From TekinHealthHealthcare Tech Field Notes

CMS released preliminary payment rates for the 2027 Medicare Clinical Laboratory Fee Schedule on September 21, 2026, showing 1,171 of 1,947 test codes would be paid less than in 2026 under the PAMA private-payer-rate reporting cycle. The agency says the cuts average about 16% overall, with molecular pathology and genomic sequencing codes down roughly 22-23%; reductions would phase in at up to 15% a year from 2027 through 2029. CMS estimates the changes save Medicare about $1 billion a year, while the American Clinical Laboratory Association says the new rates are based on private-payer data from only about 2% of labs that bill Medicare Part B. A 30-day comment period is open, with CMS expecting to finalize rates in November 2026.

Why it's notable: Lab reimbursement funds the diagnostic testing infrastructure clinics and hospitals rely on, and a sustained double-digit cut concentrated in molecular pathology and genomic sequencing codes could affect which tests labs can keep offering.

CMS announced that its ACCESS model will add tracks for heart failure, chronic obstructive pulmonary disease, substance use disorder and tobacco cessation starting in spring 2027. ACCESS is a voluntary 10-year Medicare model that pays participants for technology-enabled chronic care between visits. It launched on July 5, 2026, with tracks for hypertension, diabetes, chronic musculoskeletal pain and depression. Full payment depends on patients reaching defined outcomes, such as better blood pressure control. CMS says about 160 organizations joined at launch and that three in four Medicare beneficiaries qualify for at least one track.

Why it's notable: ACCESS is Medicare's main outcome-based payment route for virtual care, remote monitoring and connected-device programs. Adding common high-cost conditions like heart failure and COPD widens the pool of patients whose between-visit care clinics and digital health partners can be paid for.

The FDA granted De Novo authorization to Powerful Medical's PMcardio Queen of Hearts, an AI model that reads 12-lead ECGs for signs of urgent acute coronary syndrome, including STEMI and STEMI equivalents, and alerts clinicians to suspected cases. STAT reported the authorization on September 11 and the company announced it on September 15. The company says the model is backed by more than 20 peer-reviewed studies covering over 40,000 patients, found about twice as many heart attacks as standard care with up to five times fewer false positives, and is already used across more than 150 hospital locations, mostly in Europe.

Why it's notable: Many heart attacks caused by a fully blocked artery don't meet standard STEMI criteria on the ECG. A tool that flags them could get more of these patients to the cath lab faster. De Novo authorizations are rare, and this one creates a new FDA device category for AI ECG models that detect acute, life-threatening heart conditions.

The FDA's Digital Health Center of Excellence published a discussion paper on how it might regulate medical devices that use generative AI. The paper sets out a two-axis risk framework: how independently a device acts, and how much harm a wrong output could cause. It also proposes competency-based premarket testing that combines benchmarking with clinical confirmation, and it covers postmarket monitoring, foundation models and agentic systems. Comments are due October 19, 2026, under docket FDA-2026-N-7874. The FDA has not yet authorized any generative AI device.

Why it's notable: This is the clearest signal so far of how the FDA may treat LLM-based clinical tools, including chatbots and agents. Those rules will decide which of these tools can be marketed for clinical use and what evidence vendors must show.

CMS approved a New Technology Add-on Payment (NTAP) for Aidoc's CARE Body CT Multi-Triage. The FDA-cleared software flags suspected urgent findings on contrast and non-contrast CT of the chest, abdomen and pelvis. Holland & Knight says this is the first NTAP CMS has granted for an AI-enabled diagnostic tool. Starting October 1, 2026, hospitals can receive the add-on for eligible Medicare fee-for-service inpatient cases for three years, with a reported maximum of $137.53 per case.

Why it's notable: How to pay for AI has been one of the main barriers to hospitals adopting imaging AI. A Medicare add-on for an AI triage tool gives other vendors and hospital finance teams a path to follow.

Johnson & Johnson received FDA de novo authorization on July 22, 2026 for its Ottava robotic surgical system, which builds four robotic arms into the operating table. The authorization covers upper-abdominal and other general surgery procedures including Roux-en-Y gastric bypass, sleeve gastrectomy, cholecystectomy, splenectomy, appendectomy, small bowel resection, fundoplication and hiatal hernia repair. The authorization came about six months after J&J filed. The company says it will launch with select U.S. customers first, and a U.S. trial in inguinal hernia repair is ongoing.

Why it's notable: Ottava is a second major competitor to Intuitive's da Vinci in soft-tissue surgery, after Medtronic's Hugo, which gives hospitals buying robots more choice on price and operating-room footprint.

