- Kenny Lin, MD, MPH
Last week, the family medicine residency inpatient service that I supervise admitted several patients from the emergency department with acute chest pain that had resolved. Most of them had no history of cardiovascular disease, but were deemed to have enough risk factors to undergo pre-discharge cardiac stress testing after they had "ruled out" for acute coronary syndrome with normal cardiac enzymes. Rationales for the American Heart Association's recommendation for routine stress testing in patients with resolved chest pain include reducing malpractice liability, improving cardiac risk stratification, and initiating appropriate interventions earlier in high-risk patients. Although this practice is widely accepted, there is no evidence that it improves patient-oriented outcomes compared to outpatient management, and some researchers have argued that randomized trials are needed to prove that the benefits actually exceed the harms.
A recent study published in JAMA Internal Medicine adds fuel to this debate by presenting prospectively collected outcomes of adult patients evaluated in the emergency department chest pain unit of Mount Sinai Medical Center from 2004 to 2010. A total of 4181 patients underwent stress testing (512 with exercise ECG tests and the rest with nuclear perfusion imaging), and 470 tests suggested potential myocardial ischemia. 123 patients underwent cardiac catheterizations; 60 of these patients were found to have normal coronary arteries. Of the 63 patients whose catheterizations showed obstructive coronary artery disease, only 28 had lesions that warranted stenting or coronary artery bypass grafting according to expert consensus guidelines.
There are at least two ways to view this study's results. A positive interpretation is that cardiac stress testing led to in the presumptive diagnosis of coronary artery disease in more than 10 percent of patients, who could then have received medical interventions shown to improve outcomes. On the other hand, the high false positive rates on coronary angiography suggest that up to half of these diagnoses were incorrect (and, consequently, that more than 150 patients would have received therapy inappropriately). Nearly 90 percent of patients were exposed to significant radiation doses through nuclear imaging, but less than 1 percent had coronary artery lesions that warranted revascularization. So are the benefits of routine pre-discharge stress testing in patients with resolved chest pain worth the harms? If not, is reducing medical liability risk enough reason to continue a low-value practice?