The stress of stress tests
Douglas Dawley, M.D.
Cardiologist, Providence Heart and Vascular Institute
Published October 2012
Advanced stress cardiac imaging now allows us to detect heart disease with greater accuracy than ever before. But these sophisticated and expensive tests also bring a challenge: When are they necessary and when will simple exercise testing suffice?
Standard treadmill testing often is our first test for diagnosing coronary artery disease. It’s safe, accurate and inexpensive. Under current American College of Cardiology/American Heart Association guidelines, a Class I indication for treadmill testing would be an adult with baseline ST-segment depression at 1 mm or less, and at a pretest probability for heart disease based on sex, age and symptoms.
Exercise or pharmacologic stress tests combined with echocardiography or nuclear imaging provide higher sensitivity and specificity, but they’re considerably more expensive. In addition, the radiation exposure from radionuclide imaging carries more risk to the patient.
For these reasons, appropriate use criteria have been established for imaging tests. The key questions to be addressed before choosing an appropriate stress test are:
- Is the patient at risk for coronary artery disease? This can be assessed by a variety of risk calculations, such as ATP III or Reynolds Risk Score. Another way to assess pretest risk, based only on age, sex and nature of symptoms, is outlined in the table below.
- What degree of accuracy is needed? Treadmill tests are between 60 and 80 percent accurate, while stress echo and nuclear tests are 80 to 90 percent accurate.
- What are the costs? Treadmill tests cost about $300; echo stress tests run about $1,500 and nuclear tests can go as high as $3,500.
- Are there special considerations? That is, can the EKG be interpreted, or can the patient exercise?
Going one step further, the ACC/AHA in 2009 identified 67 clinical scenarios where stress imaging is appropriate. It created algorithms for patient categories, such as symptomatic patients, asymptomatic patients, prior test results, preoperative risk assessment, etc.
Based this review, they came up with “Five Things Physicians and Patients Should Question.” I’ll highlight some of their key points and add some of my own:
- Don’t perform stress imaging in patients without cardiac symptoms unless high-risk markers are present. Asymptomatic, low-risk patients account for 45 percent of inappropriate stress testing. Testing should be performed only if the patient has peripheral artery disease, is older than 40 and diabetic, or is at more than 2 percent yearly risk for a heart disease event.
- Don’t perform stress imaging for patients at low risk. Patients reporting chest pains who are at low risk, have an interpretable EKG and are able to exercise do not need stress imaging studies.
- Don’t perform radionuclide imaging as part of routine follow-up in asymptomatic patients. Performing this test routinely every one to two years after heart procedures rarely results in meaningful management changes, and may lead to unnecessary invasive procedures and excess radiation exposure. An exception to this rule would be for patients who had coronary artery bypass graft surgery more than five years earlier.
- Don’t perform stress imaging as a preoperative assessment in patients scheduled for low- to intermediate-risk noncardiac surgery. Such testing will not change the patient’s management or outcomes, but will result in increased costs. A standard treadmill test is appropriate for patients without symptoms or clinical risk factors, or who have moderate to good functional capacity.
- Use methods to reduce radiation exposure in stress imaging, including not performing such tests when limited benefits are likely. Stress echo is similar in sensitivity and specificity to nuclear testing, but is cheaper and without radiation risk. It represents a preferred stress modality as long as the echo images are interpretable (about 80 percent of the time).
One parting thought: Stress imaging may be overused in patients who for whatever reason are not good candidates for angiography. If patients have typical anginal symptoms, treat them empirically for angina. If they have atypical symptoms, look for other causes.