Large Variability of False-Positive Rate in National Lung Screening Trial

Background: The authors note that the National Lung Screening Trial (NLST) reported a 20% mortality benefit to 3 annual low-dose chest CT screenings compared with chest radiography. There were high rates (27%), however, of positive screenings; the majority (>90%) were false-positive nodules, which were found to be benign on repeated follow-up CT. Objective: To evaluate variability in radiologists' nodule detection rate, including evaluation of false-positive rates (FPRs) and sensitivity in the NLST. Participants/Methods: Participants were part of the NLST, which randomized people aged 55 to 74 years to either low-dose CT or chest radiography. All participants had a smoking history of ≥30 pack-years. A noncalcified nodule ≥4 mm was considered a positive finding. If the nodule was stable in size for 2 years, it was considered benign; therefore, a positive study could be converted to a negative final interpretation. The FPR definition was rate of positive screenings in participant more...