Keeping Oncology Practices Independent: Capital, Reimbursement, and Ambient AI with Jeff Patton, MD

Beyond the Chart Podcast Guest

Jeff Patton, MD

CEO, OneOncology
Executive Chairman of the Board, Tennessee Oncology

Key Insights from the Conversation

• OneOncology's affiliate model leaves practices in charge of their own clinics.
  Practices own 100% of their tax ID. Central services like revenue cycle are optional by design, so OneOncology has to prove its version is better before a practice will switch.
  
  
• Community practices deliver chemotherapy for far less and get paid less every year.
  On the commercial side, hospitals are paid about 300% more for the same drugs. The physician fee schedule has fallen about 30% against inflation. Dr. Patton sees 340B as "close to an entitlement" that funds hospitals more than patients.
  
  
• Veteran oncologists who dictated for years picked up ambient AI fastest. According to Dr. Patton, physicians who were used to Dragon found ambient documentation easier to learn than younger colleagues who had become fast on the keyboard.
  
  
• Most patients in an oncology office don't resemble the patients in the trials behind the guidelines. Dr. Patton points to comorbidities that exclude most real-world patients from trials. He sees AI-built patient cohorts as a way to guide their care.

What is the affiliate model in oncology?

Dr. Patton joined Tennessee Oncology in 1996, when the practice had no quality or compliance program. He was "volun-told" to build both, and later served as the practice's CEO for a decade. In 2018, Tennessee Oncology became the first practice to found and join OneOncology. After a stint as interim CEO, he was named permanent CEO more than six years ago.

"To me, independence means you own your practice and you and your colleagues set your own governance and your own kind of rules of engagement."

To Dr. Patton, independence means making your own clinical decisions and most of your business decisions — which in practice means owning your practice. OneOncology's affiliate model has a management relationship with its practices but owns no part of them. Decisions are made jointly, and OneOncology has input where it puts up the capital.

The founding practices, including New York Cancer & Blood Specialists, were never going to hand their operations to a private equity firm. What they wanted from a larger organization was aggregate buying power, shared best practices, and access to capital.

That respect for practice autonomy shapes how OneOncology offers central services. A practice can hand revenue cycle operations to OneOncology, which runs it at cost rather than charging a third-party margin. A practice can also keep running it itself. If a practice's collections lag, OneOncology will try to persuade it to switch, but it can't force it. As Dr. Patton puts it, "If you're still a functioning medical oncology practice in the world today, you got something right."

How does private equity help oncology practices grow?

Dr. Patton draws a line between private equity firms that break up assets and growth capital investors who are betting a business can outgrow its market. Most community oncology practices are PCs (Professional Corporation) or PLLCs (Professional Limited Liability Company) that zero out their books every year. That structure makes growth expensive for the partners.

A new physician who isn't yet busier than their salary dilutes everyone else's compensation in year one.
A PET/CT scanner or linear accelerator counts against the year's P&L, leaving less cash to distribute.

"It takes capital to underwrite growth. And in my opinion, if you're not growing, you're dying."

OneOncology uses private equity capital to absorb that risk, so practices can invest without cutting physician pay. Dr. Patton says practices typically earn back the management fee within the first year or two, and their incomes end up higher than before they affiliated.

Data was the other part of his 2018 pitch to Tennessee Oncology's partners. He told them "he or she who manages their data the best is gonna win," and that no single practice could afford the IT platform and analytics that would take. He didn't foresee AI or ambient listening, but he did foresee an information technology expense that would require scale. Matthew Ko sees the same pattern today: Structuring data well is becoming a competitive advantage for organizations, in oncology and beyond.

Why do hospitals get paid more for chemotherapy?

As Dr. Patton details, against inflation, the physician fee schedule has dropped about 30%, while hospital payments have risen about 4% a year. For chemotherapy on the commercial side, hospitals are paid about 300% more than community practices. On the government side, community practices are about 10% less expensive. As the gap widens, more physicians go to work for hospitals, which are the higher-cost site of service.

“We can provide the same service, the same drugs in a more convenient setting than a hospital at about a third of the cost on the commercial side.”

In terms of quality of care, Dr. Patton cites a Community Oncology Alliance (COA) study with Flatiron Health that found longer survival among community oncology patients with metastatic breast cancer and non-small cell lung cancer compared to national SEER benchmarks. On the topic of 340B, he has never met a community oncologist who disagrees with the program's original goal of giving patients access to advanced chemotherapy. Today, though, more than half of the chemotherapy given in the U.S. goes through a 340B program, and the savings flow to hospitals rather than patients. He considers the program "close to an entitlement" that hospitals now depend on.

Matt described this as an incentive pushing community oncologists toward hospital networks. Dr. Patton put it a different way: it gives hospitals an incentive to acquire distressed oncology practices.

When should cancer patients go to an academic medical center?

Dr. Patton believes patients deserve a choice: a community practice, a community hospital, or an academic medical center. Some communities have no independent option left, only the choice of "hospital or hospital." Matt noted that many families assume the best cancer care is wherever they see Super Bowl commercials. Dr. Patton has seen people in smaller towns pick whichever hospital is tallest.

