Beholder's Share can be supported in software without incurring much technical cost by supporting cosmetic configuration. Some Epic reports allow…
News 7/2/25
Top News
RCM technology vendor VisiQuate acquires Etyon, whose technology optimizes revenue cycle workflows.
Reader Comments
From Dark Daze: “Re: Sharp HealthCare. Cut 315 FTEs yesterday, 150 of them from IT, which is a 25% haircut.” Unverified, although the local TV station reports that IT was one of the departments affected and Redditors claim that IT was the hardest-hit area.
From Isthmus: “Re: Sharp HealthCare. Laid off most of their Epic IT staff with plans to have their Epic system supported by a third party.” Unverified.
Sponsored Events and Resources
None scheduled soon. Contact Lorre to have your resource listed.
Acquisitions, Funding, Business, and Stock
Premier acquires IllumiCare, which adds a financial component to clinical decision support, and will market it under its Stanson Health brand.
Sales
- Valley Health System (NJ) will use technology and services from Inbound Health to launch a hospital-at-home program.
- Emory Healthcare (GA) selects Aidin’s post-acute care transitions software.
People
Ralph Porpora (Grand View Health) joins Deborah Heart and Lung Center (NJ) as VP of IS/CIO.
Pager Health names Ken Yamaguchi, MD (Northwestern University’s Feinberg School of Medicine) chief medical officer.
BayCare (FL) names Patrick Brown, MD (Centra Health) VP/CMIO and Philip Karp, MS (NYU Langone Health) VP/CTO.
Medical error prevention pioneer Lucian Leape, MD died Monday at 94. He was involved in the Harvard study that resulted in the landmark 1999 report “To Err Is Human: Building a Safer Health System.”
Announcements and Implementations
Spartanburg Regional Healthcare System (SC) implements precision medicine technologies from 2bPrecise.
Microsoft’s AI team claims that one of the company’s diagnostic tools can outperform physicians in solving the toughest cases, correctly identifying 85% of 304 real-world cases that had been published weekly in the New England Journal of Medicine, four times the success rate of doctors. This is an interesting test, with these caveats:
- The research involved published cases, so physicians were barred from using electronic medical records, care guidelines, textbooks, colleagues, and other LLMs that could have accessed the answers.
- NEJM cases are “pedagogically curated” rather than real-world examples.
- Nobody knows whether commercial LLMs were trained on the actual published case studies, which would have given them an unfair advantage.
- The physicians who were involved had a median experience of 12 years, but still diagnosed correctly just 20% of the time, raising questions of the validity of the hypothetical cases.
- The NEJM cases are created to be interesting, challenging, and pathology-focused. They may contain red herrings and often involve obscure diagnoses. They require the reader to make a diagnosis without having access to the patient’s medical record, without being able to examine the patient or ask them questions, and without the benefit of observational findings.
- The study is an interesting proof of concept, but should not be taken to mean that doctors can’t diagnose correctly 80% of the time or that AI can do a better job in clinical situations.
Government and Politics
CMS approves Eko Health’s Sensora digital stethoscopes cardiac algorithm for payment under the Hospital Outpatient Prospective Payment system.
The Washington Post provides details from court documents of the federal government’s Operation Gold Rush, a massive fraud takedown where Medicare was billed $10.6 billion for medical devices that were not medically necessary or actually delivered to patients:
- Prosecutors say the schemers used stolen identities of 1 million Americans to flood Medicare with bogus claims, including 1 billion urinary catheters.
- The fraud was funneled through 30 Medicare-approved medical supply companies that the perpetrators had quietly acquired.
- HHS and CMS teams blocked 99% of the attempted Medicare payments, signaling a shift from the traditional “pay and chase” model to one of proactive prevention.
- Despite that success, the scammers still collected $1 billion from Medicare supplemental insurers.
- Most of the accused reportedly never set foot in the United States.
- Medicare received 400,000 complaints from patients questioning why companies like Pretty In Pink Boutique were billing for catheters they neither needed nor received.
