Biomarker testing may allay concerns yet provide little clinical benefit. Eliminating unneeded tests can save big bucks, one hospital has shown.
Inappropriate and excessive testing touches many segments in healthcare, including cardiology. Consider cardiac biomarkers, such as troponin, creatine phosphokinase (CK) alone or CK-MB, which often are ordered in the emergency department (ED) to determine if patients have acute coronary syndrome (ACS). Recent research has found that these blood tests are commonly requested even in the absence of ACS symptoms. As much as one third of 28.6 million patient visits involved inappropriate cardiac biomarker testing.
However, at least one institution found a way to scale back these tests without compromising patient care. An intervention at Johns Hopkins Bayview Medical Center in Baltimore successfully reined in the practice, trimming costs by $1.25 million.
Clinical experience drove Anil Makam, MD, MAS, to look into the prevalence of cardiac biomarker testing. “In my own practice, I saw a lot of patients getting admitted having these tests ordered. I wanted to see if it’s unique to my practice setting,” he says.
Makam and colleagues at University of Texas Southwestern Medical Center in Dallas analyzed reports submitted to the National Hospital Ambulatory Medical Care Survey for 2009 and 2010, which provides a representative sample of patients presenting to U.S. emergency rooms (JAMA Intern Med 2015;175:67-75). They found that an estimated 8.5 million patients may have been tested for cardiac biomarkers even though they lacked ACS symptoms.
Also, the study revealed that patients already receiving a high volume of testing were much more likely to get biomarker testing. “This is just not related to biomarkers. It’s a symptom of the bigger problem of over testing,” he says.
Reasons for excessive testing have to do with the environment itself. ED physicians tend to lean on diagnostic certainty to help them rule in or rule out a condition. Also, demands in an ED mean physicians don’t have the time to look at a range of symptoms in a patient that could signal a pending heart attack. “It’s easier for them to order tests. They have a lot of competing priorities,” Makam says.
These physicians are attuned to malpractice risk, too. “Some of this is defensive medicine. They are willing to do the biomarker testing even if [patients] have no symptoms,” he observes.
Johns Hopkins intervention
It was not administrators or payers but physicians who led the effort to reduce unnecessary testing at Johns Hopkins Bayview Medical Center, according to Marc Larochelle, MD. The push to reduce cardiac biomarker testing emerged during a physician-led discussion on how to address the issue of ballooning healthcare costs.
“If there is waste, I think we could be part of the puzzle,” says Larochelle on physicians’ role in curbing healthcare costs. In this case, it led to the formation of a group of physicians and residents from various departments who identified some cardiac biomarker testing as a potential target.
ED physicians often order tests for troponin, CK alone or CK-MB—at average costs per test of $51.90, $12.33 and $30.80, respectively—for patients they suspect may have ACS. But after talking with cardiologists, Larochelle and his colleagues found that troponin alone, tested at certain intervals, is sufficient to determine the presence of ACS. The group developed guidelines in concordance with the American Heart Association that called for only troponin testing with a limit of three tests evenly spaced over 18 to 24 hours.
“There was a huge opportunity. We were using three tests when one was sufficient, and we were using them more than we needed. That was what motivated us,” says Larochelle.
The intervention entailed physician education on the new guidelines paired with changes to the computerized physician order entry system (J Gen Intern Med online June 28, 2014). Thanks to the evidence-based reasoning for the intervention, and a vetting of the new practice with all departments, the intervention was implemented with little pushback, according to Larochelle.
The study, which followed a before-and-after design, utilized data extracted from their administrative database, with a primary outcome of the percentage of patients each month meeting cardiac biomarker guideline recommendations. The pre-intervention period spanned January 2009 to July 2011 (60,494 patients); the rollout from