Staged testing as a solution to the challenges of testing lower risk patients
- PDF / 587,822 Bytes
- 4 Pages / 593.972 x 792 pts Page_size
- 19 Downloads / 167 Views
b
Division of Cardiovascular Medicine, Department of Internal Medicine, University of Michigan, Ann Arbor, MI Division of Cardiovascular Medicine, Center for Outcomes & Research Evaluation (CORE), Yale University School of Medicine & Yale New Haven Health, New Haven, CT
Received Aug 27, 2018; accepted Aug 27, 2018 doi:10.1007/s12350-018-1437-y
See related article, https://doi.org/10.10 07/s12350-018-1376-7.
INTRODUCTION Several groups have published on the decreasing prevalence of reversible perfusion abnormalities in stress SPECT MPI populations, such that ischemia may be present in under 10% of tests among patients without established CAD.1,2 This trend is seen despite strong evidence that many patients with stable CAD can be managed safely and effectively with only medical therapy,3–5 prompting the question why are we testing lower and lower risk populations over time? It is likely that many factors contribute including changing demographics, protocol driven care, malpractice concerns, and decreased continuity between inpatient and outpatient encounters. Regardless, the declining prevalence of ischemia is a major challenge for nuclear cardiology labs. Chief among these challenges is a marked decrease in positive predictive value with declining prevalence (Figure 1). REVIEW OF TERMS Although likely familiar to most of the readers of the Journal of Nuclear Cardiology, a brief review of terms is in order. The pre-test risk of the entire
Reprint requests: Venkatesh L. Murthy, MD, PhD, FASNC, Division of Cardiovascular Medicine, Department of Internal Medicine, University of Michigan, 1500 E. Medical Center Dr, SPC 5873, Ann Arbor, MI 48109-5873; [email protected] J Nucl Cardiol 1071-3581/$34.00 Copyright Ó 2018 American Society of Nuclear Cardiology.
population is equivalent to the prevalence in the population (pre-test probability estimates for individuals will vary around this). The sensitivity of a test is the proportion of those patients with disease who will be correctly identified as disease. Physicians worry about this parameter greatly out of concern for missing disease. Perhaps more critical in testing low risk populations is specificity which is defined as the proportion of non-diseased patients whom the test will correctly identified as non-diseased. It therefore follows that the proportion of non-diseased patients who are incorrectly identified as diseased is given by (1-specificity) (false positives). In practice, the sensitivity and specificity are neither particularly intuitive nor useful for patients and clinicians. More practically useful are the positive predictive value (PPV) and negative predictive value (NPV). The PPV is defined as the proportion of all patients identified as abnormal by the test who are truly abnormal. When prevalence is moderate, PPV is principally affected by specificity with a very minor contribution from sensitivity. However, when prevalence drops, PPV declines precipitously, regardless of specificity (Figure 1A). Consequently, even very specific tests cannot ove
Data Loading...
