Copyright 2016 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
To the Editor Drs Polonsky and Blankstein presented an informative discussion of a patient with chest pain, but their description of the diagnostic characteristics of the ETT merits closer attention—particularly because the case was published in the JAMA Diagnostic Test Interpretation section.1 Although they reported that the sensitivity of ETT for obstructive coronary artery disease (CAD) is “approximately 68% in men and 61% in women” and the specificity “is approximately 77% in men and 70% in women,” these estimates may significantly misrepresent the test’s true performance. Because of referral bias (also known as verification bias) in studies that have evaluated ETT, high-risk patients are generally more likely to be referred to cardiac catheterization than low-risk patients.2,3 Although clinically appropriate, this referral pattern results in an observed sensitivity that overestimates the test’s true sensitivity and an observed specificity that underestimates the test’s true specificity. Studies and meta-analyses of ETT performance rarely account for this bias,3 and those that do account for it report, for example, stress test sensitivity rates as low as 32% in women and 42% in men and test specificity rates that exceed 80% in both sexes.4
Ladapo JA. Treatment of Patients With Stable Ischemic Heart Disease. JAMA. 2016;315(17):1905. doi:10.1001/jama.2016.0683
Customize your JAMA Network experience by selecting one or more topics from the list below.