Background
Low-dose computed tomography (LDCT) is increasingly being adopted in high-income countries in high-risk groups, typically in those with a history of smoking. Cost-effectiveness analyses (CEAs) are being used to inform this policy adoption. Lung cancer may differ by sex in several aspects, including incidence and survival outcomes, with screening trials suggesting potential sex heterogeneity in screening effectiveness. CEAs without sex stratification may therefore introduce aggregation bias, leading to suboptimal policies.
 
Objectives
To evaluate how sex is handled in model-based CEAs of LDCT lung cancer screening, including assessing if: studies employ sex-stratified modelling; key inputs are sex-specific; sex-specific outcomes favour women or men; and sex-specific optimal strategies differ in screening age range and/or interval.
 
Methods
PubMed, Web of Science and Embase were searched. Full-text, model-based CEAs of LDCT screening reporting health-related outcomes were included. We defined sex-stratified modelling as analyses that simulate separate male and female cohorts and report sex-specific outcomes. For sex-stratified studies, data on whether/what sex-specific values were applied to major inputs, and whether/what sex-specific optimal strategies were reported. Reporting quality and modelling stratification were assessed using a self-developed checklist adapted from the CHEERS 2022 guidelines.
 
Results
This review included 68 studies; only 15 (22.1%) reported sex-stratified outcomes. In 12/15 studies (80.0%), the optimally cost-effective strategy was the same for men and women, with one study showing sex differences only in sensitivity analysis. All 12 applied at least one sex-specific input, but parameterisation was inconsistently reported, and mostly limited to incidence and mortality. The remaining 3/15 (20.0%) reported sex-differential optima, including differences in screening interval and/or age range, and in one analysis, no screening was the optimal strategy in women while a strategy remained cost-effective in men. All these 3 studies incorporated multiple sex-specific inputs, including incidence, all-cause mortality, lung cancer mortality/survival, stage shift and participation, while none of them employed or reported sex-specific costs or histology inputs. Overall, more favourable cost-effectiveness estimates were found for women than men in 11/15 studies.
 
Conclusions
Sex stratification remains uncommon and variably implemented in CEAs of LDCT lung cancer screening. Even when sex-stratified outcomes are reported, key sex-specific inputs are inconsistently documented. Richer sex-specific parameterisation may contribute to divergent sex-specific screening recommendations. More routine use of sex-stratified model structures, together with transparent reporting of sex-specific inputs, cost-effectiveness estimates, and optimal strategies, may reduce aggregation bias and support more precise and equity-informed screening policy.