Home Discharges Does that patient really need to be admitted?

Does that patient really need to be admitted?

A new study tracks patients who are discharged home from the ED

November 2018
questions on admitting low-risk patients

Published in the November 2018 issue of Todays Hospitalist

THE MEDICAL SYSTEM puts a lot of time and energy into identifying high-risk patients who must be closely followed. Think of STEMI patients and the effort that goes into reducing door-to-balloon times to give those patients the best possible outcomes.

But hospitalist Christopher Caulfield, MD, medical director of the observation unit and associate medical director of clinical care management at the University of North Carolina Medical Center in Chapel Hill, is just as intrigued by low-risk patients.

“Those patients don’t get the discussion, documentation or research in the literature as those at high risk,” Dr. Caulfield says. “And low-risk patients are often admitted to the hospital when they may not need to be.” Referring low-risk patients for admission, he adds, may be the easiest option in busy emergency departments that don’t have the time to discuss patient risk or assess how likely reliable outpatient follow-up may be.

“Low-risk patients are often admitted to the hospital when they may not need to be.”

~ Christopher Caulfield, MD
University of North Carolina Medical Center


But low-risk patients—as well as the high-risk ones who need to be hospitalized—face the real possibility of adverse events from being admitted. That was the thought behind a study Dr. Caulfield served as lead author for that was published in the March issue of the American Journal of Managed Care.

In that retrospective, he and his colleagues looked at 2011-14 data on the outcomes of patients who’d been referred to them by their ED for admission—but who they decided instead to discharge home from the ED. Among those patients, 80% didn’t return to the ED within 30 days, although that rate varied depending on diagnosis.

Further, only 12% were hospitalized within the next 30 days, which Dr. Caulfield and his coauthors estimate saved his hospital about $1 million in health care spending over the three years studied. He says that he found that low percentage of returns to be encouraging.

“That shows there are groups of low-risk patients who can be sent home, with the understanding that they’re not likely to return with an adverse event,” Dr. Caulfield says. “I think it’s important to identify those patients and prevent hospitalizations that might not be needed.” Dr. Caulfield spoke to Today’s Hospitalist.

Of the patients the ED referred for admission, how many did hospitalists discharge from the ED instead?
It’s not a large amount, but over the course of the three years of the study, I’d say it was between 5% and 10% of the patients we assessed in the ED.

Part of that decision, obviously, is that the patient appeared to be at low risk. And when deciding to discharge patients home, I also focus on whether I can assume they’re reliable to get follow-up. If it’s during normal business hours, I try to call their primary care office, just to let them know the patient was in the hospital. And with our current EHR, we can often send e-mails or phone messages through the system to at least alert primary care doctors that we saw their patient. I’ll also try to schedule appropriate follow-up for those patients, depending on their condition.

When we discharge patients home from the ED, we bill either an outpatient consultation code for those with non-Medicare insurance (99241-99245) or an ED visit code for Medicare patients (99281-99285).

While 80% of discharged patients overall didn’t return to the ED within a month, three cohorts came back much more frequently: those with sickle cell disease (82%), alcohol-related diagnoses (44%) and abdominal pain (36%). Have you since set up resources in the ED to try to prevent these patients from needing to return?
We have. We have a sickle cell program at UNC as well as a great relationship with our hematology division, so we’ve actually developed individualized care plans for many sickle cell patients. Those include clear expectations of when they’ll be admitted for IV pain control, for example.

The same is true for patients who need an alcohol detox admission when they don’t have good primary care or appropriate follow-up. We again have clear pathways, protocols and expectations. We may care for those patients in the hospital by monitoring their CIWA scores, putting them on standing Librium or providing naltrexone. We obviously also try to set them up with outpatient resources.

The study also found that insurance status made a difference in who returned within 30 days, with Medicaid patients coming back more frequently than those with Medicare or private insurance. Does that reflect problems with outpatient access?
I do think that’s the issue, barriers with outpatient providers and clinics; the wait time to see a provider for these patients may be twice as long. That’s why, for a while now, the ED has been an easy access point for many of these patients.

Not surprisingly, among patients discharged home, the most common diagnosis (38%) was chest pain. Have you now put algorithms in place in the ED to help determine which of these patients warrant observation or admission?
Within the past couple of years, our hospital has started using the HEART score to risk-stratify chest-pain patients who either may need hospitalization or are appropriate for discharge home. We typically discharge those with low-risk scores with the understanding that they’ll receive follow-up and possible stress testing within 72 hours.

One of my colleagues, John Stephens, MD, and I published a “Things We Do For No Reason” review in the Journal of Hospital Medicine on patients with low-risk chest pain who are hospitalized for evaluation. Currently, more than 30% of patients who present to the emergency department with chest pain are hospitalized for observation, symptom evaluation and risk stratification. But a percentage of these patients are likely appropriate for discharge home from the ED with close outpatient follow-up if risk-stratified into the low-risk group at the point of entry.

The use of such risk scores is one more indication that clinicians now have a greater understanding of options available to send low-risk patients home with close follow-up.

Phyllis Maguire is Executive Editor of Today’s Hospitalist.

Notify of
Inline Feedbacks
View all comments