Evaluating the Cost-Effectiveness of the Sybil Model for Lung Nodule Surveillance in Cancer Screening
Principal Investigator
Name
Kei Chuen Ma
Degrees
MBChB
Institution
Massachusetts Institute of Technology
Position Title
Student
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-1235
Initial CDAS Request Approval
Apr 18, 2024
Title
Evaluating the Cost-Effectiveness of the Sybil Model for Lung Nodule Surveillance in Cancer Screening
Summary
Our study conducts a follow-up analysis on the cost-effectiveness concerns arising from the NLST trial, as the high false-positive rate of the screening program introduced by this trial led to invasive tests, psychological distress, radiation exposure, and increased healthcare costs. We propose that Sybil, a deep learning tool for assessing lung cancer risk in CT scans, offers a more cost-effective approach for lung cancer surveillance compared to manual detection. Our evaluation is based on the number of CT scans required (NNS) per lung cancer case identified and the incremental cost-effectiveness ratio (ICER)
Aims
- To assess the efficacy of lung nodule surveillance conducted by the Sybil model versus traditional human radiologist practices, following the Lung-RADS reporting system.
- To determine the optimal Incremental Cost-Effectiveness Ratio (ICER and the number of CT scans required per lung cancer case identified, for both the Sybil model and human practices.
- To perform sensitivity analysis and assess the robustness of the results.
Thank you very much for your consideration.
Collaborators
Dr. Frank Schuller, MIT Jameel Clinic
Prof. Florian Fintelmann, Harvard Medical School