Use and impact of risk-based eligibility models in low-dose computed tomography lung cancer screening: a systematic review

Objectives Low-dose computed tomography lung cancer screening (LDCT-LCS) significantly reduces mortality, yet identifying high-risk individuals while reducing over-screening remains challenging. Risk-based eligibility models are promising to optimize participant selection. Within the European context, we assessed the types, outcomes, and impact of these risk-based eligibility models for LDCT-LCS. Methods We systematically reviewed prediction model studies (PROSPERO CRD42025648906) across EMBASE, MEDLINE, and Cochrane Central Register of Controlled Trials. We included original research on adults aged 18+ at risk for LC, excluding East-Asian populations. Study characteristics, model type, performance and outcomes were extracted for narrative synthesis. Results The review included 46 articles (2003–2025), identifying 39 risk-prediction models. Models were primarily statistical (72%); PLCOm2012 was most frequent. Age (100%), smoking duration (91%), and intensity (72%) were the most common variables. Risk models improved eligibility and demonstrated cost-effectiveness over traditional criteria, though heterogeneity and population-specific calibration remain challenges. Conclusion Risk-based eligibility models improve LDCT-LCS efficiency by enhancing detection rates and personalization. While PLCOm2012 is prominent, addressing model heterogeneity, ensuring population-specific validation, and calibration are crucial to optimize LDCT-LCS outcomes in Europe. Systematic Review Registration Identifier CRD42025648906.