Comparison of two- vs. three-view thoracic radiographic studies on conspicuity of structured interstitial patterns in dogs

Three-view thoracic radiography is often used to evaluate patients for pulmonary metastatic disease. Although use of three views has been reported to be more sensitive than two views for focal lung disease, it also requires increased time, effort, and radiographic exposure of patients and personnel. This study was performed to evaluate the conspicuity of lesions on two-view vs. three-view radiographic procedures to determine the proportion of diagnoses that would change. One hundred three-view radiographic studies of the canine thorax were randomized, and four protocols were reviewed for each study: right lateral and ventrodorsal views, left lateral and ventrodorsal views, both lateral views, and all three views. Radiographs were interpreted as either positive or negative for structured interstitial pulmonary disease, and the certainty of the reading was recorded using a visual analog scale. There was 85-88% agreement between each two-view group and the three-view group, with the κ statistic ranging from 0.698 to 0.758. There were no differences in certainty of diagnosis among the groups, though within each group there was more certainty for positive diagnoses than negative diagnoses. These findings indicate that three-view studies should be continued when evaluating for possible structured interstitial pulmonary disease, including metastatic disease, as eliminating one view from a three-view study would change the diagnosis in 12-15% of patients.