Physician Referral Management: How to Build a Referring Physician Strategy

Why referring physicians decide your appointment volume

Your schedule is filled by other people's decisions. A primary care doctor, a specialist or an urgent care clinician evaluates a patient, decides imaging is needed, and chooses where to send them. That choice is the single largest determinant of your appointment volume, and it is made outside your building.

That dependence creates an uncomfortable asymmetry. You can control your equipment, your staffing and your turnaround times, but the demand side of your business sits with a few dozen practices whose loyalty you have to earn continuously. A referring physician strategy is how you stop leaving that to chance.

There is a second reason it matters now. Referring physicians have more choice than they used to, and switching is cheap for them. If your intake is slow, your results are late, or their office staff cannot reach a human, the next referral goes somewhere else and nobody tells you why.

What counts as a referring physician

A referring physician, sometimes called the ordering physician, is the clinician who evaluates a patient, determines that imaging is needed, and directs them to your center. Usually that is a primary care doctor, but it is just as often a specialist, an urgent care provider or a nurse practitioner with ordering authority.

The practical definition matters more than the formal one: anyone whose signature puts a patient on your schedule is a referral source, and belongs in your strategy.

Who is allowed to order an imaging exam?

Medicare answers this narrowly, and the answer decides who belongs on your referral list. Under 42 CFR 410.32(a), diagnostic tests must be ordered by the physician treating the beneficiary, meaning the clinician who uses the result to manage that patient's specific problem. Paragraph (a)(2) then extends the same status to nonphysician practitioners operating within their scope under state law and within their Medicare benefit. That list includes nurse practitioners, physician assistants, clinical nurse specialists and nurse-midwives.

The practical consequence is a reporting one. If your referral reports roll nonphysician practitioners up under a supervising physician, you are hiding the people who actually sign your orders, and you will thank the wrong person.

What referring physicians actually want, and what most centers give them

Closing that gap is what a referral strategy is for. Referring physicians rate accuracy above everything else in a radiology report, and they mostly assume they will get it. What they report as the actual problem is timeliness, and everything around the report that their office has to chase.

The pattern is consistent: every row is an operations problem, not a marketing problem. You can buy lunch for a practice every week and still lose its referrals to a competitor whose portal is faster.

What do referring physicians actually complain about?

There is real published evidence here, and it is narrower than most vendor material implies. Two studies are worth knowing, and both point at the same thing. Neither is recent and neither is specific to an outpatient imaging center, which is precisely why they are worth reading carefully rather than quoting as a benchmark. What they establish is the shape of the problem, and that shape has been stable for twenty years.

What does the research say about the report itself?

Johnson and colleagues surveyed physician faculty across a large academic medical center about what makes a radiology report good. More than 95 percent gave the top importance rating to "accurate." Between 78 and 83 percent gave the top rating to "clear," "complete" and "timely." Their conclusion is the useful part for an imaging center: the single greatest problem area in reporting is lack of timeliness, and overall satisfaction with reporting left room for improvement.

Read that as a warning about assumptions. Referrers do not treat accuracy as a differentiator because they expect it. They experience the difference in how long they wait.

What does the research say about the referral process around it?

Gandhi and colleagues surveyed primary care physicians and specialists at an academic medical center about the outpatient referral process. 63 percent of primary care physicians and 35 percent of specialists were dissatisfied with it. The leading complaint from primary care physicians was lack of timeliness of information coming back, cited by 54 percent. Four weeks after the referral visit, 25 percent of primary care physicians had still received nothing.

Both studies are single-site and old. Treat them as direction, not as a benchmark. The safe reading is that the failure mode is information not arriving, and that it has been the failure mode for a long time.

Sources: Johnson AJ et al., Journal of the American College of Radiology, 2004. Gandhi TK et al., Journal of General Internal Medicine, 2000;15(9):626-631.

The five components of physician referral management

The five components run in order, and each one depends on the one before it. Goals tell you what to count, and tracking tells you what is happening. Marketing gives you something to test, and communication is what referring offices actually experience. Relationship building is what you do with the time the first four save you. Skipping the middle two is why most referral programs cannot say whether they worked.

