2021 LCRF Research Grant on Disparities in Lung Cancer

Loretta Erhunmwunsee, MD
City of Hope Comprehensive Cancer Center
Research Project:
Using social determinants of health to predict adherence of lung cancer screening in minority high-risk smokers
Summary:
Lung cancer is the number one cancer killer worldwide. Lung cancer screening (LCS) through low dose CT (LDCT) allows early detection of lung cancer in high risk smokers. Guidelines recommend annual LDCTs in order for survival benefit to be achieved. Unfortunately, adherence (obtaining LDCTs annually) is as low as 40% in some marginalized communities. Because the burden of lung cancer is already disproportionately carried by racial and ethnic minority groups, it is important to understand what factors may prevent these communities from adhering to these guidelines. Clearly, without guideline adherence, the survival benefit from LCS will not be realized in underserved communities and non-small cell lung cancer (NSCLC) disparities may widen.
This study proposes to evaluate the individual and neighborhood level social determinants of health that impact LCS adherence in minority individuals and develop a predictive tool that uses state-of-the-art artificial intelligence/machine learning technologies to determine which individual will be at high-risk for non-adherence. We will recruit 300 minority smokers who underwent their first LDCT in 2018-2020 and collect information on whether they completed their follow up scan within 15 months later, their age, race, sex, income, education, perceived discrimination, smoking status, financial and food insecurity. We will also obtain neighborhood level socioeconomic data based on the participants’ addresses. We will then use all of this data to develop a prediction model to identify people unlikely to present for their recommended LDCT. The accuracy of the model will be evaluated in a second group of 180 minority individuals who underwent their LDCT in 2021. Completion of the current proposal will inform the development of a LCS navigation intervention for individuals identified as high-risk for non-adherence.
* * This project was awarded the LCRF William C. Rippe Award for Distinguished Research in Lung Cancer, acknowledging the investigator whose proposal not only demonstrated exceptional scientific merit but also exemplified an enduring commitment to making an impact in the field of lung cancer research.
Final report:
The goal of this project was to assess perceived discrimination and adverse social determinants of health (SDH) faced by minority smokers who have undergone LDCT between 2017 and 2021 and in year 2021 and 2022 and document their adherence to subsequent recommended scans. A total of 162 eligible patients were recruited. In the first group 62% did not adhere to the LDCT protocol and in the second group it was 54%. In a self-reported survey 66% were worried about developing lung cancer and 47% experienced personal discrimination. In Aim 2 a machine learning model was created to predict the probability that an individual would not complete the recommended follow-up LDCT after their first scan. The analysis indicated that participants with lower quality school systems, a higher percentage of individuals below the poverty line, and a higher area deprivation index in their neighborhood, lower educational attainment, and greater perceived difficulty in maintaining their current household income, were at a higher risk of non-adherence to LDCT.
Impact:
Lung cancer screening has resulted in earlier detection of lung cancers and a reduction in mortality. It is understandably very important that patients at risk for lung cancer not only be screened but follow through with the established protocol. There is great concern that certain patient groups are at higher risk with regard to following the protocol. A reliable model could be important in identifying the patients who are most likely to not comply. This information could be invaluable in developing an intervention that will help these patients follow through with their screening. This is a potentially lifesaving intervention.