CMS's proposed CY 2027 Medicare Physician Fee Schedule, released July 14, 2026, would pay for remote physiologic monitoring (RPM) and remote therapeutic monitoring (RTM) only when the clinical staff doing the work are employed by the billing practice, not by a contracted vendor. It would also limit the services to established patients, require a separately billed initiating visit before monitoring starts, and revalue the device codes. CMS is also considering bundling the codes into four new G-codes. The same rule carries forward telehealth flexibilities through 2027 and proposes a roughly 1.7% cut to the main conversion factor. Comments were due September 14; if finalized, most provisions take effect January 1, 2027.

Why it's notable: Many practices run RPM through third-party vendors, so the proposal could force them to bring monitoring in-house or drop it. It follows an OIG finding that about 43% of 2022 Medicare RPM patients did not get all three required components.

On June 26, 2026, HHS's Office of the National Coordinator for Health IT (ONC) said the number of health records exchanged through the Trusted Exchange Framework and Common Agreement (TEFCA) has grown from about 10 million to more than 1 billion in less than a year. ONC awarded Alliance Global Tech, a federal IT contractor in Columbia, Maryland, a one-year contract of about $1.3 million to check that TEFCA participants follow the network's rules. The contract can be renewed through 2031, for up to about $5.5 million in total. ONC also said it is doing extra compliance reviews of the 11 Qualified Health Information Networks, a group that now includes eClinicalWorks, Netsmart, Oracle Health and Surescripts, and of their participants.

Why it's notable: TEFCA is becoming the main national route for pulling outside records into a clinic's EHR. The new oversight follows disputes over improper record access, including Epic's lawsuit against Health Gorilla, so participants should expect closer checks on why and how they query records.

Oura introduced its Ring 5 on May 28, 2026, with plans to add a feature that tracks overnight blood pressure patterns and tells users when to see a doctor. The feature shows trends rather than systolic and diastolic readings, and it is sold as a wellness tool without FDA review under the agency's January 2026 wellness guidance. Oura said a separate feature to flag possible hypertension is being studied and will go through the FDA. The launch also added health record import and access to clinicians through Counsel Health and ResMed.

Why it's notable: Clinicians should expect patients to bring in blood pressure trends from rings and watches that no regulator has checked for accuracy. Experts quoted by STAT warned that these unvetted blood pressure features are spreading quickly.

Iowa Gov. Kim Reynolds signed House File 2635 on May 13, 2026, effective July 1, 2026. The law says insurers may not use artificial intelligence as the sole basis for denying, delaying or downgrading medically necessary care, and a human clinical reviewer must be involved. It also removes prior authorization for recommended cancer screenings, sets deadlines for insurers to tell hospitals the status of a claim, and requires clinical explanations when care is denied. The bill passed the Iowa House 87-0 and was opposed mainly by insurers.

Why it's notable: Iowa joined a growing list of states putting rules on insurers' use of AI in prior authorization, which gives clinics grounds to challenge denials that were never reviewed by a clinician.

On April 10, 2026, CMS proposed a rule (CMS-0062-P) that extends its 2024 electronic prior authorization requirements to drugs. It would apply to Medicare Advantage plans, Medicaid and CHIP, and ACA exchange plans, and it adds small-group exchange insurers. Payers would need FHIR-based prior authorization APIs for drugs. ACA exchange plans would have to decide within 72 hours for standard requests and 24 hours for expedited ones, and Medicaid within 24 hours. Payers would also have to give a specific reason for every drug prior authorization denial. Most provisions would take effect October 1, 2027, and comments were due June 15, 2026.

Why it's notable: The 2024 rule left drugs out, and drug prior authorization is one of the biggest paperwork burdens in clinics. This proposal would bring it under the same electronic process, deadlines and denial-reason rules.

On April 1, 2026, the FDA denied a petition from Sydney-based Harrison.ai. The company wanted developers that already had a 510(k) clearance to be able to market new AI tools in four radiology categories without premarket review, including triage (CADt), detection and diagnosis (CADe/CADx) software and medical image analyzers. The agency said clearing one device does not show that a company's processes will produce safe future devices, and that postmarket monitoring cannot replace premarket review. It pointed developers to predetermined change control plans instead. The FDA finalized the denial in the Federal Register on September 17, 2026.

Why it's notable: New radiology AI tools, and major new versions of them, will still need FDA review before they reach health systems. Buyers should check which software version is actually cleared.