Goal setting

Start with numbers you can be wrong about. How many new referring practices do you want this year, how much recurring volume do you want from existing ones, and what does a referral from each segment actually earn you? A strategy without a target is an outreach habit.

Referral tracking

Track referrals by individual physician, not by practice. Two doctors in the same group can have opposite referral patterns, and practice-level reporting hides both the champion and the defector. You need to see volume by physician, by exam type and by month, in something that updates without anyone exporting a spreadsheet.

Marketing

Print mailers, a provider-facing section of your website, trade shows, email, and physician liaison visits all work. Which of them works for you is an empirical question, and you will not answer it without the tracking above.

Communication

The point of communication is not to stay top of mind. It is to make sure a referring office is never wondering what happened to their patient. Confirmations, results and exceptions, delivered without being asked, do more for retention than a newsletter.

Relationship building

The tactics are unglamorous and they work:

• Introduce yourself and your center to referring practices in your catchment before you need anything from them.

• Ask about the referral process from their side, and fix what they name.

• Thank people specifically. "Thanks for the 14 referrals last quarter" lands differently from a branded mug.

• Stay in touch on a schedule, not on a mood.

• Make referring easy. Online scheduling, a referral portal and same-day results remove more friction than any amount of goodwill.

Where does federal law draw the line on referral marketing?

Two federal laws sit underneath every referral relationship in US healthcare, and the marketing literature on this topic almost never mentions them. Nothing in this section is legal advice. It is here to show you why a referring physician program needs your compliance officer and qualified healthcare counsel involved while the plan is being written, rather than after it ships.

Why does the Stark Law apply to imaging in particular?

The physician self-referral law names imaging directly. At 42 USC 1395nn(h)(6)(D), designated health services include "Radiology services, including magnetic resonance imaging, computerized axial tomography scans, and ultrasound services." HHS OIG describes the law as "a strict liability statute, which means proof of specific intent to violate the law is not required." Good intentions are not a defense, which is why the analysis has to be done by someone qualified to do it.

What does the Anti-Kickback Statute add?

The Anti-Kickback Statute at 42 USC 1320a-7b(b) is criminal, and it turns on intent. It reaches anyone who "knowingly and willfully" offers, pays, solicits or receives remuneration, in cash or in kind, to induce referrals for services payable by a federal health care program. The statute provides for a fine of not more than $100,000, imprisonment of not more than 10 years, or both. Safe harbors exist, and OIG is explicit that an arrangement "must fit squarely in the safe harbor and satisfy all of its requirements."

Is there a dollar limit on what you can give a referring physician?

There is a figure inside one narrow Stark exception, and it is not a marketing budget. CMS updates it every year by CPI-U and publishes the table. These are the CY 2026 figures.

Three things about that table matter more than the numbers. The figures change every January. The middle row applies to a hospital and its own medical staff, not to an imaging center and its referrers. And clearing a Stark exception does not answer the Anti-Kickback question, which is a separate analysis under a separate statute.

Can you give a referring practice software or an interface for free?

The honest answer is sometimes, under tightly drawn conditions. The Stark exception for electronic health records items and services at 42 CFR 411.357(w) requires the physician to pay 15 percent of the donor's cost before receiving the items. A parallel Anti-Kickback safe harbor sits at 42 CFR 1001.952(y). A separate exception for cybersecurity technology at 42 CFR 411.357(bb) has no contribution requirement but carries its own conditions. Whether any of it fits your arrangement is a question for counsel, not for a marketing plan.

Why is this worth raising before you sign anything?

Enforcement interest in referral technology is current, not historical. On 30 June 2026, HHS OIG posted Advisory Opinion 26-15, an unfavorable opinion concerning a home health agency's payment of remuneration to a vendor for the use of online referral management software. OIG advisory opinions bind only the party that requested one, and OIG issues them about a requesting party's own existing or proposed arrangements. Show your counsel that opinion before you sign a referral platform contract, and ask whether your own arrangement needs one.

