Abstract/Description

Background: In Virginia’s Hampton Roads (HR) region, minority populations experience notably higher mortality rates than white populations. Despite abundant research on the association between social determinants of health (SDOH) and cardiovascular mortality rates at the state and national levels, there is a lack of research examining this association in the HR region.

Objective: The main objective of this study is to train and validate a geospatial-temporal predictive model to examine the association between SDOH and cardiovascular mortality, in the HR region.

Methods: Utilizing a longitudinal dataset (2010–2021) from the Virginia Department of Health and the U.S. Census Bureau, we trained a geospatial-temporal negative binomial regression model to predict cardiovascular mortality. Standardized SDOH variables – such as affordability index, median income, low educational attainment, Townsend index, and low birthweight – were used as fixed-effects, while census tract and year were treated as random-effects. The model was trained on 2010–2019 dataset and validated using 2020–2021 dataset, with accuracy evaluated through root mean square error (RMSE). ArcGIS maps were generated to compare the observed versus and predicted distributions of cardiovascular mortality.

Results: Low educational attainment, Townsend index, and low birthweight were significant predictors of cardiovascular mortality (RMSE = 6.27). Correlations among the SDOH variables ranged from .003 (low) to .474 (moderate) for the model.

Conclusion: This study highlights the influence of SDOH on health disparities, demonstrating their effectiveness in predicting cardiovascular mortality. The results point to the necessity for targeted public health interventions focusing on education, income inequality, and maternal health.

Presenting Author Name/s

Mohan D Pant

College/School/Affiliation

Joint School of Public health

Share

COinS
 

Predicting Cardiovascular Mortality in Hampton Roads using Social Determinants of Health

Background: In Virginia’s Hampton Roads (HR) region, minority populations experience notably higher mortality rates than white populations. Despite abundant research on the association between social determinants of health (SDOH) and cardiovascular mortality rates at the state and national levels, there is a lack of research examining this association in the HR region.

Objective: The main objective of this study is to train and validate a geospatial-temporal predictive model to examine the association between SDOH and cardiovascular mortality, in the HR region.

Methods: Utilizing a longitudinal dataset (2010–2021) from the Virginia Department of Health and the U.S. Census Bureau, we trained a geospatial-temporal negative binomial regression model to predict cardiovascular mortality. Standardized SDOH variables – such as affordability index, median income, low educational attainment, Townsend index, and low birthweight – were used as fixed-effects, while census tract and year were treated as random-effects. The model was trained on 2010–2019 dataset and validated using 2020–2021 dataset, with accuracy evaluated through root mean square error (RMSE). ArcGIS maps were generated to compare the observed versus and predicted distributions of cardiovascular mortality.

Results: Low educational attainment, Townsend index, and low birthweight were significant predictors of cardiovascular mortality (RMSE = 6.27). Correlations among the SDOH variables ranged from .003 (low) to .474 (moderate) for the model.

Conclusion: This study highlights the influence of SDOH on health disparities, demonstrating their effectiveness in predicting cardiovascular mortality. The results point to the necessity for targeted public health interventions focusing on education, income inequality, and maternal health.