THCB

Categories

Above the Fold

A Failure to Rescue: How Predictive Modeling Can Rewrite the Story of Congenital Syphilis

By KAYLA KELLY

Every semester I have the privilege of guiding nursing students through their maternal and pediatric clinicals. At the beginning of the semester, their enthusiasm is contagious. They share stories about witnessing their first delivery, helping a new mother with breastfeeding, and practicing developmental assessments on pediatric patients. As the semester progresses, I see their demeanor shift. “You were right, we took care of another congenital syphilis baby today.” Their reflections on the clinical day are a mixture of emotions: frustration, anger, and sadness, as they watch fragile infants fighting an infection that no child should ever have to endure.

When I first tell my nursing students that they will likely care for infants born with syphilis during their clinical rotations, they look at me with wide-eyed disbelief. “Didn’t we cure syphilis in the 1950’s?” some ask. A few of my students usually recall hearing about the Tuskegee Study, but most have no idea that we are still fighting (and losing) a battle against congenital syphilis in the United States today. 

Congenital syphilis occurs when a mother transmits the infection to her infant during pregnancy or delivery. It is almost entirely preventable with timely screening and treatment, yet the number of cases continues to rise at an alarming rate. Between 2018 and 2022, the United States experienced a 183% increase in congenital syphilis cases, rising from 1,328 cases to 3,769. This national trend was mirrored at the state level, with Texas reporting 179 cases in 2017 and 922 in 2022. During those five years, the rate of infants born with congenital syphilis in Texas rose from 46.9 to 236.6 per 100,000 live births, a sharp increase that necessitates action.

Texas now has one of the highest congenital syphilis rates in the country, despite having one of the most comprehensive prenatal screening laws. According to the Texas Department of State Health Services, policy mandates syphilis screening at three points during pregnancy:

(1) at the first prenatal visit

(2) the third trimester (but no earlier than 28 weeks)

(3) at delivery

But herein lies the problem: What happens when a woman never attends prenatal care? How do we reach those who never step into an OB/GYN office during pregnancy? Screening laws only protect those who are able to access care. In 2022, over 1/3 of Texas mothers whose infants were diagnosed with congenital syphilis did not receive any prenatal care. Each of these cases represents a failure of our current medical system, a system that should be protecting the most vulnerable yet remains unable to reach those who need it most.

Continue reading…

When Your Cloud Provider Doesn’t Understand HIPAA: A Cautionary Tale

By JACOB REIDER & JODI DANIEL

Jacob: I recently needed to sign a Business Associate Agreement (BAA) with one of the large hosting providers for a new health IT project. What should have been straightforward turned into a multi-week educational exercise about basic HIPAA compliance. And when I say “basic,” I mean really basic, like the definitions in the statute itself.

Here’s what happened and why you need to watch out for this if you’re building health care technology.

I’m building a system that automates clinical data extraction for research studies. Like any responsible health care tech company, I need HIPAA-compliant infrastructure. The company (I’ll call them Hosting Company or HC) is good technically, and they’re hosting our development environment, so I signed up for their enhanced support plan (which they require before they’ll even consider a BAA) and requested their standard agreement.

The Problem

HC’s BAA assumes every customer is a “Covered Entity.” That means a health plan, a health care clearinghouse, or a health care provider that transmits health information electronically.

But that’s not me. I’m not a Covered Entity. I’m a Business Associate (BA). I handle protected health information on behalf of Covered Entities. When I need cloud infrastructure, I need my vendors to sign subcontractor BAAs with me.

The Back and Forth

When I told HC that I couldn’t sign their BAA as written, they escalated to their legal department. Days later, a team lead came back with this response:

“To HC, even if you are a subcontracted or a down the line subcontracted association. It would still be an agreement between the covered entity within the agreement and HC… So even being a business associate, it would still be considered a covered entity since it is your business that is being covered.”

I had to read it twice. This is simply wrong.

Jodi: Let me chime in here with the legal perspective, because this confusion is more common than it should be.

The terms “Covered Entity” and “Business Associate” aren’t interchangeable marketing terms. They have specific legal definitions in 45 CFR § 160.103. You can’t just redefine them because it’s administratively convenient. Generally… covered entities are (most) health care providers, health plans, and health care clearinghouses; business associates are those entities that have access to protected health information to perform services on behalf of covered entities; and subcontractors are persons to whom a business associate delegates a function, activity, or service.

