical rants anymore?

00:13:47: Exactly!

00:13:48: Procurement teams are now enforcing deadlines, which land in May for large organizations and May for smaller ones.

00:13:59: Think a VPAT like building inspectors blueprint for wheelchair ramps or fire exits but specifically your software code.

00:14:08: If your software relies entirely on color coding to alert a nurse, to critical patient status but the nurse using it happens be colorblind and you're.

00:14:16: VPAC doesn't account for screen reader compatibility or high contrast modes.

00:14:20: Your software is deemed a clinical hazard.

00:14:23: It's huge risk for them.

00:14:24: You don't just lose the deal!

00:14:25: You are a liability if you were listening while building a sauce roadmap.

00:14:29: you need freeze feature rollout check section five oh four compliance today Or your pipeline will hit brick wall.

00:14:36: And digital accessibility is just one layer of the gauntlet.

00:14:40: Information security isn't even stricter filter.

00:14:43: Ruth Oyendamola Johnson posted a highly detailed analysis after spending eighteen months reading OCR resolution agreements.

00:14:52: Oh, those are formal IPL enforcement documents

00:14:54: right?

00:14:55: Exactly She found that fines and failures aren't random.

00:14:59: There's five recurring security failures that show up repeatedly across industry.

00:15:04: Okay let me guess When we talk about these massive health care breaches, it's not usually state-of the art hackers burning through a firewall with zero day exploits.

00:15:12: It is probably something completely mundane like someone forgetting to delete an ex employee password.

00:15:18: You hit the nail on the head.

00:15:19: access controls that never get revoked when someone leaves a role Is one of the top recurring failures?

00:15:24: Of course he does.

00:15:25: Yeah And another major offender is simply missing risk analysis companies Not even knowing where their vulnerabilities are.

00:15:31: What else showed up in her analysis?

00:15:33: She also highlighted issues with marketing blind spots, failing to secure a system after an initial breach has already occurred and the business associate cascade.

00:15:42: I want to highlight that last one because it is critical for B-to-B vendors.

00:15:46: Go for it!

00:15:47: The Business Associate Cascade Is When A Vulnerability Doesn't Start With The Hospital Itself But With A Third Party Vendor.

00:15:54: They Use Maybe a scheduling app or an analytics tool, that vendor gets breached and it cascades up the chain to expose the hospital's patient data.

00:16:05: It means Hospital CISOs aren't just auditing you they're auditing everyone you do business with.

00:16:11: They are using these exact five patterns to grill vendors during security reviews.

00:16:15: You need rock solid answers for this specific vulnerabilities before your next Zoom call

00:16:21: If we connect us into bigger picture.

00:16:23: It becomes incredibly clear that regulatory compliance and deep-seated trust are the ultimate competitive motes

00:16:30: in healthcare.

00:16:31: Like Daniela Niprosh recently surveyed digital health startups trying to break into the German healthcare market... That's

00:16:36: a tough market to crack!

00:16:37: ...it really is, yeah And the hardest parts of their journey had nothing do with writing code

00:16:42: You really does.

00:16:43: What were this?

00:16:43: specific bottlenecks?

00:16:44: They

00:16:44: were trapped in a brutal catch twenty two.

00:16:47: The biggest barriers were understanding the complex data protection requirements, navigating incredibly long BDB sales cycles and securing initial pilot partners to build credibility.

00:16:59: Right look at how those factors interlock.

00:17:01: without regulatory clarity you cannot get a hospital to agree to a pilot.

00:17:06: Without a pilot partner You can't prove product market fit or build clinical trust.

00:17:11: And without that trust in those concrete references your enterprise sales stall completely.

00:17:16: It's just a loop of rejection.

00:17:18: Joshua Liu's reflection on his company, SeamlessMD perfectly illustrates the sheer endurance required to navigate that cycle.

00:17:26: He noted it took them eight long years To earn an integration partnership with Epic.

00:17:30: Eight

00:17:30: years of grinding.

00:17:31: Just get the partnership

00:17:32: Right.

00:17:33: Early-on they thought Our product improves patient outcomes.

00:17:36: Why wouldn't Epic want to integrate?

00:17:39: But in healthcare Potential does not equal integration?

00:17:42: No it doesn't.

00:17:43: They had do the grueling work approving their value as a standalone product.

00:17:47: first, epic and frankly any major health system relies on customer demand signals.

00:17:54: Unless the hospital is actively demanding your integration and willing to pay for it...it's not happening.

00:18:00: There

00:18:00: are absolutely no shortcuts in this industry!

00:18:04: It only compounds over years of successful compliant deployments.

00:18:08: It really emphasizes that the timeline of healthcare innovation is measured in years, not fiscal quarters which actually brings us to a final slightly provocative thought to leave you with today.

00:18:18: Okay let's hear it.

00:18:19: We have spent this deep dive talking about how deeply integrated AI and physical sensors are becoming in clinical settings.

00:18:27: But lets look at where hardware itself is heading.

00:18:29: Two posts point to wild convergence.

00:18:32: First, Fabia Teterubueno shared that visceral story you mentioned at the top of this show.

00:18:37: The car accident?

00:18:38: Yeah!

00:18:38: She had an accident and woke up to her smartwatch shaking violently having automatically called emergency services.

00:18:44: It

00:18:44: is the ultimate invisible safety net... ...the device took over when she was physically incapacitated.

00:18:49: Now pair that invisible AI-driven safety net with a breakthrough shared by Gia Tomar.

00:18:55: Researchers at Penn State have developed water based conductive ink that can be painted directly onto your skin.

00:19:02: Wait, paint it on?

00:19:03: Yes.

00:19:04: Once it dries which takes under ten minutes It functions as a highly flexible electrode.

00:19:09: It connects to a porous silver textile and tiny Bluetooth module And because its sits directly inside the microscopic grooves of your skin It eliminates motion artifacts and static you get with conventional bulky wearables.

00:19:24: It can stretch up to a hundred and fifty percent without losing signal quality, meaning it could measure ECG, EMG or EEG signals perfectly while you move.

00:19:32: You literally paint the medical sensor onto yourself like a temporary tattoo?

00:19:35: Exactly!

00:19:35: So here's the provocation for you to mull over... We just spent this entire deep dive discussing massive rigid procurement & compliance systems built to evaluate vendor software and physical medical hardware.

00:19:46: What happens our health data infrastructure when monitoring devices disappear entirely when the sensors are literally painted onto our skin.

00:19:55: Will current regulatory frameworks, data privacy laws and rigid hospital procurement cycles be ready for a world where patient quite literally

00:20:18: is?