How to build the strategy in four steps

1. 1

   Analyze your referral sources

   Pull at least twelve months of referral data and answer four questions: who refers, how often, for which exams, and what the trend is. Then sort the list into four groups: growing, steady, declining, and lapsed. Declining and lapsed referrers are your cheapest available volume, because the relationship already exists and something specific broke it.

2. 2

   Develop and test an engagement plan

   Take the targeted list and set a goal per segment. Growing referrers need protecting. Declining ones need a conversation about what changed. Lapsed ones need a reason to try again. Educational sessions, lunch and learns, and service-line updates all work here, but the plan should be small enough that you can tell whether it worked.

3. 3

   Remove the friction in your referral process

   This is the step most strategies skip, and the one the comparison table above argues is decisive. Referrals should arrive digitally, orders should carry their clinical data forward into intake and scheduling without re-keying, and results should return to the referring system automatically.

   Every fax that a staff member has to sort by hand is a delay the referring office experiences as your center being hard to work with. Automating that path is not an efficiency project, it is a retention project. See how this works in the CareFlow platform and in the Ora AI layer that classifies and routes inbound orders.

4. 4

   Measure by physician, then adjust

   Report referral volume by physician, month over month, against the plan you set in step one. You are looking for three things: which outreach produced measurable volume, which relationships are quietly declining, and which exam types are moving. Then change one variable at a time.

What does the patient's experience have to do with your referrals?

The referring office hears about every part of it. A patient who cannot get through to book, who arrives to fill in the same history a second time, or who forgets the appointment entirely, calls the referring office rather than yours. That is how a scheduling problem becomes a referral problem.

Four mechanics do most of the work:

• Self-scheduling. Let the patient book from the order confirmation, so the referring office never has to chase an appointment on the patient's behalf.

• Pre-registration. Collect insurance, authorization and clinical history before arrival, using the data already on the order rather than asking for it again.

• Reminders. Send SMS and email reminders on a fixed schedule, with a one-tap way to reschedule instead of simply not turning up.

• No-show recovery. Treat a no-show as a task, not a gap in the diary. Reach the patient the same day and tell the referring office what happened.

Every one of these shortens the time between the order and the report, which is the interval the referring physician is actually judging you on.

How do you spot a referrer whose volume is slipping?

You do not need new software to answer this. You need one identifier, one time window and one comparison, applied to data your RIS already holds. Most centers discover the answer was always available and simply never assembled, because referral reporting tends to be built around totals rather than around individual referrers. The three choices below are what turn that same data into something you can act on.

What is the right identifier for referral data?

Use the ordering provider's National Provider Identifier, not the name typed onto the order. Names arrive misspelled, abbreviated, entered under the practice instead of the individual, and changed after a marriage or a move. The NPI does none of that. CMS also publishes the NPPES NPI Registry as a free public lookup, so you can resolve an NPI to a current name, taxonomy and practice address when your own record has drifted.

Which comparison actually shows a decline?

Compare a rolling three months against the same three months a year earlier, per NPI. That removes most seasonality, and it survives a practice that refers in bursts. A month-over-month view will show you noise and call it a trend. Sort the result into the four groups from step one, and set a threshold you are willing to act on before you look at the numbers.

Four signals are worth a phone call, and none of them require a new system:

• Volume down against the same period last year, beyond the threshold you set in advance.

• Mix shifted, for example the MRI work has gone elsewhere while the plain film stayed.

• Gap since last order longer than that referrer's own normal interval.

• Rising cancellation or no-show rate on that referrer's patients, which usually points at your scheduling rather than at their loyalty.

Which numbers belong on the report?

Five, reported monthly by NPI and reviewed quarterly against the goals you set in step one.

The last row is the one to fix first if it is weak. It is the measure the published evidence keeps landing on, and it is the one your referrers can feel without being asked.

What good looks like after a year

A working strategy shows up as a shorter list of surprises. You know who your top referrers are without asking, and you hear about a problem from the referring office rather than from a volume report. The practices that stopped referring are ones you decided not to pursue rather than ones you lost without noticing.

---