Here’s what the regulations actually say:

Continue reading…

Sachin Jain–How do we do better?

What are the practices that we have normalized that future generations will criticize us for? Sachin Jain, CEO of SCAN Health Plan, is perhaps the leading truth teller in health care who also runs a real health care organization. I had a really fun but serious interview with Sachin about what health care people are doing, what are the bad things that happen. How are good people letting this happen? How we should be changing what we are doing?–Matthew Holt

Hospitals are Incompetent Monopolists!

By JEFF GOLDSMITH

The health policy community is obsessed with hospital mergers. In a recent paper which I critiqued, the operating thesis was that hospital mergers are conspiracies in restraint of trade, enabling hospitals to extract rent from helpless local employers and patients. This logic leads directly to advocacy (lavishly funded by Arnold Ventures philanthropy) of hospital rate controls as the only way of restraining this abuse of economic power.

The reality is, as you might expect, somewhat different. The following chart, courtesy of healthcare data firm Trilliant Health, shows that hospitals are truly incompetent monopolists. It shows the correlation between hospital operating margins and market concentration for 2023. The hospitals to the far right in this chart have 100% local market shares.

Source: Trilliant Healthcare Analysis of CMS HCRIS files (Hospital Cost Reports), 2023

Do you see a correlation? I sure don’t.

According to Trilliant, the average hospital operating margin in 336 CBSAs (markets) where hospital services are “controlled by a single firm” is -1.7%.This negative operating margin average does NOT include the operating losses on their physician practices, which are not reported on hospital cost reports, so the actual operating losses are likely much greater.

Jeff Goldsmith is a veteran health care futurist, President of Health Futures Inc and regular THCB Contributor. This comes from his personal substack

Kai Romero, Evidently

Kai Romero is Head of Clinical Success at Evidently. The company is one of many that are using AI to dive into the EMR and extract data to deliver it to clinicians. It works to get really great information from the EMR to various flavors of clinicians in a fast and innovative way. Kai leads me on a detailed exploration of how the technology gets used as a layer over the EMR. And Kai shows me the new version that allows and LLM to deliver immediate answers from the data. This is a demo you really need to see to understand how AI is changing, and improving, that clinical experience. Meanwhile Kai is fascinating. She was an ER doc who became a specialist in hospice. We didn’t get into that too much, but you can tell about her input into Evidently’s design — Matthew Holt

Life Is Geometry

By KIM BELLARD

In 2025, we’ve got DNA all figured out, right?  It’s been over fifty years since Crick and Watson (and Franklin) discovered the double helix structure. We know that permutations of just four chemical bases (A, C, T, and G) allow the vast genetic complexity and diversity in the world. We’ve done the Humam Genome Project. We can edit DNA using CRISPR. Heck, we’re even working on synthetic DNA. We’re busy finding other uses for DNA, like computing, storage, or robots. Yep, we’re on top of DNA.

Not so fast. Researchers at Northwestern University say we’ve been missing something: a geometric code embedded in genomes that helps cells store and process information. It’s not just combinations of chemical bases that make DNA work; there is also a “geometric language” going on, one that we weren’t hearing.

Wait, what?

The research – Geometrically Encoded Positioning of Introns, Intergenic Segments, and Exons in the Human Genome – was led by Professor Vadim Backman, Sachs Family Professor of Biomedical Engineering and Medicine at Northwestern’s McCormick School of Engineering, and director of its Center for Physical Genomics and Engineering. The new research indicates, he says, that: “Rather than a predetermined script based on fixed genetic instruction sets, we humans are living, breathing computational systems that have been evolving in complexity and power for millions of years.”

The Northwestern press release elaborates:

The geometric code is the blueprint for how DNA forms nanoscale packing domains that create physical “memory nodes” — functional units that store and stabilize transcriptional states. In essence, it allows the genome to operate as a living computational system, adapting gene usage based on cellular history. These memory nodes are not random; geometry appears to have been selected over millions of years to optimize enzyme access, embedding biological computation directly into physical structure.

Somehow I don’t think Crick and Watson saw that coming, much less either Euclid or John von Neumann.

Coauthor Igal Szleifer, Christina Enroth-Cugell Professor of Biomedical Engineering at the McCormick School of Engineering, adds: “We are learning to read and write the language of cellular memories. These ‘memory nodes’ are living physical objects resembling microprocessors. They have precise rules based on their physical, chemical, and biological properties that encode cell behavior.”

