Health economic evaluation of lung cancer screening using a diagnostic blood test: the Early detection of Cancer of the Lung Scotland (ECLS)

BackgroundDiagnostic blood tests have the potential to identify lung cancer in people at high risk, which is important as detecting lung cancer at an early stage is associated with survival advantages. We conducted an economic evaluation to assess the cost-effectiveness of a lung cancer screening intervention, using the EarlyCDT(R)-Lung Test with subsequent x-ray and low-dose chest CT scans (LDCT) for patients with a positive test result, compared to both usual care and LDCT screening for the whole target population. MethodsA lifetime analyses with a UK NHS and personal social services perspective was conducted using a decision model for a target population of 1,000 individuals, where model parameters came from the ECLS study and literature. The model simulated the probability distribution of stage at cancer detection (early vs. late) for each evaluated alternative. Quality adjusted life-years (QALY) assigned to patients were dependant on stage at detection during or after the screening period, costs were dependent on the diagnostic pathway followed by patients and on cancer stage at diagnosis. We estimated net monetary benefit (NMB) at policy relevant cost-effectiveness thresholds for base-case, deterministic sensitivity, and scenario analyses. ResultsThe base case incremental NMB of the ECLS intervention compared to no screening was {pound}33,179 (95% CI: -{pound}81,396.4, {pound}147,180) and {pound}140,609 (95% CI: {pound}36,255.1, {pound}316,612) respectively for a cost-effectiveness threshold of {pound}20,000 and {pound}30,000 per QALY. The same figures compared with LDCT screening were {pound}162,095 (95% CI: {pound}52,698.3, {pound}271,735) and {pound}52,185 (95% CI: -{pound}113,152, {pound}220,711). A deterministic sensitivity analysis indicated parameter values that resulted in a change to cost-effectiveness results, for example: prevalence of lung cancer (1%-4%); relative prevalence of early stage lung cancer (25%-75%); cost of the EarlyCDT(R)-Lung Test ({pound}59-{pound}201.5); test sensitivity for early stage lung cancer (25%-75%); and specificity of the test (50%-100%). A scenario analysis confirmed that the EarlyCDT(R)-Lung Test performs better than a zero-cost random test and showed that if the sensitivity of the test is assumed 25% (rather than base case 52%) the ECLS intervention would be not cost effective at a {pound}30,000 per QALY threshold. LimitationsLack of trial resource data for the within study analysis resulted in partial reliance on expert opinion. Some participants may have modified their smoking behaviour due to participation in the trial. ConclusionsThe base case analysis results estimated that the ECLS intervention is the most cost-effective screening alternative, with highest probability of being cost-effective, when compared to no screening or LDCT screening. This result may change with modifications of the parameters, prevalence of lung cancer and EarlyCDT(R)-Lung Test cost, suggesting that the three alternatives considered in the main analysis are potentially cost-effective depending on the disease risk of the target population and the cost of testing.