Daily Briefing for November 4, 2026

Show Notes

Asthma guideline ranks steroid rescue inhalers above albuterol alone

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

Get the free email briefing

The American Thoracic Society on Nov. 3 strongly recommended steroid-containing rescue inhalers over short-acting beta-agonists alone for adolescents and adults with mild asthma, in a new guideline.

The same day, the Food and Drug Administration warned that holes can form in some ButterflyBVM bag-valve-mask resuscitators and told users to remove the worst-affected lots. A National Institutes of Health cohort found that half of adults with long COVID at one year improved meaningfully by three years, and a Columbia study of nearly 1 million portal replies found that Epic's AI drafts did not shorten clinicians' overall response times.

In this episode

• Thoracic society strongly recommends steroid-containing rescue inhalers over short-acting beta-agonists alone for mild asthma
• FDA warns holes can form in ButterflyBVM bag-valve masks, letting breaths leak away from patients
• Half of adults with long COVID at 1 year improve meaningfully by 3 years, but only about a quarter fully recover, RECOVER finds
• Columbia clinicians answered portal messages 8.6% slower after Epic's AI draft replies arrived, a 983,154-message study finds
• Humana opens electronic prior authorization to more than 180,000 athenahealth clinicians nationwide
• A quarter of Kaiser emergency patients with six common conditions were treated at home, in clinics or in observation instead of a bed
• Stents and bypass surgery show matching 10-year survival for left main disease in a 5,379-patient registry

What is Physician in the Loop?

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

Voice 1: This is Physician in the Loop, the daily briefing on AI in medicine. It's Wednesday, November 4th, 2026.

Voice 1: The American Thoracic Society just strongly recommended steroid-containing rescue inhalers over albuterol alone for people with mild asthma.

Voice 2: So who exactly does that change land on, and what are they supposed to use instead?

Voice 1: Adolescents and adults at the lowest two asthma severity steps. The panel strongly recommends as-needed ICS-SABA or ICS-formoterol over a SABA reliever alone, and more conditionally prefers those over scheduled steroid inhalers plus as-needed SABA. For sicker patients, at steps three to five, they strongly recommend maintenance-and-reliever therapy with ICS-formoterol over scheduled ICS-LABA plus as-needed SABA.

Voice 2: The guideline authors wrote that steroid-containing relievers remain underutilized. Why, if the evidence points this clearly?

Voice 1: They don't say why doctors aren't prescribing it, only why they think it should change. The panel weighed preventing severe exacerbations against the risks of relying on SABA alone and the risks of extra systemic steroid exposure, and answered six separate questions with formal evidence reviews before landing here.

Voice 2: The FDA had a safety warning of its own this week, and it's a scary one. Holes can form in some ButterflyBVM bag-valve-mask resuscitators, letting a patient's breath leak away instead of reaching their lungs.

Voice 1: Which devices, specifically, and what should a hospital actually do about it?

Voice 2: The company is pulling READY lots 000109 through 000117 and standard A100 and A108 lot 000114, which had a higher defect rate. Other READY, standard and PLUS lots stay in use but get corrected instructions, and users are told to run the pre-use equipment check before every single patient.

Voice 1: Has anyone actually been hurt by this?

Voice 2: As of November 2nd, the company had reported no serious injuries or deaths. It first wrote to customers back on September 26th. The FDA's advice in the meantime is blunt: keep a backup resuscitator on hand, and check the back of the device for a hole if ventilation seems poor.

Voice 1: A quieter story, but a hopeful one. New NIH data finds half of adults still dealing with long COVID at one year had improved meaningfully by three years, though only about a quarter fully recovered.

Voice 2: Define meaningfully. What did they actually measure?

Voice 1: This is the RECOVER-Adult cohort, 1,019 adults who met the study's long COVID threshold at 12 months. By 24 months, about 36% no longer met it on two consecutive visits. By 36 months, that was about half. More than half had at least cut their symptom score in half and reported good or better quality of life, and about a quarter got all the way down to a score of zero.

Voice 2: What about the people still struggling at 18 or 24 months. Do they just stay stuck?

Voice 1: No, the study found they improved on a similar relative timeline, across 83 sites in 33 states, Washington D.C. and Puerto Rico. The authors say that's evidence of natural recovery, but they also called for what they described as the urgent need for placebo-controlled therapeutic intervention trials.

Voice 2: A very different kind of result on AI in the inbox. Nearly a million portal messages at Columbia, and Epic's AI draft replies did not shorten clinicians' overall response time.

Voice 1: Wait, didn't the drafts themselves work faster?

Voice 2: They did, per message. When a clinician actually used a draft, that reply went out nearly 30% faster. But across the whole system, after Columbia turned the tool on enterprise-wide, overall turnaround on messages rose nearly 9%.

Voice 1: So something else is eating that time savings. What's going on after hours?

Voice 2: That's the split. Draft use cut after-hours replies by about 4 percentage points for test-result questions that came in during the workday, but raised after-hours replies by about 5 percentage points for messages that arrived after hours. The researchers said that finding is sensitive to how you define working hours in the first place. The study covered first responses to more than 270,000 patients from 1,542 clinicians.

Voice 1: A tool that's faster in isolation but not faster in practice is a strange thing to roll out system-wide.

Voice 2: Before we go, something to watch. More than 500 comments have come in on the FDA's discussion paper on generative AI in devices, and they're split over its credentialing-style competency evaluation idea.

Voice 1: Split how? Who's on which side?

Voice 2: AdvaMed and the Medical Device Manufacturers Association want the agency to stick with existing risk-based frameworks. Mayo Clinic wants device competency kept separate from clinician credentialing entirely. And Anthropic is asking for benchmarks that don't come from the manufacturers or model developers themselves. The agency's Rick Abramson said back in October that draft guidance is coming in the coming months, without giving a date.

Voice 1: That's a lot of disagreement for a framework that hasn't even been drafted yet.

Voice 2: Treating AI competency like a credential for clinicians is a bigger idea than a comment period usually attracts, and it shows in who bothered to weigh in.

Voice 1: Sources and corrections are at physicianintheloop.org, where you can sign up for the free email briefing. This briefing was read by AI voices.