Is Cardiac CT Useless?
Dispatches from the April RFS Journal Club
This month we took a point-counter point approach to our topic. At the center of this discussion was the recent JAMA article “Comparative Effectiveness of Diagnostic Testing Strategies in Emergency Department Patients With Chest Pain: An Analysis of Downstream Testing, Interventions and Outcomes” by Andrew Foy, et al.
In support of the article, Amy K. Patel, MD, discussed why non-invasive cardiac testing in low-risk ED patients with chest pain may not be useful, leading to overdiagnosis and increased health care costs, as initial noninvasive testing often leads to additional testing such as cardiac catheterization and revascularization procedures. According to the study, these interventions did not lead to a reduction in myocardial infarction.
On the opposing side, Margaret M. Fleming, MD, began by evaluating the study design. In an attempt to select for low-risk patients, the study excluded patients who were diagnosed with a myocardial infarction during the hospital stay that followed the ED evaluation and those whom the non-invasive studies successfully triaged. However, by excluding these patients, the study essentially excluded the true positives: patients who were considered low risk, triaged with non-invasive testing, and found to be positive. Based on the conclusion of this study, these patients would be the ones sent home by the ER physician because they would be classified as low risk and then end up having a heart attack at home. This is one of the problems with using an administrative database versus a randomized control trial. Selecting for the correct patient population excluding/including becomes problematic.
In addition, the study was performed on patients with private insurance, which may not approximate the patient population in all institutions. Furthermore, there were significant confounding differences in the testing cohort versus the non-testing cohort, such as age and comorbidities including diabetes and hypertension.
One conspicuous limitation of the study was that it did not account for mortality. Researchers only looked at the percentage of patients admitted with myocardial infarction in 7 days and 30 days from the initial ER visit. Ultimately, the most compelling studies show that a particular intervention prevented death or prolonged life. Unfortunately, like much of what we do in health care, cardiac CT does not have the ability to prolong life. So we must find other ways to quantify the value these studies provide, whether it be decreased turnaround time for ER physicians or decreased length of stay for a patient in the hospital.
The debate turned to the value imaging can provide in the health care system as Ashley Prosper, MD, cited a JACC article demonstrating that use of cardiac CT in the ED resulted in more cost-efficient, safe diagnosis with quicker throughput of patients when compared to rest-stress myocardial perfusion.
Finally, discussion turned to why ED physicians order cardiac imaging. Our moderator, Saurabh Jha, MBBS, asserted that the heterogeneity of low-risk acute chest pain patients makes it difficult to define the risk of discharging without further testing. Moreover, the importance of patient satisfaction and reassurance of a negative test was also discussed. In turn, reducing uncertainty and better defining levels of risk allows radiologists to save the health system on tangible costs such as length of stay.
And Read This Too
By Melissa L. Chen, MD, radiology resident at University of Texas Health Science Center at San Antonio