On January 6, 2026, FDA Commissioner Marty Makary announced at the Consumer Electronics Show that the agency was relaxing oversight of some digital health products. The same day the FDA issued revised final guidance on clinical decision support (CDS) software, replacing its 2022 version, and on low-risk general wellness products. Under the CDS guidance, the FDA will generally not regulate software that gives clinicians a single recommendation, such as a specific drug to consider, if the software meets its other non-device criteria. The wellness guidance lets non-invasive wearables report readings such as blood pressure and blood glucose without FDA review, as long as they are marketed only for wellness and make no medical claims. The guidances were issued without a prior public comment period.

Why it's notable: More AI decision-support tools and consumer wearables can now reach clinicians and patients without FDA review, so validating them falls more on the health systems and clinics that adopt them.

On January 6, 2026, Utah's Office of Artificial Intelligence Policy and the startup Doctronic announced a 12-month pilot, set up through the state's regulatory sandbox, in which Doctronic's AI renews routine prescriptions for chronic conditions such as diabetes and hypertension. Reports put the list at about 192 drugs. Controlled substances, ADHD medications and injectables are excluded, and the first prescription must still come from a human clinician. The state agreed not to enforce its unprofessional-conduct rules against Doctronic as long as the company follows a contract with safety and privacy terms. Physicians review the AI's decisions at first, and uncertain cases are sent to clinicians. Doctronic told regulators its system matched physician treatment plans in 99.2% of 500 urgent care cases it reviewed.

Why it's notable: It is the first state-approved program in the US to let an AI system renew prescriptions on its own, which makes it a test case for how states may license autonomous clinical AI.

On December 22, 2025, HHS's Assistant Secretary for Technology Policy/Office of the National Coordinator (ASTP/ONC) proposed the HTI-5 rule. It would remove 34 of the 60 criteria in the ONC Health IT Certification Program and revise 7 others. One revision cuts back the Biden-era 'model card' transparency requirements for AI decision support tools in certified EHRs. The rule would also change the information blocking rules: it adds 'autonomous AI systems' to the definitions of access and use, drops the TEFCA manner exception, and narrows exceptions that let EHR developers limit third-party access. HHS estimates savings of $1.53 billion, and the rule was opened for 60 days of public comment.

Why it's notable: The certification program defines what every certified EHR must do, so this changes which features vendors are required to keep and how much they must disclose about the AI built into them. The information blocking changes are aimed at making it harder for EHR vendors to shut out AI tools and third-party apps that want patient data.

On December 1, 2025, the FDA said it had deployed agentic AI tools, which can plan and carry out multi-step tasks, to all of its employees. The agency listed uses including meeting management, premarket reviews, review validation, postmarket surveillance, inspections and compliance. Use is voluntary, and the models run in a secure GovCloud environment. The FDA says they do not train on staff inputs or on industry submissions. It also says all AI output is checked by staff before it goes into any regulatory action. The rollout builds on the Elsa generative AI tool launched in May 2025, which the agency says more than 70% of staff use.

Why it's notable: The regulator that clears drugs and devices is now putting AI agents into its own review work. That could affect how fast, and how consistently, applications from device makers and drug sponsors are handled.

CMS named the six companies that will run prior authorization reviews under its WISeR (Wasteful and Inappropriate Service Reduction) model in traditional Medicare: Cohere Health (Texas), Genzeon (New Jersey), Humata Health (Oklahoma), Innovaccer (Ohio), Virtix Health (Washington) and Zyter (Arizona). The vendors use AI and machine learning to review medical necessity for a short list of services CMS considers prone to waste, such as skin substitutes, electrical nerve stimulators and knee arthroscopy for osteoarthritis. Vendors are paid based on the Medicare savings they produce, adjusted for performance measures such as provider experience. Democrats in the House introduced a bill to block the pilot, but the model launched as planned on January 1, 2026, and runs through 2031.

Why it's notable: This brings prior authorization, run by AI vendors paid according to savings, into traditional Medicare for the first time. Clinicians in the six states now need prior approval for services that used to go through without it.

When Congress failed to pass FY2026 funding and the federal government shut down on October 1, 2025, Medicare's pandemic-era telehealth flexibilities expired with it. Medicare telehealth coverage for most services was again limited to rural patients seen at approved originating sites instead of at home, audio-only coverage ended, and CMS's Acute Hospital Care at Home waiver, used by more than 400 facilities in 39 states, also lapsed. Behavioral health telehealth was largely protected by permanent law, and CMS said providers could hold telehealth claims while Congress acted. The spending bill that ended the shutdown in November restored the flexibilities through January 30, 2026, with retroactive payment for services delivered during the lapse.

Why it's notable: For six weeks, practices and hospital-at-home programs could not be sure Medicare would pay for virtual visits, which exposed how much U.S. telehealth still depends on short-term extensions.