Author + information
- James A. Goldstein, MD, FACC and
- Gilbert L. Raff, MD, FACC⁎ ()
- ↵⁎Ministrelli Center for Advanced Cardiovascular Imaging, William Beaumont Hospital, Royal Oak, Michigan 48073
To calculate the sample size of our single-center randomized trial, the primary outcome variable used was the time to diagnosis. As part of a previous study undertaken in 70 patients, we performed coronary computed tomographic angiography (CCTA) on 27 of these patients seen in the emergency department with chest pain (1). Based on information from that initial experience, we estimated that time from admission to the emergency department to definitive diagnosis would be: 5 h for patients with normal CCTA, 9 h for patients with severe stenosis who would undergo early catheterization after CCTA, and 20 h for patients who are evaluated by the standard diagnostic protocol. To detect a 25% reduction in emergency department length of stay (until definitive diagnosis), approximately 102 patients would be required to achieve a power of 80% and an alpha of 0.05. We increased the sample size to 200 to ensure adequate statistical strength.
Although time to diagnosis was the determinant of sample size, clinically a diagnostic test for triage of acute chest pain would be unacceptable for use if there were a significant occurrence of major adverse cardiac events (MACE) in those who were discharged as normal. Although this safety variable is of overriding importance, it could not be used to determine sample size because the low incidence of MACE in this low-risk patient group would require a much larger sample. Our view was that even a 3% occurrence of unanticipated MACE in this preliminary study would cast doubt on the use of CCTA for acute chest pain. As reported, there were no MACEs in either group (2). A larger multicenter trial is required to investigate the issue of safety in a statistically valid way, and such a trial is currently underway.
As pointed out in the Discussion section under Limitations, we agree that alternatives to our “standard” diagnostic evaluation exist, including electrocardiographic stress or stress echocardiography, which do not involve radiation exposure and may provide faster diagnostic time. Also, the article discusses at some length issues related to the need for a second diagnostic test in 24% of patients. Regarding whether CCTA patients were “fast tracked” through the system, there was a uniform notification method for nuclear medicine and CCTA interpreting physicians; both studies were performed and read emergently.
- American College of Cardiology Foundation