Daily briefing

Medicare proposes making Advantage plan doctors enroll in traditional Medicare

Nov 17, 2026

Illustration: a white padlock with a dark shackle, a glowing key in the keyhole, and a blank tag hanging from it.

The Centers for Medicare & Medicaid Services on Nov. 16 proposed requiring clinicians who bill Medicare Advantage plans to enroll in traditional Medicare, in a fraud rule it says would save nearly $90 billion.

The proposal would also give every Medicare provider 30 days, down from as many as 90, to report changes to enrollment information, and would let the agency deny or revoke enrollment when an owner's or manager's identity cannot be confirmed, according to the agency's fact sheet.

In a randomized study from Germany, physicians' accuracy fell as an AI tool was made to fail, and one in eight noticed without being asked. The New England Journal of Medicine published a phase 3 trial of semaglutide in children as young as 6, and FDA eased pregnancy-testing rules for isotretinoin.

  • WatchRegulation Nov 16, 2026

    Comments are due 60 calendar days after the proposed rule is published in the Federal Register, according to CMS.

    PracticesHospitals
    The Centers for Medicare & Medicaid Services issued the Comprehensive Regulations to Uncover Suspicious Healthcare (CRUSH) proposed rule on Nov. 16. Providers and suppliers that bill Medicare Advantage plans would generally need approved enrollment in traditional Medicare, which carries identity verification, licensure review, site visits and fingerprint background checks; that requirement would start with the plan year beginning two years after a final rule is published. All providers would have 30 calendar days to report any change to their enrollment information, down from 90 days for noncritical changes, and CMS could deny or revoke enrollment when an owner, managing employee, officer or director has an identity that cannot be confirmed. CMS estimates nearly $90 billion in federal savings and about $378.51 million in costs from 2027 to 2036, and said it intends to issue program integrity rules like this one every year.
  • FYIEvidence Nov 13, 2026
    All physicians
    Physicians classifying heart sounds with an AI assistant lost accuracy as the tool was made steadily less reliable, even when their own unassisted accuracy was higher than the AI's, according to a randomized study published Nov. 13 in npj Digital Medicine. Researchers at Philipps-Universität Marburg in Germany had 48 participants, 18 of them final-year medical students, classify 40 heart sounds while the AI's reliability fell in blocks from 90% to 60%, 50% and 0%; AI-assisted accuracy dropped from 81.45% to 58.33%. Making participants commit to their own classification before seeing the AI's did not improve performance, and their confidence fell faster than their trust in the tool. Only 12.5% noticed the decline without being asked. Human oversight alone "may be insufficient to reliably detect system malfunctions," the authors wrote.
  • FYIEvidence Nov 16, 2026
    PediatricsPrimary care
    Once-weekly semaglutide lowered body-mass index 13.5% at 68 weeks in children 6 to under 12 years old with obesity, while BMI rose 2.7% on placebo, in a double-blind phase 3 trial published Nov. 16 in the New England Journal of Medicine. The STEP Young trial randomly assigned 167 children 2:1, with lifestyle counseling in both groups; 38% of children on semaglutide and 2% on placebo ended below the 95th percentile threshold for obesity. At 104 weeks the between-group difference was 17.8 percentage points. Gastrointestinal adverse events were the most common and more frequent with semaglutide, and led one child to stop the drug. Semaglutide is approved for obesity in adolescents and adults; Novo Nordisk funded the trial.
  • FYIIncident Nov 16, 2026
    Investigators at the Centers for Disease Control and Prevention counted 112 confirmed measles cases during pregnancy and 13 cases of congenital measles reported from Jan. 1, 2025, to Sept. 28, 2026, in a preliminary analysis published Nov. 16 in NEJM Evidence. Complications included maternal pneumonia, preterm delivery and neonatal complications. Of the pregnant patients, 98% had not received MMR vaccine before pregnancy or had unknown vaccination status, the authors reported.
  • FYIRegulation Nov 16, 2026
    EndocrinologyPrimary care
    MiniMed said Nov. 16 that the Food and Drug Administration cleared MiniMed Fit, a tubeless patch pump for people with type 1 or type 2 diabetes ages 7 and older that holds 300 units and can be worn up to seven days when the total daily dose is below 42 units. Its algorithm adjusts delivery automatically and includes meal detection that can add insulin when a meal dose is missed or underestimated, the company said; a phased U.S. rollout through pharmacies is expected in spring 2027. Tandem Diabetes Care disclosed clearance of a tubeless version of its Mobi automated insulin delivery system, cleared for up to seven days of wear, with full availability planned in the first half of 2027 through pharmacy channels only, MedTech Dive reported.
  • FYIEvidence Nov 12, 2026
    At least 1,031 optical coherence tomography volumes appear under different identifiers in two public Harvard glaucoma datasets that AI studies have treated as independent validation cohorts, according to an audit published Nov. 12 in JAMA Ophthalmology. The authors, from Musashino University in Tokyo, found the overlap covered at least 31% of the 3,300 eyes in Harvard-GF, and none of the 64 citing studies whose full text they obtained reported it. A classifier trained on one dataset scored an area under the curve of 0.90 on the other, where 78.5% of test eyes were training data, against 0.78 out of sample, and the inflation appeared in every racial subgroup. The datasets "were not independent cohorts," the authors wrote.
  • FYIMoney Nov 16, 2026
    HospitalsPractices
    Legacy vendors such as Microsoft's Dragon Copilot and Epic's Art captured 68% of new ambient scribing dollars in 2026 and now hold 46% of that market, while AI startups lost market share, according to Menlo's 2026 State of AI in Healthcare report, based on a survey of more than 550 U.S. health care and life sciences executives, Healthcare Dive reported Nov. 16. Spending on clinical copilots grew about fivefold to $550 million, with OpenEvidence the most popular, and providers accounted for about 77% of health care AI spending. More than two-thirds of providers said they prefer to keep coding, billing and chart review within their EHR, against 16% for patient navigation.
  • ActRegulation Nov 16, 2026

    The changes took effect Nov. 15; prescribers allowing home pregnancy tests must set up processes against misread or falsified results, FDA said.

    Changes to the iPLEDGE risk management program for isotretinoin took effect Nov. 15, the Food and Drug Administration said. Pregnancy tests before treatment must still be done in a medical setting, but prescribers may let patients take later monthly tests at home if they set up processes to limit misinterpretation and falsification of results. A patient who can get pregnant and misses the seven-day pickup window can repeat the test right away with no added wait. For patients who cannot get pregnant, monthly counseling documentation and the 30-day prescription window end. Enrolled patients, prescribers and pharmacies do not need to re-enroll or re-certify, FDA said.
    Source: FDA

Latest news

  1. At-home light and sound headset for Alzheimer's disease missed its primary endpoint in a phase 3 trial, Cognito says
  2. Acupoint thread embedding cut weight 3.0 kg more than sham at 24 weeks in women with overweight or obesity
  3. Medicare roadmap says AI tools could deliver many parts of the annual wellness visit and lead care in payment models

See all the latest news