Harmonizing High-Resolution Earth Observation and Health Data to Examine Associations between Extreme Heat and Asthma Exacerbations in Baltimore, MD

The diversity and resolution of satellite-borne remote sensing technologies has created tremendous opportunities for the application of satellite data for societal benefit across scales, including characterizing risks to human health. Earth observation (EO) measurements, when paired with health data, allow for associations between health and climate change to be better understood. With our collaborators at the Johns Hopkins Asthma Precision Medicine Center of Excellence (PMCOE), we created a database that integrates medical health records with EOs at various spatial and temporal resolutions. Using this database, we seek to characterize the effect of extreme heat on asthma exacerbation rates during the summer months of June, July, and August between the years 2016 and 2022. Patients from the Johns Hopkins Asthma PMCOE who currently reside in Baltimore City are included in our study and we identified 13,068 asthma exacerbations among 2,573 adults and 3,184 children. We use a distributed lag quasi-Poisson regression by including daily meteorological and air pollution values from 1 to 10 days prior to the exacerbation event to determine how asthma exacerbation evolves over time. Asthma exacerbation rates are modeled as a function of static physical characteristics and time-varying meteorological and air quality variables. Exacerbation data and meteorological exposures are aggregated at three spatial resolutions: block group, census tract, and county. We then perform our analysis at each of these resolutions to determine if coarser spatial resolution can capture the exposure-response relationship well. This study will allow us to determine how the strength of association changes with spatial resolution. These findings will help to inform assessments of the utility of earth observation data at different spatial scales and provide insight regarding the granularity of health data needed for meaningful health and environmental justice analysis.