2023 World Conference on Lung Cancer (Posters)-EP04.05. Improved Efficiency with Lungflag vs Opportunistic Selection in a Theoretical East Asian Lung Cancer Screening Program

EP04.05. Improved Efficiency with Lungflag vs Opportunistic Selection in a Theoretical East Asian Lung Cancer Screening Program - PDF(Slides)

Back to course

Pdf Summary

A study compared two methods for lung cancer screening in East Asia: a machine learning algorithm called LungFlag and random opportunistic selection. The study aimed to assess the efficiency and performance of these methods in detecting lung cancer cases at early stages. The population considered three groups based on smoking status: high-risk smokers meeting USPSTF2021 criteria, ever smokers not meeting the criteria, and never smokers. The simulation was conducted over a 5-year period on 1 million eligible adults aged 50 and above.<br /><br />Results showed that the LungFlag model identified significantly more cases of lung cancer than random opportunistic selection. The model flagged a total of 4,189 lung cancers, with 1,967 of them in early stages (0-II). In contrast, random selection only picked 751 lung cancers, with 233 in early stages. This represented a 5.5-fold increase in the number of cases detected and a more than 50% increase in the proportion of early-stage cancers with the LungFlag model.<br /><br />The study concluded that the LungFlag model demonstrated a higher odds ratio compared to random opportunistic selection in both ever smokers and never smokers. This suggests that the model can be used as a pre-screening tool to identify high-risk populations. Additionally, the model showed improved performance in detecting earlier-stage cancers, with a 50% improvement.<br /><br />Overall, the study highlights the potential benefits of using the LungFlag model for lung cancer screening in East Asia. With its ability to identify more cases, particularly at early stages, the model could potentially contribute to a more efficient and effective screening program for lung cancer in the region.

Asset Subtitle

Eran Netanel Choman

Meta Tag

Speaker Eran Netanel Choman

Topic Screening & Early Detection: Program Design

Keywords

lung cancer screening

East Asia

machine learning algorithm

LungFlag

random opportunistic selection

early stages

smoking status

high-risk smokers

never smokers

efficiency