Computed tomography screening for lung cancer: preliminary results in a diverse urban population

The purpose of this study was to describe the baseline characteristics and results of the initial 18 months of our clinical computed tomography (CT) lung cancer screening program in an ethnically diverse, poor, predominantly overweight, and obese population, which differs dramatically from the National Lung Screening Trial population.

All patients had a physician referral for CT lung cancer screening and met National Lung Screening Trial eligibility criteria. Infrastructure developed for the program included a standardized results report [Bronx score of 1 to 5 (modeled on BI-RADS)] for the electronic medical record and a dedicated bilingual screening coordinator. If the patient's insurance did not cover CT screening, a fee of $75 was charged.

A total of 320 patients [54% (174) men, mean age 64 y] underwent initial CT lung cancer screening from December 18, 2012 to July 3, 2014. The median pack-years was 47, and 68% (218) were current smokers. Twenty-six percent (84) were white, and 70% (223) were overweight (101) or obese (122). The lung cancer prevalence was 2.2% (7/320). Seventy-eight percent (7/9) of patients with CT findings positive for lung cancer (score 5a, 5b) had proven lung cancer; 1 had stage 1 (1B) disease, and 6 had stage IIA or higher disease. The false-positive rate for a Bronx score ≥3 was 19% (60). Medicare and Medicaid insure 80% of the institution's overall population but only 38% (121) of the CT screening patients.

CT screening is feasible in a diverse inner-city population with the support of a robust infrastructure. Further study is needed to determine whether CT screening will confer a mortality benefit in this population.