Gohagan 2005.

Study characteristics
Methods	Trial design: feasibility RCT  Duration of follow‐up: 1 year active, median 5 years with database linkage Number of trial locations: 6
Participants	Baseline characteristics Number of participants: LDCT arm (1660); CXR arm (1658) Age LDCT arm: 55‐59 years old (616), 60‐64 years old (514), 65‐69 years old (337), 70‐74 years old (193) CXR arm: 55‐59 years old (624), 60‐64 years old (500), 65‐69 years old (348), 70‐74 years old (186) Sex: LDCT arm (965 males, 695 females); CXR arm (978 males, 680 females) Smoking status: LDCT arm (961 current, 699 former); CXR arm (947 current, 711 former) Performance status: not published Ethnicity/race: not published  Environmental exposures: not published Inclusion criteria Age 55‐74 years old at time of randomisation Current or former smoker who had quit within previous 10 years  ≥ 30 pack‐year smoking history Exclusion criteria History of CT chest in the previous 24 months History of lung cancer Currently receiving treatment or any cancer other than non‐melanoma skin cancer Removal of a portion of a lung or entire lung Participation in another cancer screening trial (including PLCO cancer trial) or a primary cancer prevention trial other than a smoking cessation trial Preintervention investigations Nil
Interventions	Intervention characteristics Frequency of scanning: annual  LDCT setting: 120 kVp to 140 kVp, 60 mA Duration of screening: 2 years  Interpretation of scans Volumetric or size criteria: size Use of volumetry software: no  Criteria for significance: any non‐calcified nodule ≥ 4 mm or other abnormalities suspicious for lung cancer at the discretion of the radiologist Prespecified protocol for nodule follow‐up: no Comparison Description: annual CXR
Outcomes	Primary outcomes  Feasibility of rapidly accruing high‐risk participants who were not actively being screened with spiral CT scans into a trial of lung cancer screening Willingness of participants to be randomised to either a LDCT or CXR and undergoing appropriate examination Likelihood that participants randomised to CXR would subsequently receive a spiral CT scan on their own (and vice versa) Prevalence of abnormal findings on baseline screening Extent of diagnostic follow‐up after abnormal screening findings Secondary outcomes Nil
Identification	Sponsorship source: National Cancer Institute Country: USA Setting: University hospitals Trial start date: August 2000 Completion of follow‐up: December 2002 Trial registration number: NCT00006382 Corresponding author's name: Paul Pinksy Institution: National Cancer Institute, Bethesda, USA Email: pp4f@nih.gov
Notes	Conflicts of interest  Jennifer M Croswell reported financial relationships involving spouse (husband owns shares in Johnson & Johnson) Barnett S Kramer reported receiving money from the Journal of the National Cancer Institute. Nil other disclosures reported for other authors.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomisation stratified for age (in 5‐year categories), sex and screening centre using blocks of varying sizes. Once eligibility was established and consent was obtained by a trial centre, participants were randomly assigned to a treatment group through a single, centralized, secure, web‐based system (which generated random code) operated by the trial co‐ordinating centre. This process ensured allocation concealment for trial site investigators. Randomisation was stratified by age group (in 5‐year categories), sex, and trial centre by using variable block sizes.
Allocation concealment (selection bias)	Low risk	Once eligibility was established and consent was obtained by a trial centre, participants were randomly assigned to a treatment group through a single, centralised, secure, web‐based system (which generated random code) operated by the trial co‐ordinating centre. This process ensured allocation concealment for trial site investigators. Randomisation was stratified by age group (in 5‐year categories), sex, and trial centre by using variable block sizes.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Participants were not blinded.
Blinding of outcome assessment (detection bias) All outcomes	High risk	Assessors were not blinded to any of the outcomes. Lung cancer‐related deaths were determined by death certificate during the trial, with a registry linkage performed in 2007 for long‐term follow‐up data.
Incomplete outcome data (attrition bias) All outcomes	Low risk	91 of initial 3409 participants initially randomised to LSS (46 participants in intervention, 45 participants in control), were subsequently found to be ineligible due to participation in  the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO, Oken 2011). Analysis excluded this cohort.
Selective reporting (reporting bias)	Low risk	All outcomes reported
Other bias	Low risk	There was a change in definition of positive scan between T0 and T1 scans, however this was unlikely to have impacted outcomes.