Standardizing Methods of Reading CT Maximum Aortic Diameters Amongst Experts Reduces Variations and Discordance, Improving Accuracy.

BACKGROUND: There is no consensus on the method of obtaining abdominal aortic aneurysm (AAA) maximum diameters based on computed tomographic angiography, and the reproducibility and accuracy of different methods have recently been debated due to advancements in imaging. This study compared the two most common methods based on orthogonal planes and centerline of flow to determine the discordances and accuracy amongst experiences readers. METHODS: The computed tomographic angiography max diameters of 148 AAAs were measured by three experienced observers, including a vascular surgeon, a radiologist and an imaging cardiologist. Observers used two different methods with standardized protocols: multiplanar reformations based on orthogonal planes, and a software using 3D aortic reconstructions to create centerline flow lumen providing diameters based on cross sections perpendicular to this lumen. Agreements and reliability of measurement methods were assessed by intra-class correlation coefficient and Bland - Altman analysis. Discordances between measurements of the methods and the original reported measurement, as well as outside hospitals were compared. RESULTS: The average age of the cohort was 75 years and aortic diameters ranged from 3.8 to 9.6 cm. For orthogonal readings, there were agreements within 3 mm between 86% and 92% of the time, while centerline - reading agreement was between 88% and 94%, which was not statistically significant. The intra-class correlation coefficient was high between method type and between readers. Within methods, agreement was between 0.96 and 0.97, while within - reader agreement measures was between 0.96 and 0.98. In comparison to the original and the outside hospital reports, 10% ≥ of the original and 20% ≥ of the outside hospital reported measurements were discordant between the readers. CONCLUSION: Maximal AAA measurements can have substantial variability leading to clinical significance and change in patient management and outcomes. Based on the results, orthogonal and centerline measurement methods have equally high agreements and concordance within 3 mm and low variations at a high volume center. However, when compared to the official read reports, there is high discordance rates that can significantly alter patient outcomes. A standardized method of measurement maximum diameter can reduce variations and discordances among different methods.

Correlation of findings from iodine 123 scan and ultrasonography in the recommendation for thyroid fine-needle aspiration biopsy.

OBJECTIVE: To assess the impact of correlating findings from iodine I 123 (¹²³I) radionuclide scans and thyroid ultrasonography on the decision to perform fine-needle aspiration (FNA) biopsy of thyroid nodules. METHODS: Iodine 123 scans and sonographic images of adult patients who had both examinations performed within 6 months of each other at our institution were retrospectively reviewed. The presence of 1 or more nodules satisfying imaging-specific criteria for recommending FNA biopsy was recorded. Iodine 123 scan and sonographic images were then directly compared to determine how frequently the FNA recommendation would be affected by discordant findings. RESULTS: The study included 97 adult patients, with a total of 291 thyroid lobes (right thyroid lobe, left thyroid lobe, and isthmus). Recommendations for FNA biopsy were concordant in 231 of 291 lobes (79.4%), with both modalities recommending FNA biopsy in 55 lobes and not recommending FNA biopsy in 176 lobes. A discordant recommendation occurred in 60 of 291 lobes (20.6%). Using only ultrasonography findings, a recommendation for FNA biopsy was not indicated for 11 of the 291 lobes (3.8%) with functioning nodules. Using only ¹²³I findings, a recommendation for FNA biopsy was not indicated for 23 of the 291 lobes (7.9%); 13 had nodules, but none that fulfilled sonographic criteria, and 10 had no identifiable nodule on ultrasonography. Iodine 123 scan did not identify 26 lobes with nodules (8.9%) for which FNA biopsy was recommended based on ultrasonography findings. CONCLUSION: Recommendations for FNA biopsy should not be based on the presence of hypofunctioning regions on ¹²³I scan without sonographic confirmation.