Comparative cost-effectiveness of screening for lung cancer using different risk scores and thresholds: the Yorkshire Lung Screening Trial
Yingying Zhang, Sebastian Hinde, Catriona S Marshall, Neil Hancock, Irene Simmonds, Marta Soares, Rhian Gabe, Daniel Vulkan, Philip A Crosbie, Matthew E J CallisterBackground
Lung cancer (LC) is the most common cause of cancer death in the UK and worldwide, but screening with low-dose CT (LDCT) reduces LC deaths. The UK National Screening Committee has recommended nationwide roll-out of LDCT screening, but the optimal risk thresholds for eligibility remain uncertain.
Methods
We conducted a cost-effectiveness analysis in the Yorkshire Lung Screening Trial (YLST) population comparing three eligibility criteria: US Preventive Services Task Force (USPSTF) 2013 , Prostate Lung Colorectal and Ovarian study (PLCO) M2012 ≥1.51% and Liverpool Lung Project model (LLP) v2 ≥5%. A Markov model estimated a no-screening counterfactual. Scenario analyses assessed how increasing PLCO M2012 (1.51%–7%) and LLP v2 (5%–9%) thresholds affected LC detection, costs and quality-adjusted life years (QALYs). Payouts per detected LC were calculated using mortality and utility estimates from the literature and cost data from the trial.
Results
Incremental cost-effectiveness ratios (ICERs) versus no screening were £3949 (USPSTF 2013 ), £3797 (LLP v2 ≥5%) and £4013 (PLCO M2012 ≥1.51%). PLCO M2012 yielded the largest numbers screened and LCs detected, most QALYs gained and highest incremental net monetary benefit. Raising LLP v2 and PLCO M2012 thresholds reduced both ICERs and QALYs gained.
Conclusion
All three screening eligibility criteria are cost-effective according to the UK’s willingness to pay threshold of £20 000/QALY. Within the range of thresholds observed in YLST, PLCO M2012 thresholds between ≥1.51% and ≥4% offered the most efficient cost–benefit trade-offs. Evidence suggests that lowering thresholds further would detect more LC cases while remaining cost-effective. These findings support the criteria implemented in YLST, but do not by themselves identify the optimal screening threshold for the wider UK population.
Trial registration number