Reproducibility of the main pulmonary artery to ascending aorta ratio (PA:A) in real-world, non-research chest CT scans

Recent studies have shown research CT scan PA:A ratios >1 to correlate with pulmonary hypertension (PH) and be predictive of acute exacerbations of COPD (AECOPD). This predictive ability has yet to be demonstrated using non-research CT scans at the community level. Before determining if the PA:A of real-world CT scans could be used to predict AECOPD in community practice, we felt it important to measure the reproducibility of the PA:A as determined by non-radiologists in the community setting. From a database of 1040 clinical scans, 33 scans from 6 subjects (4 COPD, 1 asthma, 1 PH) were selected to measure the PA:A. A medical student (Rater 1) and a mid-career gynecologist (Rater 2) performed 3-4 blinded randomized determinations of the PA:A, and a general radiologist completed 1 determination. Bland-Altman analysis, intraclass correlation coefficient and Cohen’s kappa were determined (Table 1).In this feasibility study using real-world CT scans, we found the PA:A had very good intra- and inter-rater reliability in the range of published studies using research CT scans. The non-radiologist raters improved with experience and achieved results similar to the radiologist after 3-4 determinations. Our results demonstrate community CT scan PA:A is reproducible suggesting the PA:A from real-world CT scans may be predictive of AECOPD. Confirmation of this notion awaits further study.
Source: European Respiratory Journal - Category: Respiratory Medicine Authors: Tags: Imaging Source Type: research