How Accurate Are Lung Cancer Risk Models for Selecting Ever-Smokers for Screening?
Posted: Tuesday, July 10, 2018
A recent study, published in the Annals of Internal Medicine, compared nine lung cancer risk models and found widely different U.S. screening populations. Of these models, four of them—the Bach; the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial Model 2012 (PLCOM2012); the Lung Cancer Risk Assessment Tool (LCRAT); and the Lung Cancer Death Risk Assessment Tool (LCDRAT)—were the most accurate in predicting risk of lung cancer and “performed best” in selecting ever-smokers for screening.
“Effectively and efficiently targeting lung cancer screening to persons at highest risk can further reduce lung cancer mortality,” noted the study authors, led by Hormuzd A. Katki, PhD, of the National Cancer Institute. “[The four lung cancer risk models identified] should be further refined to improve their performance in certain subpopulations.”
The nine lung cancer risk models were the Bach model, the Spitz model, the Liverpool Lung Project (LLP) model, the LLP Incidence Risk Model, the Hoggart model, the PLCOM2012, the Pittsburgh Predictor, the LCRAT, and the LCDRAT. Model performance was evaluated using data from two cohorts: 337,388 ever-smokers in the National Institutes of Health–AARP Diet and Health Study and 73,338 ever-smokers in the Cancer Prevention Study II Nutrition Survey.
The models chose U.S. screening populations ranging from 7.6 million to 26 million ever-smokers, at a 5-year risk threshold of 2%. However, the four best performing models demonstrated better agreement on the size of the screening population (7.6 million to 10.9 million) and achieved consensus on 73% of individuals chosen. Compared with these models, the other five generally overestimated risk and had lower AUCs.