Noninvasive stress testing varies considerably following PCIs at VA hospitals

An analysis of veterans undergoing PCIs at Veterans Affairs (VA) hospitals across the U.S. found the rates of noninvasive stress testing following the procedures varied considerably. The frequency of stress testing was not significantly associated with all-cause mortality or MI.

During the study, which lasted from October 2007 through June 2010, 21.8 percent of patients had a stress test within a year of their PCIs and 37.9 percent had a stress test within two years. The risk-standardized rates at the 55 hospitals ranged from 28.5 percent to 55.2 percent, according to lead researcher Steven M. Bradley, MD, MPH, of the VA Eastern Colorado Health Care System in Denver.

Bradley and his colleagues published their findings online in Circulation: Cardiovascular Quality and Outcomes on July 21.

Previous research found that approximately 20 percent of patients have recurring symptoms within a year of PCIs but more than 35 percent undergo stress tests. The discrepancy brought upon concerns that noninvasive stress tests after PCI were overused. Bradley et al suggested that a fee-for-service healthcare model incentivizes doctors to order too many stress tests because they are paid for each test.

They decided to study stress tests after PCIs in the VA because it is the largest integrated healthcare delivery system in the U.S. They evaluated data in 10,293 patients who were not Medicare eligible during the follow-up period.

The types of testing included stress echocardiography, ECG or pharmacological stress tests, myocardial nuclear imaging or stress MRI.

Of the patients who underwent stress testing, 16.2 percent were performed without imaging, 79.8 percent were performed with nuclear imaging, 3.8 percent were performed with echocardiography and less than 0.1 percent were performed with MRI.

Between 60 days and two years after the PCI procedures, 5.1 percent of patients died and 4.5 percent had an MI.

The researchers said there were few differences in patient and hospital characteristics that explained the variation in stress testing among hospitals.

“These findings suggest factors other than patient characteristics, reimbursement structure, and care delivery integration influence practice variation in the use of stress testing,” they wrote. “This conclusion is further supported by recent studies that have demonstrated variation in the use of coronary procedures within the capitated healthcare delivery setting of Medicare Advantage. Together, these findings suggest new models of care delivery that emphasize care integration and reimbursement design may be insufficient to eliminate unwarranted variation in care delivery.”