University of Minnesota doctors are raising alarms about “diagnostic discordance,” a little-discussed but potentially dangerous problem that occurs when a patient is transferred from one hospital to another but ends up with a different list of diagnoses after arriving.

In one of the first studies to examine the phenomenon, the doctors found discrepancies in 85 percent of transfers. In some cases that was because patients developed new conditions, but in others it resulted purely from miscommunication between hospitals.

Worse yet, they found that patients with inconsistent diagnostic records were more likely to die in hospital care.

“There’s frequent communication breakdowns, and that results in real-world problems for patients, including mortality,” said Dr. Michael Usher, an assistant professor in general and preventive medicine at the U who led the research.

Their concern comes amid a reported increase in the number of patient transfers, which can occur when, for example, a heart attack patient is triaged at a community hospital then moved to a cardiac center for surgery, or when a mentally ill patient gets held in one hospital’s emergency room until a psychiatric bed opens up in another.

One in 20 hospital patients now gets transferred, according to 2015 inpatient data from the federal Agency for Healthcare Quality and Research. Mergers have fueled the increase, as small hospitals consolidate services and send complex patients to larger affiliated hospitals.

Simply transferring a patient presents risks, according to a 2017 study by Stanford University researchers. They found that patients who move from one hospital to another experience longer stays, more medical mistakes and greater odds of dying in care.

Exactly how inconsistent records aggravate the problem for transferred patients is unclear. Patients who get transferred tend to be sicker and have more complicated medical records, which increases the chances for discrepancies. Their frail conditions also increase their risk of dying in care. So its possible that diagnostic discord isn’t a cause of in-hospital deaths, but rather a red flag of the risk.

However, Usher said he has seen the consequences of miscommunication in his career as a hospitalist, a physician who makes rounds of patients while they are hospitalized. He recalled one patient who was accepted for a heart procedure, then suffered a cardiac arrest that could have been prevented if the previous hospital had forwarded information about his severe kidney failure.

Some variations in diagnostic records are understandable; a patient might present with chest pain at one hospital, and then be diagnosed at the next with the underlying condition causing that pain. And sometimes patients get better, allowing hospitals to remove diagnoses from their records.

To identify the problematic cases, the U’s researchers reviewed billing records for more than 180,000 transfers among hospitals in five states. Their findings, published in the Journal of Geriatric Internal Medicine, focused on patients with diagnoses such as chronic obstructive pulmonary disease that are incurable and should always appear in medical records. Usher said it was troubling how often this type of diagnostic information wasn’t passed along.

“It’s not that the diagnosis was being lost,” he said. “It was the information being lost in the transfer from one hospital to another.”

Testing solutions

Identifying the scope of the problem is step one. The University of Minnesota Medical Center is testing a pilot approach to transfer information immediately along with patients. Researchers also are examining whether patient information gets lost in other transfers, such as from hospitals to nursing homes, and whether patients’ insurance status affects whether their information gets passed along.

While all hospitals in Minnesota now have electronic medical records, the ability for them to be interoperable — meaning they can exchange information with one another — is inconsistent.

Only 63 percent of hospitals routinely have access to electronic records from outside hospitals or clinics, according to survey data from the Minnesota Department of Health.

That’s a problem considering the number of patients who use doctors and hospitals from different organizations, said Jennifer Fritz, the health department’s health information technology director.

“We’ve made really good progress in the state in the adoption of that technology,” Fritz said. “But the interoperability of that technology? We’re still working on that.”

Hospitals have little problem sharing information if they are part of the same organizations or use the same software. Most large hospitals in Minnesota use record-keeping systems made by Wisconsin-based Epic. But those that do not find it harder to share information with those that do, the state survey data shows.

Usher said he hopes the U research draws attention to the issue by identifying the potential harms to patients.

The consequences also show up in malpractice claims, said Trish Lugtu, a manager of advanced analytics for Constellation Mutual, which provides malpractice insurance to doctors. When claims against hospitals or clinics allege diagnostic problems, 42 percent of them involve follow-up care or coordination between different medical providers.

“Any time you have disparate organizations, when you don’t have the communication process down,” she said, “that might happen.”