Variability in Reporting of Incidental Findings Detected on Lung Cancer Screening

Abstract

To the Editor:Information on incidental findings (IFs) detected on lung cancer screening (LCS) in real-world settings is limited.The American College of Radiology (ACR) developed the Lung Computed Tomography Screening Reporting and Data System (Lung-RADS) lexicon to standardize reporting of lung nodules and clinically or potentially clinically significant IFs using an S modifier.The estimated prevalence of Lung-RADS S modifiers is 10% (1); however, "significant" is not defined, and there are scant data on the use of the S modifier in practice (2).Hence, we sought to 1) describe the most commonly reported IFs and 2) compare variability in reporting of IFs using the S modifier with and without IF-specific follow-up recommendations, overall and by patient-level (age, sex, race, smoking status, Lung-RADS assessment), provider-level (radiologist training), and facility-level (imaging location, facility type) characteristics.We hypothesize that radiologists report significant IFs by using the S modifier or recommending IF-specific workup in the radiology report and that use of the S modifier will differ by patient and radiologist characteristics, with IFs more common in older versus younger patients and those who currently versus formerly smoked.Characterizing the reporting of IFs in subgroups is needed to understand how Lung-RADS is used in clinical practice. MethodsWe conducted a prospective cohort study of individuals undergoing baseline LCS at six imaging locations in North Carolina from 2014 to 2020.From the radiology report, we abstracted LCS examination date, interpreting radiologist (general or cardiothoracic imaging fellowship trained), type of IF, Lung-RADS assessment, and recommended follow-up.We categorized IF on LCS examinations as follows: 1) the radiologist used the Lung-RADS S modifier and recommended IF-specific follow-up, 2) the radiologist used the S modifier and did not recommend IF-specific follow-up, 3) the radiologist did not use the S modifier but recommended IF-specific follow-up, and 4) the radiologist did not use the S modifier and did not recommend IF-specific follow-up.We focused on these IF categories to capture the IFs most likely to be actionable.We assessed variability in IF reporting using these four definitions by age, sex, race, smoking status, chronic obstructive pulmonary disease (COPD), hypertension, diabetes, radiologist training, year of LCS examination, Lung-RADS assessment, imaging location, and facility type using 10,000 Monte Carlo simulations to estimate Fisher's exact test (3).We described the types of IFs reported on the basis of the S modifier or recommended IF-specific follow-up.The Institutional

Topics & Concepts