Emergency Department Imaging, Testing Rises, Increasing Length of Stay

Nancy A. Melville

October 06, 2010

October 7, 2010 (Las Vegas, Nevada) — A significant increase in imaging and diagnostic testing in the emergency department (ED) is leading to longer stays and is likely having a direct effect on ED overcrowding, according to research presented here at the American College of Emergency Physicians 2010 Scientific Assembly.

A retrospective study looking at ED trends with data from the National Hospital Ambulatory Medical Care Survey found that the rates of imaging tests such as magnetic resonance imaging (MRI) and computed tomography (CT) scans in the ED were as much as 3 times higher in 2007 than in 1998, and obtaining the tests was associated with longer ED stays. The rate of procedures to treat conditions remained steady.

"In the past decade, the odds of getting a treatment in the emergency department have been flat, while the odds of having a test there have increased," said Keith E. Kocher, MD, lead author and clinical instructor at the University of Michigan's Department of Emergency Medicine Injury Research Center, in Ann Arbor.

"The intensity of that testing has also increased, and this activity is associated with a longer length of stay," he said. "So we're probably contributing to crowing conditions with our activity."

The study found that among the more than 1.1 billion ED visits between 1998 and 2007, about 432 million (39.1%) involved 1 of 3 diagnostic tests (complete blood count [CBC], urinalysis, or electrocardiogram); 462 million (41.9%) involved an imaging study such as x-ray, ultrasonography, CT scan, or MRI; 280 million (25.3%) involved dispensing 3 or more medications; and 410 million (37.1%) involved 1 of 7 treatment procedures (cardiopulmonary resuscitation, endotracheal intubation, intravenous [IV] fluids, bladder catheterization, nasogastric tube placement, wound care, or orthopedic care).

The adjusted odds ratio (OR) of obtaining at least 1 of the 3 diagnostic tests in 2007, relative to 1998, was 1.15 (95% confidence interval [CI], 1.04 - 1.28); the OR of obtaining an imaging study was 1.15 (95% CI, 1.07 - 1.23); and the OR of dispensing 3 or more medications was 1.37 (95% CI, 1.21 - 1.56).

The OR of performing at least 1 of the selected procedures, however, was 0.99 (95% CI, 0.88 - 1.11).

Imaging and diagnostic tests were more likely to be performed in 2007 than in 1998, including MRI (OR, 3.38; 95% CI, 2.11 - 5.41), CT scan (OR, 3.17; 95% CI, 2.75 - 3.66), ultrasonography (OR, 1.56; 95% CI, 1.27 - 1.92), IV fluids (OR, 1.39; 95% CI, 1.17 - 1.66), and CBC (OR, 1.31; 95% CI, 1.14 - 1.50).

The adjusted OR of obtaining 2 or more of the diagnostic tests during a patient's ED stay was greater in 2007 than in 1998 (adjusted OR, 1.32; 95% CI, 1.18- 1.47), as was the OR of having 2 or more different imaging modalities (adjusted OR, 2.11; 95% CI, 1.85 - 2.42).

The increase in testing was linked to longer ED stays. Linear regression modeling for 2006/07 showed that obtaining 1 of the 3 diagnostic tests was associated with a 127.1% (95% CI, 113.5% - 143.4%) longer stay in the ED.

Obtaining an imaging study extended the ED stay by 80.8% (95% CI, 70.0% - 90.8%), dispensing 3 or more medications resulted in a 31.6% longer stay,(95% CI, 24.5% - 40.8%), whereas performing any of the 7 procedures extended the stay by 17.5% (95% CI, 10.5% - 24.3%).

A benefit of the additional testing might be a reduction in hospitalizations and their associated costs, but that doesn't necessarily mean there isn't unnecessary testing, Dr. Kocher said.

"We may be diagnosing conditions better and faster than in the past, and this may be saving hospitalizations," he said.

"However, you also have to wonder whether there is more unnecessary testing now than before, and perhaps we are inadvertently treating people to longer stays due to all of this activity."

One important reason for the increase in testing is the pressure hospitals face on admitting patients, said Brent R. Asplin, MD, the session moderator and chair of the Emergency Medicine Department at the Mayo Clinic, in Rochester, Minnesota.

"I think one of the drivers of these trends is the intense scrutiny hospitals and physicians face over admissions," he said.

"A great deal of the patient's diagnostic evaluation is done before they leave the emergency department. Although this may help us discharge some patients who otherwise may have been admitted, I think it is safe to say that these imaging changes have had a major impact on emergency department waiting times and overall operations."

"We need to continue developing better decision rules to guide imaging decisions and reduce variability."

Dr. Kocher and Dr. Asplin have disclosed no relevant financial relationships.

American College of Emergency Physicians (ACEP) 2010 Scientific Assembly: Abstract 16. Presented September 29, 2010.