Tailored Patient Navigation to Improve the Uptake of Lung Cancer Screening in Tribal Communities in Western Washington State, SACRED LUNGS Trial

Will be compared between arms using the Cochran-Mantel-Haenszel (CMH) test, accounting for the stratification variables used during randomization. Completion rates will also be estimated in each arm and 95% confidence intervals for the rates and their difference will be calculated using Wilson's score method. The single primary outcome will be tested using two-sided α = 0.05.

Will be compared between arms using the CMH test. Will be tested using two-sided α = 0.05 without adjustment for multiplicity.

Will be compared between arms using a stratified Wilcoxon rank-sum test (van Elteren test). Will be tested using two-sided α = 0.05 without adjustment for multiplicity.

Will be measured using the Short Form 36 Health Survey questionnaire. Will be compared between arms using a stratified Wilcoxon rank-sum test (van Elteren test). Will be tested using two-sided α = 0.05 without adjustment for multiplicity.

The impact of navigation on tobacco cessation will be explored by the rates of a reported 30-day cessation period during the intervention period as measured by self-reported 30-day floating abstinence.

Quantitative analyses will use structural equation models (SEM) to understand the relationships between measured and latent variables representing barriers, facilitators and social determinants of health and the outcome of CT completion.

Quantitative analyses in this aim will use SEM to understand the relationships between measured and latent variables representing barriers, facilitators and social determinants of health and the outcome of navigation care engagement. Measure of reach will be operationalized as an ordinal scale based on care completion activities along the LCS continuum as documented in the electronic health record.

The rate of chest CT completion during the maintenance phase in the control arm (after navigation is offered) will be estimated and compared with the rate during the trial phase (before navigation was offered) using Poisson regression with the log link. The phase (trial versus maintenance) will be modeled using a time varying covariate, where each participant will contribute an outcome for each phase in which they are still "at risk" for completing CT (i.e., until first CT completion or end of follow-up). Models will include additional key covariates, including clinical site, smoking status, and prior reported LCS. Generalized estimating equations will be used to account for multiple periods per participant and potential overdispersion.

Adherence to follow-up will be assessed as binary completion of the next step in follow-up among those with a completed chest CT during the randomized controlled trial. Appropriate next step in follow-up is determined through the Lung CT Screening Reporting and Data System (Lung-RADS) system, with most participants having negative initial CT (Lung-RADS 1 or 2) with follow-up assessed at 11-15 months. Adherence rates will be estimated as a proportion with 95% confidence intervals, summarized overall and stratified by Lung-RADS.