What Hospitals Get Wrong When Hiring a Radiology Group

Choosing a radiology partner is one of the highest-leverage decisions a hospital makes. Radiology touches nearly every service line: the ED can't disposition patients without it, surgeons can't schedule without it, and oncology can't stage without it. Yet the selection process at most hospitals looks remarkably like buying a commodity — an RFP, a rate comparison, a few reference calls, and a contract that everyone hopes they won't have to read again for three years.

Having sat on both sides of that table over long careers in radiology and health system operations, we've watched the same avoidable mistakes repeat. Most of them trace back to a single blind spot: hospitals evaluate the proposal in front of them and ignore the ownership structure behind it.

Over the past decade, radiology has been through a wave of consolidation driven by private equity and venture-backed roll-ups. Independent, radiologist-owned groups were acquired by investor-owned platforms at a remarkable pace, and many hospitals signed contracts with those platforms without ever asking what the change in ownership would mean for them.

We now have enough history to answer that question, and it isn't encouraging. The investor-owned model has a structural problem that no amount of good intentions fixes: the people who own the practice and the people who read the studies are different people, with different incentives, on different timelines. Investors need a return, typically within a fund cycle. That return has to come from somewhere — and in a professional services business, "somewhere" is ultimately the radiologists and the hospitals they serve.

The mechanics are well documented across the industry. Acquisitions that pay radiologists an upfront sum are not generosity; they're the physician's own future income pulled forward and discounted, in exchange for permanently reduced compensation afterward. Staffing models get leaner. Compensation formulas get rewritten — including, most recently, schemes that quietly recapture the productivity gains from AI tools for the platform rather than the physicians using them. And when the economics stop working, the platform's obligations to its lenders come before its obligations to your medical staff.

None of this stays inside the group. It arrives at your hospital as turnover, as unfilled subspecialty coverage, as a nightly rotation of unfamiliar names, and — in the worst cases — as a contract held by an entity in financial distress.

The physicians themselves reached this conclusion first. Over the past few years, a genuine movement has formed inside radiology — national communities of thousands of radiologists organizing around physician empowerment, new radiologist-owned groups launching specifically as an alternative to investor-owned platforms, and a steady flow of physicians leaving those platforms to join or found independent practices. When the people with the best view of a business model are voting against it with their careers, hospitals should pay attention.

We'll say plainly what many hospital leaders have concluded privately: physician-owned, independent practice is the benchmark for alignment between imaging and the hospital. When radiologists control the organization, answer to their peers and their patients, and keep the full value of their own work, quality and service are not line items to be managed down — they are the whole business. Ask directly: who owns the group, who makes staffing and investment decisions, what happens to our contract if the group is sold, and what percentage of the group's economics goes to people who have never read a study? If the answers make the room uncomfortable, you've learned what you needed to know before the signature instead of after.

Every proposal you receive will promise excellent turnaround times. The numbers on the page will look nearly identical, because everyone knows what the RFP wants to hear.

The question that actually separates groups isn't "what's your TAT?" It's "what happens when you miss it?" Ask who reviews performance, how often, and with whom. Ask whether the group's physician leadership — not an account manager, not a regional VP three organizational layers from the reading room — will sit across from your medical executive committee and own the numbers. Ask for the metric definitions in writing: a group that measures turnaround from "assignment" rather than "exam completion" can hit any target you name while your ED physicians wait. But more importantly, ask about medical staff alignment and participation. Ask about who gets on the phone with physicians who have questions. Ask if the physicians are compensated fairly for their work.

Accountability follows ownership. A radiologist-owner answering for the group's performance is defending their own name and their own equity. An employed medical director answering for an investor's platform is delivering a message. Both may be sincere; only one can act.

Hospitals scrutinize the entity and ignore the individuals. But your emergency physicians, hospitalists, and surgeons don't work with a corporate entity — they work with specific radiologists, at 2 a.m., on cases where a phone call changes the plan.

Two questions reveal more than the entire RFP response. First: will we have named radiologists who consistently cover our facility, or a rotating national queue? A well-aligned group will volunteer the answer before you ask — providing biographies and direct contact information for the radiologists covering your facility, because it wants your medical staff to know its physicians by name. Consistency is how radiologists learn your medical staff, your protocols, and your patient population — and it's how trust forms.

