Daily Update for September 20, 2026

Show Notes

Allina physicians strike over pay and AI; UC San Diego finds AI portal drafts slow complex replies; physician pay outpaces productivity; CPT panel weighs AI codes

The voices on this episode are generated, and so is the script, from the day's post at Physician in the Loop. Every fact in the post links to its source.

Read the post and its sources

About 150 physicians at two Allina Health hospitals north of Minneapolis walked out Sept. 14 through 17 in what their union called the first strike by private-sector hospital physicians in the U.S. Among the union's stated demands was a say in what the hospitals' AI systems are allowed to suggest. A SullivanCotter survey of about 235,000 employed physicians, released Sept. 17, showed pay rising faster than productivity, with radiology pay up 12% to 18% since 2024. And a year of data from UC San Diego Health, published in NEJM AI, found that edits requiring clinical judgment, such as interpreting a radiology result, add 60% to 70% to the time a physician spends on an AI-drafted portal reply.

The Current Procedural Terminology (CPT) Editorial Panel of the American Medical Association (AMA) met Sept. 17 to 19 in Minneapolis on a 94-item agenda that included seven proposals for codes describing AI services. The Consumer Technology Association (CTA), the consumer electronics trade group, asked Congress in a Fierce Healthcare op-ed for a single national health-AI framework in place of the more than 240 state bills it counted this year. California Gov. Gavin Newsom issued an executive order on frontier models on Sept. 18 that does not address the four health bills on his desk; those face a Sept. 30 deadline. STAT published an interview with Dr. James Deardorff on why an accurate model can still be used badly; Deardorff, a geriatrician at the University of California, San Francisco, wrote the commentary on Epic's mortality score covered in Thursday's post.

In this episode

• About 150 Allina physicians strike for four days at Mercy and Unity in what union calls a private-sector first; a say over AI among demands
• Year of edits by more than 1,000 UC San Diego physicians shows AI portal drafts that need clinical judgment take 60% to 70% longer to finish
• Pay outpaces productivity in SullivanCotter survey of 235,000 physicians; radiology up 12% to 18% since 2024, 65% of employers plan to hire
• CPT Editorial Panel takes up seven AI-service code proposals in Minneapolis, including algorithmic STEMI identification and a mammography risk score
• Consumer Technology Association CEO asks Congress for one federal health-AI framework, citing more than 240 health AI bills in 43 states this year

What is Physician in the Loop?

How technological innovation is changing medicine for patients and the people who practice on the front lines.

Nora: This is Physician in the Loop, the daily briefing on AI in medicine. I'm Nora Davis.

James: And I'm James Hudson. It's Sunday, September 20th, 2026.

Nora: About 150 physicians at two Allina Health hospitals north of Minneapolis walked out for four days last week, in what their union calls the first strike by private-sector hospital physicians in the country.

James: And a say over what the hospitals' AI systems can suggest was one of their demands?

Nora: I've got that strike, plus a new survey on physician pay outpacing productivity.

James: I'm covering a UC San Diego study on AI portal drafts, and what's on the table at this week's CPT meeting.

Nora: The union represents hospitalists, OB-GYNs, psychiatrists and critical care doctors at Allina's Mercy campus in Coon Rapids and Unity in Fridley. Doctors Council SEIU filed its strike notice September 5th, after three years without a first contract.

James: What were they asking for, besides pay?

Nora: Sick leave they say state law requires, an offer they call pay cuts, and more input into patient care decisions.

James: That AI piece is the interesting one.

Nora: Dr. John Wust, an OB-GYN on the bargaining team, told MPR News: "We have already seen AI starting to suggest diagnostic codes for patients, and we're concerned that it may start to suggest treatments."

James: CBS Minnesota put it more broadly, didn't they? Physicians wanting a voice in AI that suggests billing and treatment without them.

Nora: Right. The Minnesota Reformer counted about 73 physicians on the picket line, 37 who crossed, and 40 who weren't scheduled that week.

James: Ten percent walking out, and the hospitals call that normal operations.

