Search
Close
Search
Advanced Search
Search Menu
AI Discovery Assistant

Extract

We read with interest the recent meta-analysis of coronary CT angiography (CCTA) in the emergency department (ED) by D'Ascenzo et al.1 Our group has conducted a similar analysis2, and some important comparisons should be discussed.

First, in regards to data pooling, the authors state that ‘no difference was noted between random and fixed effects models’. However, the endpoint of increased invasive coronary angiography (ICA) after CCTA is significant in a fixed effects model. Such a method would be more appropriate in the setting of low or no statistical heterogeneity and would result in a more precise estimate (smaller confidence interval). Additionally, they evaluated downstream ED visits and hospital admissions but not downstream ICA and coronary revascularization. This is of clinical significance as many chest pain patients may be ‘ruled out’ in the ED, but undergo ICA and/or revascularization for a significant coronary lesion in close outpatient follow-up. If including downstream ICA, most of which will be driven by the index CCTA, the odds of ICA becomes significant [1.36 (1.02–1.83, P = 0.03)] whether by fixed or random effects and the odds of revascularization remains significant.2 This finding challenges the hypothesized but unproven benefit of CCTA—avoidance of downstream ICA due to its higher sensitivity, which may not hold true in all populations studied.3 We feel the conclusion that CCTA ‘seems not to increase subsequent invasive coronary angiographies’ is inconsistent with the trend of increased ICA in all studies, overlooks ICA in near term follow-up, and, thus, might misguide the reader. Of the multiple ways to evaluate ICA after CCTA based upon these data, the authors' method is an outlier in that it does not reach significance.

Issue Section:
Letter to the Editor
You do not currently have access to this article.
Download all slides