Ascertaining Multilevel drivers of hEad and Neck cancer Disparities (AMEND)

About This Grant

Project Summary/Abstract Modifiable multi-level determinants of head and neck cancer (HNC) outcomes have persisted for decades, and groups have differed in survival rates and outcomes. Unfortunately, most HNC studies have focused on individual-level factors to explain outcomes, and there has been a paucity of research to understand the complex, but modifiable multi-level determinants that underline HNC outcomes. To address this research gap and in direct response to PAR-25-095; the Ascertaining Multilevel drivers of hEad and Neck cancer-related health Disparities (AMEND) project will utilize a cellsto- society framework, and robust secondary, longitudinal, and linked datasets (SEERMedicare and SEER-Medicaid), to identify multi-level determinants of head and neck cancer (HNC) outcomes. Advanced statistical techniques will be employed to conduct these analyses, including highly novel, deep-learning, time-to-event causal survival models. To achieve the overall project objective, we will predict multi-level (individual sociodemographic, regional and geographic) determinants of HNC outcomes; use group-based trajectory modelling to map long-term population trajectories in HNC; and quantify the effect and relative contribution of modifiable upstream and downstream factors (e.g. smoking, alcohol and other lifestyle factors, stage of presentation, treatment modality, rural/urban residency) and time-to-event factors in overall and disease-specific mortality among patients with HNC. Importantly, knowledge from our study will provide scientific evidence to guide both policies and multi-level interventions to target drivers of HNC outcomes, which aligns with the administration’s promotion of healthy living, and the Healthy People 2030 objective of dismantling the impact of multilevel determinants of health outcomes.