ASPO Abstracts
Racial differences in lung-RADS in the National Lung Screening Trials: A Propensity Score Matching Approach
Category: Cancer Health Disparities
Conference Year: 2019
Abstract Body:
Rationale: Lung cancer continues to be the leading cause of cancer deaths in the United States. While lung cancer incidence and mortality rates have decreased for all races, major disparities persist among Blacks and Whites with early stage lung cancer. In the landmark National Lung Screening Trial (NLST), low-dose computerized tomography (LDCT) screening reduced lung cancer death by 20%. Moreover, screening with LDCT had a greater impact on reduction of lung cancer mortality in Blacks than in Whites in the NLST. However, in the NLST, patient baseline characteristics between Whites (90.9%) and Blacks (4.5%) are not comparable. Few studies have addressed the impact of race on positive results of LDCT screening. The purpose of this study is to examine racial differences in lung-RADS using propensity score matching (PSM).Methods: Approval for this project was obtained from the NCI’s Cancer Data Access System on October 16, 2017 (NLST-361). The NLST established a study population of 53,452 participants between the ages of 55-74 with at least 30 pack-years of smoking. We used PSM method to obtain a matched sample of Blacks and Whites with similar distribution of matching variables (age, gender, education, smoking status, marital status, BMI, family history of lung cancer, diagnosis of COPD). PSM was conducted using psmatch2 in Stata. Lung-RADS categorized into a binary outcome (0=negative; 1=positive). Multivariate logistic regression was used to estimate the effect of race on lung-RADS results. Results: Of a total of a LDCT arm of the NLST (n=26,722), 26,309 received their initial screening. A matched sample was selected, using k nearest neighbor matching (k=2) within a caliper of width of 0.01. The matched sample was composed of 1153 Blacks and 2071 Whites. In regression analysis, Blacks were 34% less likely to have positive results of Lung-RADS than Whites using the matched sample (OR=0.66, 95% CI= 0.53-0.84). Conclusions: In consideration for widespread use of LDCT screening in clinical practice, the definition of a positive result in LDCT screening is needed for the appropriate management of positive screening results. This suggests that more research is needed to explore racial disparities on the positive results of LDCT among vulnerable populations.
Keywords: racial disparitylung cancerNLST