Field 2021.

Study characteristics
Methods	Trial design: pilot RCT  Duration of follow‐up: median 7 years   Number of trial locations: 2
Participants	Baseline characteristics Number of participants: LDCT arm (2028); control arm (2027) Age: median age  LDCT arm (median age 67 years old): 50‐59 years old (44), 60‐69 years old (1295), 70‐76 years old (689) Control arm (median age 67 years old): 50‐59 years old (58), 60‐69 years old (1291), 70‐76 years old (678)  Sex: LDCT arm (1529 males, 499 females); control arm (1507 males, 520 females) Smoking status: LDCT arm (2 never‐smokers, 777 current, 1249 former); control arm (0 never‐smokers, 791 current, 1236 former) Performance status: not published  Ethnicity/race: LDCT arm (1992 white, 18 non‐white, 18 missing data); control arm (1992 white, 19 non‐white, 16 missing data) Environmental exposures: LDCT arm: asbestos (763); control arm: asbestos (763) Inclusion criteria Risk criteria based on the Liverpool Lung Project (LLP) Risk Prediction Model (includes age, sex, smoking duration, history of previous pneumonia, history of previous cancer, family history, exposure to asbestos) Males and females aged between 50 and 75 years old Fully informed written consent given Exclusion criteria Unable to give consent Comorbidity which would unequivocally contraindicate either screening or treatment if lung cancer were detected A CT scan of the chest performed within 1 year of the invitation to be screened Any condition precluding written informed consent Inability to lie flat Weight > 200 kg  Preintervention investigations Nil
Interventions	Intervention characteristics Frequency of scanning: annual  LDCT setting: 90 kVp to 140 kVp, mA setting adjusted to achieve volume CT dose index  Duration of screening: 1 year Interpretation of scans Volumetric or size criteria: both  Use of volumetry software: yes Criteria for significance: solid intraparenchymal nodules ≥ 15 mm3 or pleural or juxtapleural nodules with a maximal diameter of ≥ 3.1 mm. Part solid nodules. Prespecified protocol for nodule follow‐up: yes Comparison Description: no screening
Outcomes	Primary outcomes  To establish the impact of preclinical detection of lung cancer mortality by comparing lung cancer mortality between the control group and the screened groups combined To establish if there is a lung cancer mortality benefit from CT screening Establish total mortality benefit Cost‐effectiveness of a national lung cancer screening programme Secondary outcomes  To determine the physical morbidity associated with lung cancer screening To determine the resource implications of screening and the resulting intervention To assess the feasibility of population screening for lung cancer as reflected by uptake of invitations and compliance rates with annual screening Establish a blood and tissue bank for the future assessment of early detection diagnostics and novel tumour biomarkers
Identification	Sponsorship source: NIHR Health Technology Assessment programme, NIHR policy research program, Roy Castle Lung cancer foundation, Royal Liverpool & Broadgreen University Hospital Trust (UK) Country: England  Setting: hospital  Trial start date: October 2011 Completion of follow‐up: February 2020 Trial registration number: ISRCTN78513845 Corresponding author's name: John Field Institution: The University of Liverpool Cancer Research Centre, Liverpool, UK  Email: j.k.field@liv.ac.uk
Notes	Conflicts of interest  John K Field reported receiving fees from AstraZeneca (Speaker's Bureau) and advisory boards of Epigenomics; NUCLEIX Ltd. AstraZeneca, iDNA; Grant Support: Janssen Research & Development, LLC.  Robert C Rintoul reported being on the advisory boards of AstraZeneca and Roche.  David R Baldwin reported receiving speaker remuneration from AstraZeneca, Roche, MSD, BMS, Johnson and Johnson.  Kate E Brain reported receiving personal fees from Astra Zeneca outside this work. Tim Eisen reported receiving research support from AstraZeneca, Bayer, Pfizer; being employed by Roche (from March 2020) and was employed by AstraZeneca (to March 2020) and having stock in AstraZeneca and Roche; was a trustee of Macmillan Cancer Support.   Arjun Nair reported having current grants and contracts with BRC, DART; Honoraria Aidence BV, AstraZeneca; Support from BLF, and as the clinical lead for NTLHC.  No competing interests were reported from other co‐authors.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomisation using a computer‐generated random number algorithm
Allocation concealment (selection bias)	Low risk	Allocation concealment via UKLS database management system
Blinding of participants and personnel (performance bias) All outcomes	High risk	Participants were not blinded.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Low risk for primary outcome lung cancer mortality. Acknowledging the limitations of determining lung cancer mortality from death registry data without blinded committee review. All‐cause mortality and lung cancer incidence were determined without knowledge of trial allocation, since these came from routine cancer registration and death certification. Interpretation of LDCT was performed by two separate radiologists (one local and one central). The central radiologist had access to the local radiologist's report.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Missing data were not significant and participants lost to follow‐up had contact attempted by site primary investigator and if unsuccessful, the trial team contacted the participant's general practitioner for follow‐up information. It should be noted however, that 87 patients were excluded due to no consent for data linkage or having censoring events after consent. Authors were contacted, and censoring events were clarified as data were unavailable via national databases.
Selective reporting (reporting bias)	Low risk	All reported
Other bias	High risk	Minor amendments including changing nodule protocol to include new nodules detected on subsequent scans and clarification of exclusion criteria as recent CT chest.  Authors reported a computer error which used LLP risk model version 2 instead of version 1. Trial reported to use intention‐to‐treat analysis, although the 87 participants were not included in long‐term mortality and cancer incidence analysis.