He estimates that somewhere between 5% and 15% of patients need quaternary care at an academic center. Pediatric oncology belongs there because the cancers are rare. Highly technical surgeries belong there because surgical skill depends on volume: "The more Whipples you do, the better you are at Whipples."

For established therapies, though, "chemotherapy is chemotherapy."

Dr. Patton's advice to fellows is to train wherever they want, and then spend time in a community practice before choosing where to work. Some OneOncology practices run more clinical research than the academic centers in their area. But for physicians who want to teach and train, academic medicine is the right fit.

Which oncologists adopt ambient AI fastest?

There's a history of technology adoption in oncology and for Dr. Patton the starting point is drugs, oncology's longstanding main technology. Some oncologists change their standard of care as soon as an ASCO abstract comes out, and others wait years. The result is a bell curve — ambient AI adoption across the OneOncology network followed the same shape.

The adoption pattern people usually might expect for each generation didn't hold. Oncologists around Dr. Patton's age had spent years using Dragon and other dictation tools, so talking through a visit came naturally. Physicians who had become very fast on the keyboard felt ambient AI slowed them down. Dr. Patton predicts everyone will eventually use ambient AI, because it will be more useful than any other way of working with the EHR. (Larry Bilbrey, OneOncology's Senior Director of Digital Innovation, described a similar experience in his Beyond the Chart episode.)

Matt shared DeepScribe's own lesson. The company first built the most clinically accurate model it could, gave it to doctors, and heard "This sucks." What fixed adoption was adapting the technology to each clinician's workflow through customized models and experiences.

How do oncologists keep up with guideline changes?

When Dr. Patton was a chief medical officer, he audited charts by hand. He came away convinced that the best tool to manage physicians is data. Once a group agrees on a standard of care, a report card comparing each physician is "kind of public shaming, but it works." Physicians who believe they follow a standard often find in an audit that they don't.

"I've talked to thought leaders who just do breast cancer or just do lung cancer. That's all they have to keep up with, and they can't keep up with the information even within their discipline and their tumor type."

He thinks oncologists adopt new therapies faster than most specialists, partly because cancer care moves too fast to wait. For a general oncologist, decision support is now a necessity. Matt noted that the NCCN guideline for non-small cell lung cancer alone went through 11 versions in 2024.

Dr. Patton detailed a few instances of OneOncology helping their oncologists keep up. The organization helped Flatiron Health develop Flatiron Assist, which embeds OneOncology's clinical guidance in the EMR. When an NCCN guideline lists several options, Flatiron Assist narrows it to one recommendation. It can also show a payer's formulary or biosimilar preference before an order goes out, so prior authorizations aren't denied. OneOncology also gives physicians access to OpenEvidence inside the EMR.

What comes after guideline-concordant care?

Matt said oncologists tell him only about 15% of patients fit neatly into a guideline, and that the real goal is patient-centered care. Dr. Patton agreed that those patients aren't the ones he sees most. To enroll in a randomized trial, a patient has to be healthy enough to be "eligible to join the Marines;" most patients in a community practice have obesity, heart disease, diabetes, or another condition that would exclude them.

Dr. Patton believes there are enough patients with lung cancer, COPD, and diabetes to build real-world cohorts that could guide their care. Socioeconomic factors are known to affect outcomes, but no guidelines say how to adjust therapy for them. He hopes AI can help build datasets large enough to fill that gap.

Matt added a detail from a physician who wanted DeepScribe to flag when a Muslim patient was about to start fasting for Ramadan, since that could affect the timing of a new therapy. That kind of information rarely reaches a trial or the chart, but it shapes the decision. Dr. Patton offered safety checklists as another example: pilots run through them, physicians often skip them, and AI could do it for them. For Matt, the design challenge is making that information welcome rather than an interruption, as most doctors don't want to be told what to do.

What should technologists understand about oncologists?

Dr. Patton thinks most healthcare technology has been a hindrance to physicians. He wants oncologists to "get over their PTSD and embrace it's a new world" with AI and ambient listening. For people building the software, his advice is to start with how clinicians think.

"You've got to get in the skin of the providers and see what their workflow is, and see how they think and see how they make decisions. Design technology to them, not them to technology."

Matt agreed. He added that the technology available today is the worst it will ever be, and recalled the first time he shadowed a doctor for a day in clinic, being shocked by what it means to really understand a clinical workflow and see the work from an oncologist's viewpoint. Dr. Patton's response: "It's not what you think."

‍
More from Beyond the Chart: Exploring the AI Frontier for Oncology

Beyond the Chart features timely conversations with oncology leaders, clinicians, researchers, and innovators working to improve the future of cancer care. Subscribe to the podcast for more perspectives on how AI is impacting oncology today, and how it should in the future.

You may also like:

The Missing Data that Could Change Oncology | Dr. Caroline Chung, MD Anderson

How Do You Build Technology that Oncologists Actually Trust? | Lavi Kwiatkowsky, Canopy

‍

‍

‍

‍

‍

‍