“Living, breathing computational systems”? “Microprocessors”? This is DNA computing at a new level.

The study suggests that evolution came about not just by finding new combinations of DNA but also from new ways to fold it, using those physical structures to store genetic information. Indeed, one of the researchers’ hypothesis is that development of the geometric code helped lead to the explosion of body types witnessed in the Cambrian Explosion, when life went from simple single and multicellular organisms to a vast array of life forms.

Coauthor Kyle MacQuarrie, assistant professor of pediatrics at the Feinberg School of Medicine, points out that we shouldn’t be surprised it took this long to realize the geometric code: “We’ve spent 70 years learning to read the genetic code. Understanding this new geometric code became possible only through recent advances in globally-unique imaging, modeling, and computational science—developed right here at Northwestern.” (Nice extra plug there for Northwestern, Dr. MacQuarrie.)

Coauthor Luay Almassalha, also from the Feinberg School of Medicine, notes: “While the genetic code is much like the words in a dictionary, the newly discovered ‘geometric code’ turns words into a living language that all our cells speak. Pairing the words (genetic code) and the language (geometric code) may enable the ability to finally read and write cellular memory.”

I love the distinction between the words and the actual language. We’ve been using a dictionary and not realizing we need a phrase book.   

Continue reading…

Waiting for Codman: 100+ Years of Profits > Patients

By LEONARD D’AVOLIO

I’m in the waiting room of the New England Baptist Hospital. They just wheeled my father to the OR. It’s strange to be back. 

Once upon a time, their Chief Medical Officer, Dr. Scott Tromanhauser asked for my help. He was interested in improving the outcomes of total knee replacement surgeries. Nearly 20% of all knee replacements do not improve outcomes. The greatest opportunity for improvement is reducing unnecessary surgeries. 

This seems straightforward enough to the casual reader but in the upside down that is US healthcare, very few surgical centers in this country bother to learn if their surgeries make things better or worse. Doing anything that threatens to reduce volume is bad for business. 

We pitched a concept to his Board of Directors. 

“What if,” we proposed, “we could measure 1 year post-operative outcomes of every total knee replacement? We could share that data with our surgeons and see – for the first time – how our patients fared. With enough data, we could make personalized predictions of outcomes during a pre-operative consult visit. We could give people the information they need to make good medical decisions.” 

They supported the idea. Yes, it might lead to fewer surgeries – but these were the surgeries that shouldn’t be conducted. Plus, it might be an edge during price negotiations with payors. Beyond that, they concurred, it was the right thing to do. 

Scott and I celebrated the approval with a walk through the Mount Auburn Cemetery to visit the grave of Dr. Ernest Codman. It was his idea after all. 

Dr. Codman, was a surgeon at Mass General Hospital in 1905 when introduced his “End Results System.” In it, he proposed that every hospital capture data before, and for at least one year, after every procedure. This was to find out if the procedure was a success and if not, to ask “why not?” Codman wanted patients to have this information. How else would outcomes improve? How else would patients make good medical decisions?

Now, more than 100 years later, we would bring his idea to life, just miles down the road from where he introduced it.

Continue reading…

Artificial Intelligence Renders the FDA’s Current Drug Approval Process to be Obsolete

By STEVEN ZECOLA

Artificial intelligence (“AI”) has taken root in the field of drug discovery and development and already has shown signs of running past the traditional model of doing research.  Congress should take note of these rapid changes and: 1) direct the Department of Health and Human Services (“HHS”) to phase down the government’s basic research grant program for non-Ai applicants, 2) require HHS to redirect these monies to fund nascent artificial intelligence applications, and 3) require HHS to revamp the roadmap for drug approvals of AI-driven trials to reflect the new capabilities for drug discovery and development.

Background

There are four distinguishing features of the U.S. healthcare industry.

First, the industry’s costs as a percentage of GNP have increased from 8% in 1980 to 17% today, and are expected to exceed 20% by 2030.  The federal government subsidizes roughly one-third of these costs.  These subsidies are not sustainable as healthcare costs continue to skyrocket, especially in the face of an overall $37 trillion federal deficit.

