Community-partnered research to enhance lung cancer screening in the BJC Collaborative: A stepped-wedge Protocol

Authors: Salazar AS; Sekhon SK; Nuako, A; Lu E; BJC Collaborative; Colditz GA; James AS

Category: Early Detection & Risk Prediction
Conference Year: 2019

Abstract Body:
Purpose: Lung cancer is the leading cause of cancer-related death in the US. Although 8 million Americans qualify as high risk for lung cancer, only 2% were screened with low-dose computed tomography (LDCT) in 2016. Because prognosis is driven by stage at diagnosis, early detection is crucial to reduce mortality. This community-based study aims to assist primary care providers to increase referral for lung screening and to guide high-risk patients into appropriate care pathways. Methods: On average, monthly initial screening across participating screening centers reaches 30 patients per site. To double this rate, we will create, implement and assess a community-based intervention (toolbox). Toolbox elements will address known barriers to screening and referral, as well as required elements for screening (shared decision-making and smoking cessation counseling). They will be designed to be adaptable to the needs of different referral sites, including rural and non-academic centers. Each BJC Collaborative member health system will identify 10 primary care sites that are willing to implement the toolbox. Six providers per step per site and 9 clusters achieves 90% power to detect the monthly rate difference between intervention and control group. Using a stepped-wedge implementation design, the toolbox will be sequentially assigned to each screening site. Results: A hierarchical multilevel model including fixed and random effect, variation between clusters and between times within a clinic will be used to evaluate the intervention effect. Descriptive statistics, diagnostic plots of sites, and patient characteristics will be summarized at each time point for each cluster and the entire sample separately. The estimated intervention effect will be reported as coefficient, standard deviation, and p-values. Conclusions: Lung cancer screening rates remain unacceptably low despite guidelines recommendations. Community health systems are an important venue for building capacity and providing accessible tools to increase appropriate referral for LDCT. This multidisciplinary collaborative effort will encourage buy-in and acceptability by using local knowledge, enhancing culturally-competent communications, and promoting partnership between care providers and patients.

Keywords: Lung cancer screening, early detection, LDCT, community-based trail