Second: How will you gauge radiologist interpretation quality and hold it to a standard that is best for patient care? That second question is where ownership shows up again. The best radiologists in this market have abundant options, and the sustained migration of talent out of investor-owned platforms and back toward independent practice is one of the clearest signals in the field right now. High quality radiologists stay where they have a voice, a stake, and colleagues they chose. A group that can't retain physicians or recruits subpar talent will be re-introducing itself to your medical staff every year, or even worse missing critical findings that forces your medical staff to look elsewhere for answers.

"Full subspecialty coverage" appears in every brochure. The gap between the claim and the reality shows up in the details: who actually reads your PET/CT, your pediatric studies, your MSK MRI? During what hours? What percentage of studies in each subspecialty are read by a fellowship-trained radiologist in that subspecialty — and can the group produce that report from its own data?

If the answer involves phrases like "access to a network of," probe further. Access is not coverage. You want to know whether the subspecialist is in the workflow or on a brochure — and whether the group's economics allow it to recruit and keep subspecialists, or whether that line of the budget is where the investor's return comes from.

Nearly every group now claims to be AI-enabled, so the question isn't whether — it's how, and for whose benefit.

Ask which tools are deployed, validated against what evidence, governed by whom, and with what role for the physician on every final read. Then ask a question almost no hospital asks: when AI makes your radiologists more efficient, where does that efficiency go? In investor-owned platforms, the honest answer is increasingly "to the owners" — productivity gains are measured, work units are repriced downward, and radiologists run faster to stand still. That treadmill is not your problem on paper, but it becomes your problem in practice: it burns out the physicians reading your studies and drives the good ones toward the exits.

In a physician-owned group, the same technology points the other direction. Efficiency gains belong to the people doing the work, which means AI gets deployed where it genuinely improves quality, speed, and radiologist focus — not where it maximizes extraction. The quality of a group's answer on AI governance tells you how it makes decisions generally.

While you're in technical due diligence, bring your CIO in early. Ask how the group integrates with your specific EHR and PACS, how long their last three implementations actually took, and how they handle downtime. A modern, cloud-based group should describe onboarding in weeks and demonstrate it with references. A group carrying decades of legacy infrastructure — or a platform mid-way through integrating its latest acquisitions — will describe a "project."

The per-read rate is the most visible number and the least meaningful one. The real costs of a radiology relationship live elsewhere: length of stay added by slow reads, ED throughput lost to disposition delays, referral leakage when your medical staff loses confidence, recruitment problems when surgeons can't get the imaging partnership they expect, and the disruption of a re-procurement when a distressed platform stumbles or sells your contract to the next buyer.

Investor-owned groups can sometimes quote a lower rate, precisely because the model is built to compress the cost of the professional work. But a rate that is achievable only by underpaying and overloading radiologists is not a savings — it's a deferred cost, and hospitals pay it with interest in service failures and turnover. Model the decision the way your CFO models everything else: total impact over the life of the relationship, not unit price.

Every mistake on this list reduces to the same root error: evaluating a radiology group as a vendor rather than as a clinical partner — and evaluating the proposal while ignoring the ownership behind it.

Vendors are selected on price and promises. Partners are selected on alignment: of incentives, of governance, of the people who will actually show up. In radiology, alignment has a simple test. When the physicians reading your studies own the practice, control its decisions, and keep the value of their own work, their interests and your hospital's point the same direction. When the practice exists to generate a return for owners who will never meet your medical staff, they don't — and no contract language fully papers over that gap.

The good news is that alignment is checkable, and the well-aligned groups tend to share visible markers. They put their radiologists forward by name, with bios and direct lines to your medical staff. They carry their own malpractice coverage — a group that stakes its own insurance on its physicians is telling you something about its hiring standards. They flex to your actual needs, whether that's full coverage, partial coverage, overnight, or subspecialty overflow, rather than forcing you into the contract structure that suits their platform. And they can onboard in weeks, because modern cloud infrastructure and clean HL7 integration were designed in from the start.

So, ask who owns the group. Ask who will read your studies and whether they'll still be there in two years. Ask where the AI efficiency gains go. Ask what happens when performance slips, and what happens if the group is sold. Ask your CIO and your medical staff to interrogate the details before the signature, not after.

The groups that welcome those questions are the ones worth hiring. The groups that deflect them have already answered.

This article is part of a series from the founders of Sublimity Radiology, a 100% physician-led, physician-owned, independent radiology practice headquartered in Greensboro, North Carolina. We believe thriving independent practice is the benchmark for the field — and we built our practice to prove it. Learn more at sublimityradiology.com.