By Anna Wilde Mathews, Christopher Weaver, Tom McGinty & Mark Maremont

Millions of times each year, insurers send nurses into the homes of Medicare recipients to look them over, run tests and ask dozens of questions.

The nurses aren’t there to treat anyone. They are gathering new diagnoses that entitle private Medicare Advantage insurers to collect extra money from the federal government.

An investigation of insurer home visits found the companies pushed nurses to run screening tests and add unusual diagnoses, turning the roughly hourlong stops in patients’ homes into an extra $1,818 per visit, on average, from 2019 to 2021. Those payments added up to about $15 billion during that period, according to an analysis of Medicare data.

Nurse practitioner Shelley Manke, who used to work for the HouseCalls unit of UnitedHealth Group, was part of that small army making home visits. She made a half-dozen or so visits a day, she said in a recent interview.

Part of her routine, she said, was to warm up the big toes of her patients and use a portable testing device to measure how well blood was flowing to their extremities. The insurers were checking for cases of peripheral artery disease, a narrowing of blood vessels. Each new case entitled them to collect an extra $2,500 or so a year at that time.

But Manke didn’t trust the device. She had tried it on herself and had gotten an array of results. When she and other nurses raised concerns with managers, she said, they were told the company believed that data supported the tests and that they needed to keep using the device.

“It made me cringe,” said Manke, who stopped working for HouseCalls in 2022. “I didn’t think the diagnosis should come from us, period, because I didn’t feel we had an adequate test.”

Other nurses interviewed, said many of the diagnoses that home-visit companies encouraged them to make wouldn’t otherwise have occurred to them, and in many cases were unwarranted.

Last month, it came to light, that insurers received nearly $50 billion in payments from 2019 to 2021 due to diagnoses they added themselves for conditions that no doctor or hospital treated. Many of the insurer-driven diagnoses were outright wrong or highly questionable.

The diagnoses added after home visits accounted for about 30% of that total. More than 700,000 peripheral artery disease cases diagnosed only during home visits added $1.8 billion in payments during that period.

In the Medicare Advantage system—conceived as a lower-cost alternative to traditional Medicare—private insurers get paid a lump sum to provide health benefits to about half of the 67 million seniors and disabled people in the federal program. The payments go up when people have certain diseases, giving insurers an incentive to diagnose those conditions.

To find out how insurers use home visits to add diagnoses; nurses, patients, home-visit managers and industry executives were interviewed and hundreds of pages of internal documents from home-visit companies were reviewed. They described a system that used nurses, software and audits to generate diagnoses.

“They do the job with a purpose, and it pays off for the Medicare Advantage plans,” said Francois de Brantes, a former executive at Signify Health, a company that does home visits for insurers. “Identifying the diagnoses, that’s the job.”

Insurers, including UnitedHealth and CVS Health, owner of both Signify and Aetna, said the house calls help patients by, among other things, catching diseases early and making sure people are taking their medicine properly. The insurers said they relay home-visit findings to primary-care doctors.

Nurses who made visits said they felt they were helping some patients with advice about medications, performing needed tests and sometimes reporting health emergencies.

For UnitedHealth, the parent company of the largest Medicare insurer, each home visit was worth about $2,735 in extra Medicare payments during the three years covered by the data. That’s nearly three times the average for all other Medicare Advantage insurers.

UnitedHealth’s chief physician, Dr. Wyatt Decker, attributed the disparity to what he said was UnitedHealth’s sicker patient population and its nurse practitioners being so effective at their jobs.

Sixty percent of UnitedHealth home visits generated at least one new revenue-producing diagnosis of a condition no doctor was treating, the analysis showed. Home visits by Humana, the No. 2 Medicare insurer, did so 39% of the time.

Medicare data covering the home visits was reviewed under a research agreement with the federal government. The data doesn’t include patients’ names, but covers details of doctor visits, hospital stays, prescriptions and

The home-visit industry has grown in recent years. UnitedHealth’s HouseCalls sent nurse practitioners to the homes of more than 2.7 million people last year. CVS’s Signify performed about 2.6 million home visits in 2023.

Step one is getting Medicare Advantage recipients to agree to a visit, especially patients whom insurers deem most likely to have undiagnosed conditions that would garner extra payments.

Two former managers who oversaw home visits for Humana, and a third who worked for both Humana and Signify, said insurers used an internal scoring system to identify prospects. Under the Medicare Advantage system, diagnoses have to be documented every year to trigger the extra payments, so people who had an earlier home visit that produced extra payments were particularly valued, the managers said.

Insurers also considered other factors, including how likely patients were to agree to a visit, some home-visit executives said.

Call centers bombard Medicare recipients with offers of home visits—in the case of Humana, autodialing them as many as 10 times, according to the former managers. Agents sometimes offered the Medicare recipients incentives such as Walmart gift cards.

A Humana spokesman said the company is committed to accurately identifying patients’ health conditions, and that its home-visit vendors don’t use software to suggest diagnoses for its patients.