Second, the industry is regulated under a system that results in an average of 18 years of basic research and 12 years of clinical research for each drug approval.  The clinical cost per newly approved drug now exceeds $2 billion.  The economics of drug discovery are so unattractive to investors that the federal government and charitable foundations fund virtually all basic research.  The federal government does so to the tune of $44 billion per year.  When this cost is spread among the 50 or so drug approvals per year, it adds a cost of roughly $880 million to each drug, bringing the total cost to over $3 billion per drug approval. Worse yet, the process is getting slower and more costly each year.  As such, drug discoveries under the current research approach will not be a significant contributor to lowering the overall healthcare costs.

Third, the Trump administration has undercut the federal government’s role in healthcare by firing several thousand employees from HHS.  Thus, the agency can no longer effectively administer its previously adopted rules and regulations, and therefore, cannot be expected to shepherd drug discovery into lowering healthcare costs.

Fourth, on the positive side, artificial intelligence software combined with the massive and growing computational capacity of supercomputers have shown the potential to dramatically lower the cost of drug discovery and to radically shorten the timeline to identify effective treatments.

Enter Artificial Intelligence (AI) into Drug Discovery

For the past decade, a handful of companies have been exploring advanced automation techniques to improve the many facets of the drug discovery process. Improvements can now be had in fulfilling regulatory documentation requirements, which today add up to as much as 30% of the cost of compliance.  More significantly, Ai can be used to accurately create comprehensive clinical documents from raw data with citations and cross-references – and continually update and validate the documentation.

The top Ai drug discovery companies include Insilico Medicine, Atomwise, and Recursion, which leverage Ai to accelerate various stages of drug development, from target identification to clinical trials. Other notable companies are BenevolentAI, Insitro, Owkin, and Schrödinger, alongside technology providers like Nvidia that supply critical Ai infrastructure for the life sciences sector.

Continue reading…

The ‘After Phase’ Is Missing: Why Every GLP-1 Prescription Needs an Exit

By HOLLI BRADISH-LANE

I’ve seen clients start GLP-1 medications full of hope—and stop them feeling betrayed by their own biology.

Some reached their limit with side effects: relentless nausea, fatigue, or the quiet loss of joy in eating. Others simply couldn’t afford to stay on. A few never saw the promised results at all. But for nearly all of them, the story ended the same way—one step forward, five steps back.

We celebrate the success stories of GLP-1s, but we rarely talk about the crash that follows when treatment stops. And it’s not just psychological. The body rebounds fast—hunger, weight, and metabolic chaos rush back in.

The problem isn’t the medication itself. It’s that we’ve built an elegant on-ramp for GLP-1s—and almost no off-ramp at all.

The Evidence Is Already Warning Us

The data couldn’t be clearer. In the STEP-1 extension trial, participants who stopped semaglutide regained roughly two-thirds of the weight they had lost within one year. Their blood pressure, cholesterol, and blood-sugar levels slid back toward baseline.

A nearly identical pattern appeared in the SURMOUNT-4 trial for tirzepatide: those who continued therapy maintained—or even deepened—their weight loss; those who stopped rapidly regained.

Meanwhile, the SELECT cardiovascular outcomes trial showed semaglutide reduced major cardiac events in people with overweight and obesity. That’s a major win—but also a reminder that stopping abruptly can erase much of the benefit.

Both the American Diabetes Association 2025 Standards of Care and the American Gastroenterological Association guidelines now emphasize continuing anti-obesity pharmacotherapy beyond initial weight loss goals.

The implication is simple: for most patients, GLP-1s are not a 12-week intervention—they’re chronic therapy.

Yet in real life, chronic use isn’t always realistic.

Why So Many Will Stop Anyway

Insurance coverage ends. Supplies run short. A job changes, or a deductible resets. Some patients plan a pregnancy, experience intolerable side effects, or simply want to know who they are without the injection. Others plateau despite perfect adherence and feel the drug has stopped working.

In each case, the result is the same… withdrawal without a plan.

Continue reading…

Justin Schrager demos Vital.io

Justin Schrager is the CMO of Vital.io. Their technology sits in the hospital telling patients what is going on with their care while they are in the hospital, particularly in the ER. Justin showed a deep demo about the patient experience of using Vital.io which includes what the patient can expect and guides them through the confusing workflow. It allows the patient to make requests, and also lots of guidance about what is happening to them, or for example what lab results might mean. It goes as far as helping people book appointments for follow up with the right doctor. We had a great chat about the product and also about the realities of running a tech company that has to integrate with Epic and many other EMRs.–Matthew Holt

assetto corsa mods