DOI: 10.1148/radiol.2441061198
(Radiology 2007;244:319-320.)
© RSNA, 2007
Are Two Tests Always Better than One?
Theodoros D. Karamitsos, MD, and
Joseph B. Selvanayagam, MBBS, FRACP, DPhil
Department of Cardiovascular Medicine, University of Oxford, John Radcliffe Hospital, Oxford, Oxon, OX3 9DU, England
e-mail: joseph.selvanayagam{at}cardiov.ox.ac.uk
Editor:
We read with great interest the article by Dr Cury and colleagues (1) in the July 2006 issue of Radiology. The authors studied the accuracy of a combined magnetic resonance (MR) approach (stress first-pass perfusion imaging followed by delayed enhancement imaging) for depicting clinically significant coronary artery stenosis in patients suspected of having or known to have coronary artery disease. The patient population consisted of 47 patients, of whom 14 had experienced a previous myocardial infarction (30%). The prevalence of significant (>70% stenosis) coronary artery disease in the population studied was 65%. By studying this patient cohort by using the "believe the positive" rule, the authors found a slightly better diagnostic accuracy for the combined stress first-pass perfusion imaging and delayed enhancement imaging than for first-pass perfusion MR imaging alone.
It is disappointing that the positive and negative predictive values of either cardiovascular MR approach are not clearly stated in the article. In assessing the utility of any diagnostic test, we need to know the probability that the test will give the right diagnosis in an individual patient—quoting the sensitivity and specificity does not give us that information (2). Furthermore, we find the percentage of the study population with previous history of myocardial infarction to be low (14 of 47 patients). As with the performance of any diagnostic test, the diagnostic accuracy of the combined perfusion imaging and delayed enhancement approach is largely affected by the patient population studied and the prevalence of the disease in the study population. We believe that in a "real life" situation with a larger percentage of patients with previous myocardial infarction, or previous revascularization, the combined perfusion and delayed imaging approach would result in increased numbers of false-positive results, with subsequently decreased specificity. Hence, not withstanding the results of another recent article (3), we believe that the actual value of adding delayed enhancement to first-pass perfusion imaging is still unproved and should be tested in a larger and more representative cohort of a typical catheterization laboratory.
 |
References
|
|---|
- Cury RC, Cattani CA, Gabure LA, et al. Diagnostic performance of stress perfusion and delayed-enhancement MR imaging in patients with coronary artery disease. Radiology 2006;240:39–45.[Abstract/Free Full Text]
- Altman DG, Bland JM. Diagnostic tests 2: predictive values. BMJ 1994;309:102.
- Klem I, Heitner JF, Shah DJ, et al. Improved detection of coronary artery disease by stress perfusion cardiovascular magnetic resonance with the use of delayed enhancement infarction imaging. J Am Coll Cardiol 2006;47:1630–1638.[Abstract/Free Full Text]
Response
Ricardo C. Cury, MD
Cardiac MRI-PET-CT Program, Department of Radiology, Massachusetts General Hospital, Harvard Medical School, 14 Fruit Street, Boston, MA 02114
e-mail: rcury{at}partners.org
I read with great interest the comments from Drs Karamitsos and Selvanayagam regarding our article published in the July 2006 issue of Radiology (1). Although we did not report the positive and negative predictive values of the different cardiac MR approaches tested for detection of myocardial ischemia, one can calculate these values by using a simple 2 x 2 table given the prevalence of disease, sensitivity, and specificity through probability revision. With this method, the positive and negative predictive values of the combined approach by using stress perfusion and delayed enhancement imaging are 85% and 91%, respectively, whereas the positive and negative predictive values of the stress and rest perfusion approach are 82% and 87%, respectively. When one is reporting diagnostic performance of a test, the most valuable test characteristics are sensitivity and specificity because they are independent values from the prevalence of the disease. Moreover, if you know the sensitivity and specificity of a diagnostic test and the pretest probability of the disease, you can calculate the posttest probability by using the Bayes formula (2).
The percentage of patients with myocardial infarction was not quite low in our study population and may represent a typical cohort of a catheterization laboratory. We detected delayed enhancement not only in all 14 patients from group 2 with prior history of myocardial infarction or coronary artery bypass graft placement but also in seven patients with unsuspected myocardial infarction in group A (total, 21 of 46 patients). This finding demonstrates the value of delayed enhancement technique. Furthermore, the number of false-positive results is unlikely to be increased with the presented approach. The diagnosis of inducible ischemia will be made only when there is stress perfusion defect in a myocardial segment without delayed enhancement. If there is matching perfusion defect and delayed enhancement or if there is only delayed enhancement, the myocardial segment will be considered to have a fixed defect or prior infarct but no ischemia. Finally, I agree that more research is necessary, but the presented MR imaging approach looks very promising, as another recent study demonstrated similar results (3).
|
References
|
|---|
- Cury RC, Cattani CA, Gabure LA, et al. Diagnostic performance of stress perfusion and delayed-enhancement MR imaging in patients with coronary artery disease. Radiology 2006;240:39–45.[Abstract/Free Full Text]
- Hunink MGM, Glasziou PP, Siegel JE, et al. Decision making in health and medicine: interpreting diagnostic information. Cambridge, England: Cambridge University Press, 2001;136–143.
- Klem I, Heitner JF, Shah DJ, et al. Improved detection of coronary artery disease by stress perfusion cardiovascular magnetic resonance with the use of delayed enhancement infarction imaging. J Am Coll Cardiol 2006;47:1630–1638.[Abstract/Free Full Text]
TOP
References
References 
