top of page

Let’s have better conversations about the physician landscape: Five suggestions

The 'physician landscape,' meaning how physicians are organized, ownership and contract structures, and so on, is a particularly hot topic right now (and perpetually, if we're being honest, but especially right now). Key ongoing themes include the role of private equity in health services, the rise of concierge practice, and the impact of vertical integration.

It makes sense to focus on these and other physician-related trends because they deeply affect not only physicians themselves, but a whole host of related businesses—as go physicians, so goes healthcare access, quality, and utilization patterns, including downstream use of hospitals.

But in our experience, discussions of the physician landscape tend to be particularly difficult to navigate in a productive way—they can be ambiguous and ill-defined (at best) and heated and emotional (at worst): Imprecise terminology can lead to confusion and misunderstanding. Hypotheticals are stated as fact. Conversations are framed in moralistic terms, with various trends and players painted as wholly good or bad—depending on who’s talking, and often without a lot of data that specifically ties a physician ownership model to differences in outcomes.

As informed consumers of conversations about the physician landscape—and physician alignment trends in particular—we’ve observed several areas in particular where these discussions have a tendency to go off the rails. Here are five suggestions about how to keep physician landscape conversations on track, enabling smarter, more productive discussions about this critical part of the industry.

1. Any conversation about physician alignment should begin by specifying whether the focus is individual physicians or entire physician practices (or both).

I’ve been in plenty of healthcare strategy conversations about what “physicians” want: which types of corporate entities they are likely to align themselves with (or not), what they’re looking for from those potential partners, and how that alignment might impact the larger healthcare ecosystem—only to realize at some point that the discussion was missing important context about whether we’re talking about individual physicians or entire practices—a consideration that might lead to very different conclusions.

There are major differences in the incentives, motivations, and behaviors of individual physicians vs. entire practices. For example, a newly graduated physician in search of his or her first source of employment may have very different criteria for a choice of future employer vs. a large practice governed by physicians nearing retirement who are looking to sell. Or a particularly timely example: physicians as individuals likely support the FTC’s proposed ban on non-competes, which would make it easier for them to switch employers without having to move to a new geography and/or give up their patient panels. But physician practice leaders, at least while wearing their employer hat, likely oppose such a ban.

2. We no longer live in a binary world of “employed” and “independent” physicians; we all need to get more nuanced about practice ownership (and also, embrace a world of shades of grey).

Until very recently, the presumption was that a physician practice was either independent, or employed by a health system-owned medical group. But as other entities such as health plans and private equity firms have broadened their participation in the physician market, now we need more nuanced data.

We are just beginning to get data that separates out hospital employment and practice ownership from that of ‘non-hospital corporate entities.’ This is helpful in enabling us to see differences in behavior between hospitals and non-hospitals. Take a look at these figures on employment and practice acquisition trends from Avalere:

Data that breaks out physician employment by hospitals vs. non-hospital companies shows that even as the overall employment trend increased, hospital employment of physicians had essentially plateaued in the two years leading up to the pandemic, holding steady at around 47% of the physician market. But something—whether Covid-19 or the perceived need to keep up with non-hospital corporate entity employment of physicians—appears to have reinvigorated the trend. By mid-2021, over half of physicians were employed by hospitals.

And we can see that non-hospital firms now own a larger share of the physician practice market than hospitals (27% vs. 26%). This has taken hold rapidly: as recently as 2019, non-hospital corporations lagged hospitals in practice ownership by a full 10 percentage points.

All this is helpful; but at some point we are probably going to need to get even more granular. Not all hospital systems are the same, but they likely behave more similarly to one another than say, private equity behaves relative to an insurance- or retail-owned conglomerate. In the era of vertical integration and thoroughly diversified corporate entities it may be a lost cause to try to further categorize among them—but on the other hand, wouldn't it be good to know if insurer-dominated companies manage physicians differently from, say, PE or Amazon?

3. We need to scrutinize the impact of physician employment/ownership by non-hospital entities with the same level of data-driven rigor that we’ve applied to hospital-based health systems, rather than relying on hypotheticals or one-off examples.

If we can separate out more categories of ownership, we can start expanding the world of research studying access, quality, and cost results.

Because hospital employment of physicians and ownership of physician practices is the longest-standing trend, it’s where we have the highest-quality data at this time—i.e. academic studies, from credible sources, conducted in multiple markets/geographies. A review of this academic literature shows a lot of support for hospital employment of physicians having a noticeable upward impact on prices, and an increase in referrals to hospital-based settings, which are themselves higher-priced. On the other hand, there is little evidence that hospital employment has negatively impacted quality (but also, little evidence that it improves it).