Dave Sherwood, a 68-year-old retired accounting executive in Williamsburg, Va., joined a UnitedHealth Medicare Advantage plan last year. Earlier this year, he got a flurry of calls from representatives of the insurer. When he finally answered one, he told them he didn’t want a home visit.

Months later, he started getting the calls again. Finally, he said, he picked one up and told the agent to stop calling.

When patients agree to a visit, home-visit companies send nurse practitioners or, less frequently, doctors or physician assistants. Some are full-time employees, others contractors who get paid around $100 or more per visit.

At each home, the nurses run through a series of questions covering medical history and medications, as well as doing a physical assessment and some basic testing.

In the HouseCalls system, nurses feed the information into a laptop or tablet, and the software suggests diagnoses. They automatically appear in a “diagnosis cart” on the side of the screen, according to training documents from last year that were reviewed.

Kristen Bell, a nurse practitioner who left HouseCalls in May after doing home visits for seven years, said the prompts were one way to prod nurses to add diagnoses. They also got regular training about conditions they could record, she said. She characterized the message from management as: “I’m not going to beat you up about this, but I want you to go in this direction.”

Secondary hyperaldosteronism, a condition in which levels of the hormone aldosterone rise, is rarely diagnosed in traditional Medicare patients. HouseCalls documents show that its software would suggest the diagnosis if a patient had a history of heart failure or cirrhosis, and either took certain drugs, such as diuretics, or had swelling due to fluid retention. Nurses weren’t required to confirm the diagnosis with a lab test.

“In a million years, I wouldn’t have come up with a diagnosis of secondary hyperaldosteronism,” said Bell, the former HouseCalls nurse.

UnitedHealth diagnosed it 246,000 times after home visits, leading to $450 million in payments over the three years of analysis. All other Medicare insurers combined collected $42 million from making that diagnosis after home visits.

“It is a vastly underdiagnosed condition that is super valuable to call out,” said UnitedHealth’s Decker.

To find more cases of peripheral artery disease, both HouseCalls and Signify used the testing device that nurse Manke said she distrusted, company manuals show.

The Food & Drug Administration said the device, called the QuantaFlo, “is not indicated for use as a stand-alone diagnostic device but as an adjunct to the diagnostic process.” Medical guidelines recommend against widespread screening for the condition.

A HouseCalls training manual advised nurses to diagnose peripheral artery disease based on results from the device. Managers at Signify told nurses they were required to use the device to test the patients of most insurers, a 2020 email shows. Nurses diagnosed the condition after 568,000 home visits to UnitedHealth patients in the period analyzed, adding up to nearly $1.4 billion in additional payments.

UnitedHealth’s Decker said the company expected clinicians to use their judgment in making peripheral artery disease diagnoses. A spokesman for Signify-owner CVS said medical providers decide when the test is appropriate on a case-by-case basis, and that the 2020 email was “not clearly worded” and didn’t reflect company policy.

Renae Cormier, chief financial officer of QuantaFlo maker Semler Scientific, said the device assesses risk for the disease.

Dr. Amy Chappell, a 73-year-old neurologist in Naples, Fla., was surprised when a nurse sent to her house earlier this year by UnitedHealth pulled out a QuantaFlo device. “She had no reason to think I had peripheral artery disease,” said Chappell, who says she has had no symptoms of the condition and is an avid runner and tennis player.

Chappell tested positive, although the nurse didn’t do any other standard exam to check for symptoms of the disease, Chappell said. Her primary-care doctor, Dr. Rebekah Bernard, said in an email that the diagnosis was inaccurate.

“From what I do know of the case, it’s an example of where we could have done better, and we need to own that,” said UnitedHealth’s Decker. The company later confirmed the diagnosis was a mistake and said it corrects such errors.

The Medicare Payment Advisory Commission, a nonpartisan agency that advises Congress, has recommended that diagnoses from home visits shouldn’t count toward extra payments to Medicare insurers. The inspector general that oversees the Medicare agency has said it should reconsider the use of such diagnoses.

A spokeswoman for the Centers for Medicare and Medicaid Services, said the agency recently ramped up audits to verify diagnoses. The agency also is eliminating some diagnoses from those that qualify for extra payments, including peripheral artery disease.

Nathanael Lacaria, a nurse practitioner who did home visits in Colorado for CVS’s Signify unit in 2019 and 2020, said he didn’t feel it was appropriate to make definitive diagnoses based on one visit.

He said he didn’t realize insurers were submitting some of his tentative diagnoses to Medicare for billing purposes until a woman he had visited called the company to complain. “What’s this depression diagnosis in my chart?” she asked, according to Lacaria.

When he visited the woman, whose husband had died, Lacaria said, he recorded her answers to a standard depression screening tool, but he hadn’t actually diagnosed depression. “These visits were definitely used to jump to conclusions I wouldn’t have arrived at,” he said.

The CVS spokesman said Signify’s clinicians independently determine which conditions a patient has.