BNP test utilization in the ED—it’s not what
  you think

May 2007
Feature Story

Anne Ford

Some say it’s human nature to assume the worst. If nothing else, that theory may explain the widespread belief that BNP and NT-proBNP are often used improperly. “People say, ‘It’s an expensive test. Our clinicians insist on using it. There’s probably a fair bit of inappropriate use of it,’” says Richard C. Friedberg, MD, PhD. Dr. Friedberg is chair of the Department of Pathology at Baystate Medical Center, medical director of Baystate Reference Labs, Springfield, Mass., and professor of pathology at Tufts School of Medicine, Boston.

That suspicion isn’t entirely unfounded. When BNP and NT-proBNP were first rolled out, “definitely there was a lot of marketing involved,” says Elizabeth A. Wagar, MD, laboratory medical director for UCLA Clinical Laboratories and vice chair of clinical affairs for the Department of Pathology and Laboratory Medicine at UCLA’s David Geffen School of Medicine.

“All of a sudden I had in my face this new test,” she says, “and it had not really been analyzed to the extent that other tests are sometimes analyzed in the laboratory world in terms of its value clinically.” Not only that, but “there’s a fair amount of question as to whether laboratory tests in general are appropriately ordered and utilized. Some say that lab utilization contributes to a lot of medical costs, and others say, well, that’s not really proven.

“I thought it would be good—given this new test with some very discrete indications—to see if utilization was really as bad as some people implied,” she continues. Those people, she says, were relying on “just their sense of things,” not concrete data. Hard numbers were in order.

Sounds like a case for the CAP’s Q-Probes. In the recent study, “BNP Utilization,” 119 institutions reviewed 7,041 BNP test orders and reported the rate of appropriate use in the emergency department. (Test orders from freestanding emergency care clinics were also included.) For purposes of the study, the term “BNP” refers to both BNP and NT-proBNP. “We didn’t want to get into a Coke vs. Pepsi argument” by comparing the two markers, Dr. Friedberg says. “This isn’t a question of which test is the better test.”

Participants compared the ICD-9 or clinical diagnosis code associated with each test order against a list of codes and diagnoses for which BNP is indicated. For test orders that didn’t include a code or diagnosis, participants reviewed the emergency department chart and determined the differential diagnosis. The study also asked participants to supply receipt-to-report turnaround time for BNP tests performed in the clinical laboratory rather than the ED.

“Every effort was made to include all appropriate diagnosis codes and clinical diagnostic descriptions that support BNP test ordering,” the authors wrote. Among those 28 codes/diagnoses were chest pain, shortness of breath, heart failure, and pneumonia.

The study did not examine BNP test orders from inpatients, long-term care facilities, nursing homes, or urgent-care facilities that do not accept life-threatening cases. All study participants were located in the United States. Teaching hospitals represented 26 percent of participants, while 9.9 percent had a pathology residency program. Just over 86 percent had been inspected by the CAP within two years.

And the results? “We found that by and large, most of the testing was certainly appropriate,” Dr. Friedberg says. Specifically, the median institution reported an appropriate use rate of 95 percent. “I was pleased to see that things, at least in terms of the ordering data we collected, were a lot better than we anticipated,” he adds. “Sometimes things are being done correctly, despite what we expect.”

As Dr. Wagar points out, however, those findings apply only to the emergency department, not to inpatients or to outpatient clinics. “Only 40 percent of the total BNP tests performed were on specimens from the emergency room or the emergency department. So we really can’t say, of total BNPs performed at the various labs, whether all were performed for appropriate reasons or diagnostic codes,” she says. “Utilization may not be as good if you look at every BNP ordered in an institution.”

Dr. Friedberg counts the relatively low rate of ED-originated BNP test orders as “one surprise” of the study. “I think most of us would have expected to see most [of the orders] coming from the ED and not from inpatients,” he says. He suspects that “people simply look at BNP as a way of assessing fluid status in all patients, not just congestive heart patients. I expect you’re going to see BNP added as a routine test.”

And that, of course, could have financial implications for hospitals. BNP was originally touted as a cost-savings measure, Dr. Friedberg says, the idea being that even though it was expensive, it would save institutions money in the long run because it would allow them to avoid the expense of admitting patients or of having them take up ER space. “Typically what happens with a lot of these ‘spend $10 here, save $100 there’ kind of promises is, you spend $10 here and you don’t save $100 there because it’s tough to get things out of the system that are already well entrenched,” he says.

“Not that there’s anything wrong with adding BNP as a routine test. But the argument to make it a routine test—and the argument to justify ordering it on every ICU patient every day, for example—has nothing to do with the argument for why it’s cost-effective inside an emergency room setting.”

The study also examined turnaround time for non-point-of-care BNP testing, and found a median TAT of 40.5 minutes. The authors could find no correlation between particular demographic or practice variables and appropriate BNP use. However, they found that lower turnaround time goals, BNP reports that did not include recommendations or comments other than reference ranges, and lower occupied bed sizes were associated with lower median turnaround time.

One additional finding: the most commonly used diagnostic codes or diagnoses associated with BNP, namely, chest pain (786.50–786.59); shortness of breath, tachypnea, wheezing (786.05, 786.06, 786.07); and heart failure (428.0–428.9). “Those are all very appropriate,” Dr. Wagar says, “which brings you to question some people’s criticism of laboratory utilization as less efficient than it should be. I think we proved that laboratory utilization can be pretty good, contrary to many misperceptions.”