Now we’ve entered a world in which plenty of observers are concerned about the potential for health plans and private equity to also shift physician behavior in ways that might similarly change cost and quality outcomes—but without the data to inform these conversations. If these structures are creating problematic access, cost, or quality issues, that’s something the industry and policymakers need to know. And if instead non-hospital corporate entities tend to be more focused on diversification than true integration (as we believe), then those practice acquisitions are unlikely to translate to wholesale changes to clinical practice.

4. Don’t assume that loss of ‘independence’ (in the ownership sense) necessarily equals loss of physician autonomy.

This so-called “corporatization” of medicine naturally prompts worry (and hope) about how physician alignment with larger entities will directly impact physician behavior. This is a longstanding historical issue here in the U.S.; insiders know both that 'corporate practice of medicine' is illegal in some states (and also, that physician integration with non-physician entities happens in those states anyway, just through affiliation models that do not include direct employment). Suffice it to say: there is broad agreement that studying these arrangements for positive or negative effects—and taking action if needed—is important.

But until and unless we have that concrete data to act on, it’s important to remember a few key caveats about physician behavior and ownership arrangements. Specifically: physician behavior, in total, is not that easy to inflect simply by virtue of any kind of contractual/employment/integration structure, including employment. Again, their autonomy, in some cases, is protected by law. Employers are constrained in many ways, including what incentives they can offer; to take only one example, it is not legal to directly incentivize physicians for actions designed to reduce cost of care for Medicare patients.

In general, physicians have a complex hierarchy and operate (no pun, sorry) by a complex set of rules. And they tend to be more willing to make some changes than others—which means that some effects are much likelier to be realized than others.

The easiest thing for an acquirer to change would be the business practices and operational behavior of a medical group: invest in marketing, make revenue cycle more efficient, etc. In many ways, this is the traditional PE playbook: streamline and scale where possible.

Referral behavior is more difficult to influence. Physicians choose referral targets reasons ranging from the clinically valuable (physicians have their own actual lived experience with quality) to the clinically dubious (personal relationships) to somewhere in between. Health systems certainly have managed it to some extent (as the data indicates), and as plans diversify their care delivery networks and acquire assets, adding urgent care centers and ambulatory surgery centers, presumably this will become their playbook too. They will use levers ranging from marketing the services on the basis of superior quality, lower cost, and better experience—to simply making it far easier for care to stay “in house.”

Hardest of all to influence is clinical behavior and care decision-making within a physician's own practice. Health systems have had some success with care variation reduction, and this is a big part of the potential health plan value proposition when it comes to vertical integration. But for good or for ill, considerable practice variation remains.

Are we saying there’s no need to worry about influence of ownership on physician practice patterns? Not at all. Just beware of foregone conclusions. Open the question, and insist on concrete data: Are practice pattern changes actually being observed?

5. Let’s all keep an eye on the trends that cut across multiple types of practice ownership and employment settings—for example, the growing interest in low-panel models.

So much of the commentary about the physician landscape is focused on trends in employment and ownership, and for good reason. But a laser focus on those dynamics risks missing out on some of the larger forces reshaping the physician landscape and practice patterns; some of which might arguably be even more important.

One of the most critical trends, to my mind, is the growing interest among physicians and multiple physician employers/owners in low-panel models. At a surface level, it’s easy to relegate this trend to the tiny corner of the market comprised of concierge medicine: physicians who deliver a high level of service and experience to patients, in exchange for a fee (either on top of, or in some cases, instead of, traditional insurance). To deliver that enhanced experience, these physicians accept a limited number of patients, often overseeing panels that are several orders of magnitude smaller than the industry standard of a couple thousand.

The (limited) market sizing numbers out there suggest that concierge models are a relatively small part of the market, which can make it an easy trend to dismiss (and an easy one to criticize, as one that only appears, at first glance, to be available only to wealthy consumers). But concierge models are not the only practices embracing lower panel sizes: “value-based” medical groups such as ChenMed and CareMore also emphasize that smaller panels are key to improving patient outcomes and reducing cost, while also enhancing the physician experience of practicing medicine. There’s also growing momentum behind more accessible “concierge-lite” type models which just got a huge shot in the arm with the Amazon' acquisition of One Medical.

I underscore this trend because it cuts across multiple types of ownership and practice settings. Many ‘VBC-oriented’ practice owners, from hospitals to insurer-owned corporate entities, to VC- and PE-backed companies, are experimenting with these models. And, it could meaningfully impact both cost and quality, potentially for the better (more patient attention, better management of complex conditions)—or potentially for the worse, if it is adopted at a scale large enough to reduce overall physician supply, and therefore access.


As always, we love to hear from you. What do you think the critical questions regarding the physician landscape are? What’s missing from the conversation today? You can reach me at, or our entire team at

Join our mailing list to see future posts

Thanks for submitting!

bottom of page