Interobserver variability of radiographic pulmonary nodule diameter measurements in dogs and cats.

Abstract

The purpose of this study was to determine the interobserver variability of radiographic pulmonary nodule diameter measurements among readers with varying levels of experience. Because interobserver variability may lead to inaccurate estimations of nodule growth on repeat radiographic assessment, an incorrect presumption of malignant etiology or misclassification of tumor response to treatment may result. The maximum diameters of 47 pulmonary nodules from 22 dogs and 7 cats were measured. Measurements were performed using one digital thoracic radiographic projection by eight clinicians. The eight clinicians included two interns, two residents, two board-certified veterinary specialists, and two board-certified veterinary radiologists. A mixed-effect analysis of variance model was used to evaluate the contribution of reader, experience level, patient, nodule, and nodule size to the overall variability in mean pulmonary nodule diameter. The interobserver variability in diameter measurement for any given nodule was 16%, and experience level and nodule size classification did not contribute to measurement variability. Linear measurements of the diameter of a pulmonary nodule can vary significantly among a group of clinicians; however, depending on the criteria used to evaluate nodule growth or tumor response, the 16% interobserver variability reported here is likely not clinically significant.