番茄社区

The overall decline in diagnostic cardiovascular testing over the past decade may be attributed to drops in low-value tests, not high-value ones, researchers said -- seen as a win for Medicare payment reforms beginning 15 years ago.

Cardiovascular diagnostic testing rates increased from 275 per 1,000 patient-years in 2000 to 359 per 1,000 patient-years in 2008 (P<0.001), then fell to 316 per 1,000 patient-years in 2016 (P<0.001), reported a group led by Vinay Kini, MD, MSHP, of the University of Colorado Anschutz Medical Campus, Aurora, in .

Two kinds of low-value testing similarly showed significant rises and falls over the years: cardiovascular tests before a low-risk noncardiac surgery (2.4% in 2000 up to 3.8% in 2008, then down to 2.5% in 2016) and routine stress testing within 2 years of coronary revascularization (47.4% in 2000 up to 49.2% in 2003, then down to 30.8% in 2014).

All the while, the high-value practice of left ventricular systolic function assessment increased steadily for cases of acute MI (85.7% in 2000 to 89.5% in 2016, P<0.001) and heart failure (72.6% to 80.1%, P<0.001).

"During the period from 2008 forward in which overall testing rates fell, continued increases in high-value testing for patients with acute MI and heart failure occurred, as well as declines in low-value testing rates before low-risk surgery (reversing the prior trend) and after revascularization (continuing a trend that began in 2003)," Kini's team said.

"Our findings suggest that during a period of Medicare reimbursement changes intended to reduce spending on overall testing, rates of low-value testing declined considerably while guideline-concordant testing among patients with acute MI and heart failure was not adversely affected."

The Medicare payment changes in question had started in 2004 as a series of reductions in the physician fees for inpatient and outpatient testing and the facility fees for office-based testing. Additionally, facility fees for testing in hospitalized patients were removed from prospective payments for hospitalizations based on diagnosis, according to Kini and colleagues.

Ultimately, based on the present findings, these payment changes did not have the unintended consequence of reducing guideline-recommended cardiovascular testing, they concluded.

But whether they were responsible for the drops in low-value tests is another question.

The decline in routine cardiovascular testing after revascularization had started a year before the policy changes of 2004, with no change in slope after the release of appropriate use criteria from cardiovascular professional societies and the Choosing Wisely campaign from the American Board of Internal Medicine.

"Other changes during the study period such as implementation of public reporting of hospital outcomes, vertical integration of cardiology practices into health systems, value-based physician incentives, or other clinical practice changes not addressed in guideline documents could also have contributed to the observed changes in testing rates," the authors suggested.

Their retrospective cohort study tracked age- and sex-adjusted rates of high- and low-value cardiovascular testing from 1999 to 2016, using a 5% random sample of Medicare fee-for-service beneficiaries ages 65-95.

The two most common tests were echocardiography and nuclear SPECT. Other tests included stress ECG, left heart catheterization with left ventriculography, nuclear PET, coronary CT with angiography, and cardiac MRI.

Average age was 75 years at the beginning (2000-2003) and at the end (2012-2016) of the study period. The proportion of women decreased from 63.2% to 57.2% and the prevalence of most comorbid conditions increased during these years.

Kini and colleagues acknowledged that they did not assess other scenarios of other low- and high-value tests, and their retrospective study might have unmeasured confounding. Furthermore, the results are of questionable generalizability to other patient populations.

Comment