Privacy and Security
Central Maine Healthcare restores its IT systems and resumes normal operations after a June 1 cyberattack forced it to downtime procedures.
Other
In South Africa, the Gauteng health department denies reports that a hospital morgue was backed up with two weeks’ worth of bodies because it couldn’t print death certificates due to a shortage of printer cartridges.
Sponsor Updates
- A new analysis from Black Book Research reveals a sharp escalation in state-level legislative actions that are reshaping revenue cycle management frameworks.
- Nordic releases a new episode of its “Designing for Health” podcast featuring UXonSpec founder and CUO Randolph Bias.
- CereCore releases a new podcast episode titled “Building Trust: From Expanse Rollout to High Reliability Organization.”
- Vyne Medical releases a new case study titled, “How Mayo Clinic Reduced Lab Data Entry Time by 50% While Increasing Data Volume.”
- AdvancedMD adds health insurance contracting and consulting services firm NGA Healthcare to its partner integration marketplace.
- Agfa HealthCare releases a new guide titled “Purchasing a Cloud-Hosted Enterprise Imaging Solution.”
- Clinical Architecture releases its “2025 Healthcare Data Quality Report.”
Blog Posts
- Enhancing Clinical and RCM Efficiency with Coding Automation (AGS Health)
- 4 Takeaways from Rooted in Purpose: Advancing Whole-Person Care for Individuals with IDD & Autism (Netsmart)
- New Nursing Charge Automation Adds $45K in Weekly Revenue (CereCore)
- EPRs at the heart of urgent and emergency treatment centres: Delivering safer, faster care (Altera Digital Health)
- How Vanderbilt University Hospital Tackled Readmission Rates By Leveraging Artera for Post-Discharge Communication (Artera)
- Set the foundation: How to prepare your data for AI transformation (Philips Capsule)
- From Texas to Wisconsin, Cardamom’s David Gibson shares his winding journey in health IT (Cardamom)
- MUSE 2025: Innovation, Connection, and Giving Back (Consensus Cloud Solutions)
- Why It’s Time to Rethink Prior Authorization—Especially for Specialty Medications (DrFirst)
Contacts
Mr. H, Lorre, Jenn, Dr. Jayne.
Get HIStalk updates.
Send news or rumors.
Follow on X, Bluesky, and LinkedIn.
Sponsorship information.
Contact us.
Re: NEJM study of AI vs. Physician Diagnostic Abilities
Call me a skeptic, but this seems to me a very “Sherlock Holmsian” way to study diagnostic abilities. As such it is both unrealistic and measures the wrong things.
I mean seriously, the physicians had no access to the patient or the patient’s medical record? This seems a limitation designed to put the physician on an equal footing with the AI, but it veers too far away from how care is actually delivered.
This also brings me to my critique of the Sherlock Holmes approach. In my experience, people typically overrate their abilities to bring pure logic to bear on most problems. In addition, the facts in any case are often not facts at all, because they aren’t correct. Or the facts may be correct but they aren’t determinative.
How can facts be incorrect? Just one example, but test result uncertainties come to mind, with both false positives and negatives.
How can facts fail to be determinative? Again just one example, but often the real thing you’d like to know isn’t available for some reason, so you use a proxy result instead because it is available. And the proxy is imperfect.
In the world of crime & detective literature, Agatha Christie was very effective at using some of these ideas in her novels. Use of the “unreliable narrator” was a key plot point in a couple of her stories, which seeded the plots with dead ends, false leads, and red herrings.
On the positive side, the NEJM study did attempt to capture some of this, as their cases “… may contain red herrings …”
There is an analog in computing. I had my mind blown once by discovering data in a database, that was larger than the defined maximum field length. What I had assumed was an absolute, inviolable rule, turned out to be more of a best-efforts guideline. There was an enforcement mechanism which had holes in it. Oops! Yet when you call it a “maximum”, that leads to pretty reasonable expectations.