Chronic diffuse infiltrative lung disease: comparison of diagnostic accuracy of CT and chest radiography.

TL;DR: In this paper, the accuracy of chest radiography and computed tomography (CT) in the prediction of specific diagnoses in 118 consecutive patients with chronic diffuse infiltrative lung disease (DILD) were compared.

Abstract: The accuracies of chest radiography and computed tomography (CT) in the prediction of specific diagnoses in 118 consecutive patients with chronic diffuse infiltrative lung disease (DILD) were compared. The radiographs and CT scans were independently assessed by three observers without knowledge of clinical or pathologic data. The observers listed the three most likely diagnoses in order of probability and recorded the degree of confidence they felt in their first-choice diagnosis on a three-point scale. Confidence level 1 (definite) was reached with 23% of radiographic and 49% of CT scan readings, and the correct diagnosis was made with 77% and 93% of those readings, respectively (P less than .001). The correct first-choice diagnosis regardless of the level of confidence was made with 57% of radiographic and 76% of CT scan readings (P less than .001). The CT scan interpretations were most accurate in silicosis (93%), usual interstitial pneumonia (89%), lymphangitic carcinomatosis (85%), and sarcoidosis (77%). Observers correctly predicted whether a transbronchial or open lung biopsy was indicated with 65% of radiographs and 87% of CT scans (P less than .001). It is recommended that CT be performed before lung biopsy in all patients with chronic DILD.