This is a Physician in the Loop special topic.

Equal pay for equal results

Medicare now pays technology firms for results, and a smart-ring maker will promote one of them to its members. Primary care deserves the same deal, or Medicare will pay twice for fragmented care.

On September 29 investors are due to put a price on Oura, the maker of a smart ring with 5 million paying members. At the top of its range the offering would value the company at $14.1 billion, and it has drawn about four times as many orders as there are shares. Oura's prospectus describes a market that "begins with wearables and extends into preventative healthcare," and part of that extension runs through Medicare. On September 17 Counsel Health, which calls itself "the AI-native primary care company," said it was taking part in ACCESS, a Medicare program that pays for measurable results in chronic disease. From early next year its doctors, working with its AI, will treat high blood pressure, obesity, high cholesterol and prediabetes at no out-of-pocket cost to eligible patients. Oura will promote the service in its app, and its chief medical officer says the program will let the company "study that model at scale."

The Counsel deal is small beside Oura's listing, but it belongs to a pattern. Whoop and Withings, two other makers of consumer health devices, have physician groups on Medicare's list of accepted participants, alongside WeightWatchers, Noom and Alphabet's Verily; about 160 organizations are taking part, and health plans covering 165 million Americans have pledged to pay in a similar way. A second front door to chronic care is being built, and it opens on a phone. Medicare pays the firms behind it for results and keeps paying the patient's own primary care practice by the visit. Paying for results is the right idea, but paying only newcomers for them will fragment care, add to spending and leave primary care with the patients hardest to help. Medicare should offer practices the same terms.

ACCESS, which began in July, pays a fixed yearly fee for each patient, in monthly installments, and holds back half of Medicare's share until the year's results are in. In Counsel's track the fee is $360 for a patient's first year. To be paid in full, at least half of a firm's patients must finish the year and meet their targets, such as a systolic pressure below 130 or a drop of 15 points. Firms may waive the part of the fee that patients normally owe, which lets them offer care at no out-of-pocket cost. Patients sign up by phone or online, without a referral; their doctor gets an update once care begins.

Medicare assures primary care doctors that the firms "do not replace your role as the patient's primary clinician." It pays that role far less. A doctor who reviews a firm's update and acts on it may bill $30, up to three times a year, which, with a small one-time fee for helping a patient get set up, comes to little more than a quarter of the firm's first-year fee. The visits at which she checks the same blood pressure are still paid one by one, from a fee schedule whose basic rate Medicare has proposed to lower next year. And the firm's fee shrinks if too many of its patients get listed services for the same condition from other clinicians, so care that another clinician starts, the family doctor included, can count as duplication.

A practice could in principle join ACCESS itself, since almost any organization enrolled in Medicare's Part B may apply. But a participant, and any practice affiliated with it, may not bill Medicare's ordinary fees for its enrolled patients during their year of care, and claims systems will block such bills automatically. The rule suits a firm with no clinic. A practice that sees the same patients in person would have to give up their visit fees to collect the outcome payment. Aledade, whose network is made up of primary care practices and health centers, has two entities on the list. Still, the design favors the firm that reaches a patient's phone over the doctor who knows her history.

Dr. Mehmet Oz, who runs Medicare and Medicaid, has been frank about how payment shapes what technology does. "Short term, AI is going to be inflationary because it's going to turbocharge the ability of the current billing systems to work more effectively," he told an Oracle conference in September. The Blue Cross Blue Shield Association says hospitals' AI coding added $942 million to its plans' costs over two years with no matching change in care, which hospitals dispute; Hartford HealthCare says autonomous coding, piloted with ambient documentation, raised its primary care revenue by 17 percent, and complaints rose with patients' bills. At an event for ACCESS on September 15, Dr. Oz said Medicare was paying "for outcomes, not for effort or process or some surrogate," and that "the most expensive health care is bad quality care."

He is right on both counts, which makes the design of ACCESS odd. In practice it hands the outcome contract to newcomers and leaves the practices that already manage these diseases on the fee-for-service billing he expects AI to inflate. It even reaches into his preferred remedy, the accountable care organization, which is rewarded when its patients' total costs fall. From 2028 the firms' fees will count in the benchmarks and spending by which such organizations are judged, and analysts at Milliman, an actuarial firm, call that a risk for them unless the firms' patients end up costing less.

Whether the firms earn their fees depends on what they do. The Peterson Health Technology Institute, an independent evaluator, found that hypertension programs that manage patients' medication lower blood pressure quickly and meaningfully, better than usual care, and should save money in the long run; programs that only transmit readings or coach behavior do not, and all of them raise spending at first. In its review, digital diabetes tools showed no meaningful clinical benefit and higher net spending. The evidence Counsel offered with its announcement was about satisfaction: members who saw its doctors rated their care 4.6 out of 5. Satisfaction matters, but Medicare is paying for blood pressure.

History suggests that convenient care is more often added to other care than substituted for it. When RAND studied direct-to-consumer telehealth for respiratory infections, only about one visit in eight replaced a visit elsewhere, and spending rose. Enrollment in Medicare's remote patient monitoring grew more than tenfold in three years, often without all the parts of the service. The Congressional Budget Office found that Medicare's innovation center, which runs ACCESS, raised federal spending by $5.4 billion in its first decade instead of cutting it. On the evidence, ACCESS money is well spent when a firm adjusts the drugs of patients whose disease is out of control, and wasted on monitoring, coaching and convenience for patients who were doing well.

Paying for results has its own trap: it rewards whoever enrolls the patients easiest to improve. The companies say a fifth of Oura's members aged 65 and over are obese; among all Americans aged 60 and over, the CDC counts nearly two in five. People who buy a smart ring and pay for a membership are not typical of Medicare's patients, and a firm that recruits them will hit its targets more easily than a practice that treats everyone, including the patient with five diseases and no smartphone. Yet ACCESS asks the same share of patients to reach their targets whoever a participant enrolls. Unless that changes, Medicare will be paying a bonus for selection.

The strongest objection is that primary care has had its chance: only about one American adult in four with high blood pressure has it under control. If a firm with a ring, an app and doctors on call does better, patients gain. Eighteen clinical and patient societies have backed Medicare's push for technology-supported chronic care, and the American Medical Association's chief executive called ACCESS "an important step." That is a case for paying for results wherever they are achieved. A payment that makes a firm control blood pressure can make a practice do so too, and the practice already has the patient's history, medication list and trust. Primary care gets about 3.4 percent of Medicare's spending; paying others for its most measurable work will not raise that.

Medicare should let practices earn the same outcome payment for their own patients while still billing for their visits. It should adjust targets for the patients each participant enrolls, and show in the directory of results it plans to publish how each participant's patients were recruited. It should require firms to tell a patient's clinician before taking over her blood pressure, and pay enough for acting on their updates to cover the work. Before the fees count against accountable care organizations, it should publish an independent count of what ACCESS adds to total spending. Counsel and Oura should say whether and how Oura is paid for promoting a Medicare service; their announcement did not. And doctors whose patients enroll should read the updates, bill for acting on them and remain the clinician who sees the whole patient.

Dr. Oz has asked why Medicare should have doctors "check boxes" when it could ask whether they found a problem, treated it and made it better. Under ACCESS a firm that met its patient through a ring will be asked that question and paid for the answer. The doctor who found the problem in the first place will still be checking the boxes.

This special topic was read by an AI voice. Its sources are linked at physicianintheloop.org.