Nora: That's Allina's framing. They said the strikers were about 10% of the physicians at the two campuses, and that both hospitals ran normally.

James: A year of data out of UC San Diego Health, published in NEJM AI, tracked over a thousand physicians editing AI-drafted portal replies. Researchers sorted every edit into a 15-category taxonomy built with language models and expert review, then used response time as the workload measure.

Nora: What took the most time to fix?

James: Edits that needed real clinical judgment. Interpreting a radiology result added 70.1% to the time spent on that message.

Nora: And the rest of the judgment edits?

James: Clarifying or ruling out a diagnosis added 63.9%. Interpreting lab results added 60.8%.

Nora: So the hardest calls cost the most time per message.

James: Right, but they weren't the most common edits.

Nora: What was?

James: Scheduling changes, in 38.5% of edited messages. Lifestyle advice came next at 18.2%, then added empathy at 16.1%.

Nora: And the rarest?

James: Stopping or tapering a medication, at just 2.4%.

Nora: So which burden matters more, the rare hard edit or the constant small ones?

James: Becker's Physician Leadership and Healthcare Innovation both covered it. Their read: judgment edits cost the most per message, but the high-volume administrative edits cost the most across the whole system.

Nora: SullivanCotter released its 2026 Physician Compensation and Productivity Survey on September 17th, covering 575 organizations and about 235,000 physicians.

James: And pay's outrunning productivity?

Nora: Work relative value units grew 0.8% in primary care and 2.8% in adult medical specialties.

James: Pay grew faster than that, I take it.

Nora: Pay in adult medical specialties rose 7.2% in a year, and 25.2% over five years.

James: Any specialty stand out?

Nora: Anesthesiology's up 14% to 16% since 2024. Pay for certified registered nurse anesthetists is up nearly 12%.

James: What about radiology, given all the AI tools cleared for it?

Nora: Radiology's up 12% to 18% across diagnostic, interventional and mammography work. It's the specialty with the most FDA-cleared algorithms.

James: Does the survey connect that to AI directly?

Nora: Its only mention of AI is a managing director saying shortages and AI tools together mean compensation plans need to be "agile."

James: Agile. That's one word for it.

Nora: Nearly 65% of the organizations plan to expand their physician workforce next year.

James: Anything about hiring incentives?

Nora: Nearly 95% of new hires got a sign-on bonus, and clinical outcome measures now show up in 55% of incentive plans.

James: Is this the only survey saying pay's climbing?

Nora: No, it's the third dataset this year, after Doximity and the Medical Group Management Association, all showing physician pay continuing to rise.

James: The AMA's CPT Editorial Panel met September 17th to 19th in Minneapolis, the last meeting before the 2028 code set.

Nora: How big was the agenda?

James: 94 items, and seven of them describe AI services.

Nora: What kind?

James: Four are seeking permanent codes: quantitative MRI software for liver tissue, arrhythmia analysis from cardiac simulations, a noninvasive measure for heart failure, and impedance spectroscopy of skin lesions.

Nora: And the other three?

James: Category III codes: algorithmic analysis of prostate slides, a breast cancer risk score from mammograms, and noninvasive identification of STEMI heart attacks.

Nora: There's a new vocabulary for this too, right?

James: The AMA's Appendix S taxonomy, updated the day the meeting opened. Assistive software provides data, augmentative software provides a score or parameter, and autonomous software makes the interpretation itself.

Nora: Where do these seven proposals fall?

James: Three are labeled augmentative. None is labeled autonomous.

Nora: Any bigger structural change buried in there?

James: Tab 94 would rewrite the pathology and lab code application to ask about software and its validation.

Nora: Sounds like the one that outlasts any single code.

James: That's exactly how Madhani Healthcare Consulting and the Discoveries in Health Policy blog put it.

Nora: Do we know how the votes went?

James: Not yet. The panel doesn't publish a summary for weeks, and whether Medicare pays for any of these gets decided later, in the 2028 fee schedule.

Nora: Sources and corrections are at physicianintheloop.org. I'm Nora Davis.

James: I'm James Hudson. See you tomorrow.