Risk prediction for lung cancer screening: a systematic review and meta-regression
Rather than developing new lung cancer prediction models, research should focus on validating and comparing the 91 existing models for practical use.
This updated systematic review cataloged 91 lung cancer risk prediction models developed between 2020-2026 covering pre-screening eligibility and post-screening nodule classification, finding discrimination ranges from AUC 0.70 to >0.90 with biomarker-enhanced models often superior. The authors call for prioritizing external validation and head-to-head comparison of existing models rather than continued development of new ones, given that most lack implementation evidence.
What the study was
- Study design
- systematic_review_meta_regression
- Category
- early_cancer_detection
- Maturity
- Validated
- Journal
- Eur Respir Rev
Why it surfaced
Timely synthesis of the expanding lung cancer screening model landscape following recent US/European guideline expansions; identifies critical quality gaps (lack of external validation, inconsistent calibration reporting) for health system decision-makers; Eur Respir Rev quality.
A plain-language summary of published research — not medical advice. Talk to a clinician about your own care.