Medscape's New DPC Column Puts the Failure Rate at 10%. It Doesn't Say 10% of What.

The most reassuring number a physician considering direct primary care can hear ran in Medscape on Oct. 19: a 10% failure rate.

It comes from Doug Farrago’s column “Direct Primary Care: A New Hope for Family Medicine,” which Medscape sets up as a recurring feature. An editor’s note invites readers to send DPC questions for future columns. Farrago is a retired family physician who helped found the DPC Alliance and now runs DPC News, a relationship the column discloses.

“I conducted a year-long study on 59 failed DPC practices and found a 10% failure rate,” he writes. “That is extremely impressive as compared to other industries, but it is not guaranteed you will succeed.”

The column doesn’t say what the 10% is a share of.

Two counts, three years apart

Farrago has been studying closures for a while, and he has described the work in public before.

In a November 2023 Medscape feature by Jennifer Nelson, he said he had “interviewed about 60 of the 155 failed DCP practices that he found in his research.” The most common reasons, per that article, were poor or stagnant growth, personal or family health problems, pricing, personality issues and a physician whose heart wasn’t in it.

Four months later, on DPC News in March 2024, he wrote that he had spent nine months researching why practices close. He declined to share the findings there. “Am I going to give them here? No,” he wrote, saving them for a keynote at that year’s DPC Summit.

So the figures line up this way:

• 155 failed practices Farrago said he had found, as of late 2023.
• About 60 of those he interviewed then.
• 59 failed practices in the year-long study he cites now.

The 59 is close to the roughly 60 interviews he described in 2023. Neither number, by itself, is a rate.

What a rate needs

A failure rate is a fraction. The top is closures. The bottom is every practice that could have closed over the same stretch of time. Medscape’s 2026 column gives the top and leaves the bottom out.

Here’s why the bottom matters. That same 2023 Medscape article cited an estimate of 2,060 DPC practices operating in the United States. If the 155 closures Farrago described that year are the numerator, a 10% rate implies about 1,550 practices underneath it. Measured against 2,060, by DPC Insider’s arithmetic, 155 would be about 7.5%.

That’s only a guess at how the figure was built, and Farrago may have used a different count, a different period or a different definition of closure. The point is that each of those choices moves the result, and the column doesn’t state any of them.

The current column also says there are “more than 3100 DPC practices” in the country, linking to DPC Frontier’s mapper. That map describes itself as a listing of “all known DPC practice locations,” and it includes hybrid practices that also bill insurance and on-site clinics that usually serve one employer’s workforce. A count of locations is a different unit from a count of practices, and a clinic that bills insurance alongside a membership might close for different reasons than a cash-only one.

There’s a quieter problem too. A practice that closes in its first year may never reach a public map or a conference hallway. A study that finds failures by asking around will tend to find the ones that lasted long enough to be known. That would push a count of failures down, not up.

The useful part was always the interviews

None of this makes Farrago’s work worthless. Tracking down dozens of physicians whose practices closed, and getting them to talk, is the harder half of the job.

The reasons he collected are concrete enough to plan around. Stalled growth and pricing are things you might model before you sign a lease. Health and family emergencies argue for a cash reserve and a backup arrangement for coverage. “Heart wasn’t in it” is a question you can only answer yourself.

The column’s own pricing history helps here. Farrago writes that he charged $75 a month for one person, $125 for a couple and $150 for a family. He filled his practice with 600 patients in 18 months after starting with $20,000 in credit card debt. That is one practice’s path, told by the person who walked it.

The open question

What DPC Insider couldn’t establish is the thing a physician weighing a launch most wants to know: out of every DPC practice that opened in a given year, how many were still open three or five years later?

The Medscape column doesn’t answer that. Neither does the 2023 article or the 2024 post. In 2024 Farrago said his findings were headed for a DPC Summit keynote, and the column’s 10% arrives without the count it divides by.

Farrago writes that he’ll use future columns to take on arguments against the model. A column on how the 10% was built would be a strong place to start. Until then, the 59 interviews may be worth more to you than the rate they produced.