Unveiling the Impact of Social and Environmental Determinants of Health on Lung Function Decline in Cystic Fibrosis through Data Integration using the US Registry

Abstract

Background: Integrating different data resources is a challenging task, but it provides a more comprehensive view of the patient information that could improve healthcare decision-making compared to relying solely on electronic health records. For example, in Cystic Fibrosis (CF), which is a genetic disorder mainly affecting the lungs, biomarkers that track lung function decline serve as important predictors for assessing disease progression and predicting hospitalization, transplantation, and mortality outcomes. It has been shown that including data sources incorporating location-specific social and environmental determinants significantly improves the accuracy of disease progression prognostication, particularly in characterizing lung function decline.

Methods: Even though people with CF lung disease represent heterogeneous socioeconomic groups from different geographic regions, lung function decline could differ in these groups. To explore these disparities in the progression of lung disease, we integrate patient registry data from the US Cystic Fibrosis Foundation with information on social and environmental health. In particular, we focus on the relation between lung function, measured as forced expiratory volume in 1 s of % predicted (FEV1) and the community deprivation index, a marker derived from six variables of the American Community Survey. Given that both outcomes are time-dependent, our methodology is built upon an extension of multivariate mixed-effects models. This approach is designed to model multiple longitudinal outcomes, incorporating varied functional forms to establish their connections. We use the area under the deprivation index curve specified at different periods in the submodel of FEV1. Examining various periods would enable us to investigate whether this relationship differs based on the duration of patients’ exposure to areas with a high deprivation index. Additionally, considering the geographical variations in both lung function decline and community deprivation, we explore this relationship within each state of the US.

Results: A strong association is observed between FEV1 and the area under the deprivation index curve across all states. The strength of this association diminishes when considering a two-year time window preceding the FEV1 measurement, as opposed to the patient’s entire medical history. Furthermore, we conducted a sensitivity analysis to explore different ways of linking these outcomes.

Conclusion: Incorporating environmental and socioeconomic markers in clinical decision-making strategies is expected to provide more insights into the progression of the disease.