Trials, including the Dutch-Belgian lung cancer screening trial (NELSON), have demonstrated that low-dose CT screening reduces lung cancer mortality among long-term (former) smokers. However, a significant proportion of individuals will not derive benefit from screening. Some may have a low risk of developing lung cancer despite their smoking history and age, while others may have limited life expectancy, precluding any survival advantage. Therefore, optimizing the selection of individuals who are most likely to benefit remains a critical challenge. Additionally, many screened individuals undergo unnecessary diagnostic procedures for nodules that are ultimately benign.

In recent years, risk models have gained increasing recognition for improving CT screening efficiency by identifying individuals at the highest risk of lung cancer while ensuring adequate life expectancy, as well as for assessing the malignancy risk of detected nodules.

There is growing interest in incorporating novel data sources to enhance risk models for both participant selection and nodule evaluation. Potential improvements include genetic markers and CT-based biomarkers beyond lung nodules. Recently, air pollution has also been identified as a potential predictor of lung cancer risk. However, no validated risk prediction model currently integrates these biomarkers.

NELSON-POP aims to improve the efficiency of lung cancer screening by identifying high-risk participants, and by improving cancer risk discrimination for nodules, by developing multi-source data models.