Referral Management for Independent Networks

If your network carries any total-cost-of-care risk, every referral that lands at a higher-priced site works against your shared savings, whether anyone chose it on purpose or not. Referral management is not an administrative courtesy for a network. It is where your network's pricing gets decided, one referral at a time.

That matters because most of the referral conversation is still stuck at the practice level: faxes, missing consult notes, patients who never show. A network sees a bigger problem underneath it.

How much more does a hospital outpatient referral cost than an office visit?

Across 57 MedPAC-identified site-neutral services, hospital outpatient prices exceeded physician office prices in every service category and every year measured, reaching 1.27 to 13.5 times higher by 2022 (Health Care Cost Institute, 2025). That finding comes from commercial employer-sponsored insurance claims, not Medicare, and the multiple varies widely by service, so it is a directional finding about where money goes, not a single conversion rate you can apply to any one referral.

What it tells a network is simpler than the range makes it look: the site of service is a price decision, and referral routing is what makes that decision for the patient. Most practice leaders think of a referral as a scheduling handoff, get the patient to a specialist and get the notes back. For a risk-bearing network, a referral is a routing decision with a price attached, and that price flows straight into whatever total-cost-of-care number the network reports at the end of the year. Seen this way, referral management is not a workflow fix waiting on better software. It is the architecture of what your network's care actually costs.

Why hasn't better EHR technology closed the coordination gap between referring and receiving physicians?

Because the gap was never a technology problem to begin with. A 2019 replication of a national survey of 4,754 primary care physicians found that 22% sometimes or seldom send clinical information at the time of referral, and 33 to 35% sometimes or seldom receive useful information back from the specialist, a pattern the study's authors called "remarkably similar" to what a comparable survey found more than a decade earlier, despite the widespread adoption of EHRs in the years between (Timmins et al., Annals of Family Medicine, 2022).

Eleven years of electronic health records did not move that number. That is the strategic finding here, not a footnote: if the industry's single biggest coordination investment of the last two decades did not change the completion rate of a basic care loop, the problem was never about whether the software exists. It is about whether anyone in the clinically integrated network is accountable for what happens between the referral and the follow-up.

How many of your network's referrals are actually completed, and how would you know?

It depends entirely on how you measure it, and that gap in measurement is itself the finding. In a 30-state prospective study, 79.2% of referred patients had seen the specialist by the referring physician's own report (Forrest et al., Annals of Family Medicine, 2007; this cohort draws on 1997 to 1999 data, so treat the level as dated even where the mechanism still holds). In a separate records review of one primary care network within a single health system, covering 20 specialties, only 34.8% of referral scheduling attempts were documented as completed inside the EHR, and 38.9% of those scheduling attempts had no appointment date logged at all (Patel et al., Journal of General Internal Medicine, 2018; July 2015 to June 2016 data).

Those two numbers are not competing estimates of the same thing, and a network should never treat them that way. One measures physician-reported outcomes, the other measures what got documented in a system of record. Read together, they suggest that a meaningful share of the apparent completion shortfall is a documentation failure, not an absence of care. That distinction matters for a network deciding where to spend its attention: the fix may be less about chasing patients and more about making sure your closed-loop referral documentation reflects what already happened.

You have likely seen a widely quoted dollar figure for what referral leakage costs a hospital or a physician every year. No primary-sourced dollar estimate for referral leakage appears in the peer-reviewed literature; the figures in circulation trace to vendor content and search summaries, not to a study. That absence is worth saying out loud rather than working around: if no credible study has priced the problem, the honest starting point for a network is the site-of-service and completion evidence above, not a number nobody can source.

One finding in this body of research is genuinely actionable. When the appointment is scheduled inside the referring office rather than left to the patient, completion odds rise 1.91 times, and a physician-patient relationship of five years or longer raises them 2.14 times (Forrest et al., Annals of Family Medicine, 2007; same 1997 to 1999 cohort, caveat as above). The lever a network can pull today is not a leakage estimate. It is who books the appointment.

Does referral destination affect a network's total-cost-of-care performance?

Yes, and the evidence points toward independent-led networks doing this better than hospital-led ones. The Congressional Budget Office's 2024 review of Medicare ACOs found that ACOs led by independent physician groups generate "substantially larger" savings than hospital-led ACOs (CBO, April 2024). CBO did not attach a dollar figure to that finding, and neither should you: it is a directional result about who tends to manage cost of care well, not a benchmark to plug into a model.

The gap shows up again in how specialists get integrated into risk arrangements at all. A survey of 101 respondents representing 174 Medicare ACOs found specialists make up 66% of ACO medical providers program-wide, yet only 11% of employed specialists and 7% of contracted specialists were rated highly aligned with the ACO's goals, and 62% of respondents called specialist alignment a high priority they had not solved (Mechanic et al., AJMC, 2024). A network that has not built a deliberate specialty referral network strategy is running its total-cost-of-care number on the two-thirds of its provider base it has the least visibility into.

What you can do this week

Start by finding out which of your two completion numbers you actually have. Pull your EHR's referral-completion documentation rate and compare it, even roughly, to what your specialists report closing the loop on. If the two disagree sharply, you likely have a documentation gap before you have a leakage problem, and that is a cheaper fix.

Next, look at who is booking the specialist appointment today. If it is left to the patient after they leave the exam room, that is the single highest-leverage change available to you, backed by the largest effect size in this research. Build the in-office scheduling step into your referral workflow before you build anything more expensive.

Finally, treat your specialist network the way you treat your primary care network: as a managed relationship with defined expectations, not a directory. A network that can name its high-volume, high-cost referral pathways, and knows the price differential across sites for each one, is managing cost of care on purpose instead of by accident.

Connected Care. Shared accountability. Better outcomes.

Ready to see where your network's referral pathways are creating cost you cannot see yet? Talk with TRIARQ about building a specialty referral network designed for total-cost-of-care accountability.