The $500 Million Idea

We all know there are plenty of patients we transport to emergency departments just because we have to. That’s the only way Medicare will reimburse us, even for those who could be helped more cheaply and appropriately elsewhere. A new paper in the journal Health Affairs takes the latest look at how many of our low-acuity patients might be safely treated otherwise, and how much Medicare might save if allowed to reimburse for other kinds of dispositions.

The big takeaway: If Congress let CMS (the Centers for Medicare & Medicaid Services) reimburse EMS for taking patients to destinations like physicians’ offices or health centers or even just treating them on scene, it could save $283–$560 million or more a year. What’s more, if private third-party payers (who often take cues from CMS) did the same, those savings might double.

“I think there’s really been a legitimate question of, if EMS did all these things—and they would really represent a fairly significant change—would it really save money?” says coauthor Gregg Margolis, PhD, NREMT-P, director of the Department of Health and Human Services’ Division of Healthcare Systems and Health Policy. “I think the evidence from this paper and others is that it likely would. But the other piece to emphasize is that this isn’t just about saving money. It’s really about a more medically appropriate disposition to the broad range of 9-1-1 calls. A sizable proportion of people call 9-1-1 for things that would be better managed in a different setting. So this is really about better care at a lower cost.”

Reaching a Figure

To figure how much alternative management of Medicare patients might save, Margolis and his colleagues first had to determine how many patients could be affected, then how much spending they accounted for under current policies.

They started with a random 5% sample of Medicare claims data for 2005–09. Ambulance claims yielded information on total costs of transports, service levels provided, origins/destinations and miles traveled. That represented nearly 4 million records. Then they linked each claim by beneficiary number and service date to associated Medicare claims for emergency department care and everything else covered. That provided a pretty accurate idea of total costs for ambulance transport and subsequent treatment.

The authors tossed out interfacility transfers and everything else except payments to ambulance services for 9-1-1 responses resulting in ground transports to emergency departments. That winnowed the total to roughly 1.8 million cases and let them extrapolate that over the five years studied, Medicare reimbursed an annual average of more than 7.1 million transports to EDs.

From those they struck transports that resulted in hospital admissions, since those patients generally wouldn’t be candidates for care in other settings. That left roughly 811,000 transports, to which they applied an algorithm, developed by health-policy expert John Billings, that classified the visits in one of four ways based on discharge diagnosis:

• Nonemergent;

• Emergent and primary care-treatable;

• Emergent, ED care needed, and preventable or avoidable; and

• Emergent, ED care needed, and not preventable or avoidable.

ED visits related to injuries, mental health issues and drug/alcohol use aren’t covered by this algorithm and were classified separately.

From the algorithm the authors determined an overall proportion of Medicare-reimbursed EMS transports that might be nonemergent or emergent but treatable by primary care. To calculate the costs of taking those low-acuity cases to EDs, they added the ambulance transport, ED facility and physician payments, weighted by the percentage in those two classifications.

What all this revealed is that about 34.5% of 9-1-1 EMS transports of Medicare beneficiaries not hospitalized were relatively low-acuity cases that could possibly be managed outside an ED. That’s around 15.6% of all Medicare-covered 9-1-1 transports to EDs, a figure in line with some earlier research. “That’s kind of the low end of the range that had been previously reported,” says Margolis, “but clearly within experience and other published literature on the subject.”

Around $1 billion a year is spent on these patients, with about $380 million of that going to ambulance services. Managing them in less-expensive settings could save an estimated $560 million a year.

Limitations

There are a few limitations to this study. First, while the algorithm it used to classify patients—constructed from ED visits in New York City—was the best-suited tool available, it was designed to evaluate community access to primary care and is not a triage tool.

“It’s not a tool that’s intended, on an individual basis, to identify patients who might be candidates for alternative destinations,” explains Margolis. “It’s an algorithm that’s applied post hoc to a population of individuals who showed up in emergency departments to quantify the potential number who could have gone to alternative destinations, if they were available.”

The algorithm assumes they are, but of course alternative destinations aren’t always available, accessible or open. Some keep limited hours; some won’t take uninsured or Medicare patients; some communities lack them altogether. Conversely, EDs are 24/7 and take all comers. So that may contribute to overestimation of the number of Medicare patients treatable elsewhere.

The paper notes some other limitations. For one, the NYC-based algorithm might not be generalizable to other areas. For another, ED costs for patients considered emergent are on average higher than those for patients thought nonemergent or emergent but primary care-treatable, so using mean costs might overestimate potential savings. Nonetheless, it concludes, “Even if the actual savings were half as large as our baseline estimate…the potential savings are still large enough to justify prospective research to assess the feasibility and safety of a change in policy.”

What Can Paramedics Tell?

All of this points, however, to another big question yet unanswered: Can paramedics safely identify in the field those patients who might safely be treated alternatively? While some pilot programs have shown promise, the literature on the whole isn’t especially convincing.

“I think there is some honest concern for patient safety,” says Margolis, “in that, while there may be 15% of patients in this mix, EMS has not yet demonstrated the ability to  safely identify them with the tools available in the field.”

To broaden that range of possible dispositions, then, systems would likely also have to enhance training, oversight, protocols, clinical decision-making support and quality assurance, among other areas, as well as relationships with those other potential destinations.

Obviously, no small task. Achieve it, and you might just change what EMS in America fundamentally is. 

“Any time a patient calls 9-1-1, the goal should be to get them care in the optimal setting,” says Margolis. “This really reframes the role of EMS in the nation’s healthcare system. EMS is not just an extension of the emergency department, but a networked part of the healthcare system. Every call to 9-1-1 represents an opportunity to connect a patient to the healthcare resources they need.  It might be an emergency department or trauma center, but it might be something else. Making this change would be more patient-centered and probably save a lot of money.”

Reference

Alpert A, Morganti KG, Margolis GS, Wasserman J, Kellermann AL. Giving EMS flexibility in transporting low-acuity patients could generate substantial Medicare savings. Health Affairs, 2013 Dec; 32(12): 2,142–8.