The Geospatial Health Research Group at Saint Louis University uses location technology and nontraditional health-related data sources to better “know” communities in ways that have been siloed in their understanding of determinants of health.
We leverage social and geographic determinants of health to identify patterns of attitudes, behaviors, and outcomes to provide guidance on how and where to intervene in order to improve community health outcomes. Disparate research has illuminated opportunities to intervene, yet few have been synthesized and operationalized in public health and healthcare systems—to identify how to respond to community needs. Geospatial tools and science have an important role to play in the indicators of the health of a community.
The Geospatial Health Data Analytics Core is a service core with the Washington University Institute of Clinical and Translational Sciences. This service core is able to consult, provide complementary data, and provide geospatial analyses to expand the social and geographic determinants of health in related studies. This consultation offers insights about current opportunities and future research questions that could be answered in collaboration with geospatial insights. Further, the data sources that we complement depend on the questions being asked. It is the nontraditional use that elevates these opportunities for answering how communities play a role in individual health behaviors and outcomes.
1. Leveraging Technology to Provide Early Sensing for Predicting an Infectious Outbreak or Pandemic
The advancements of technology have the potential to provide nontraditional public health and healthcare data from disparate sources to inform in an early sensing system that would provide real-time COVID-19 risk assessments from measures of community mobility including app-based symptom tracking and contact tracing, anonymized smartphone data, geolocated social media mentions, satellite imagery analyses to identify vehicle traffic patterns of health care locations, and geolocated search terms. Synchronizing these data sources and fusing them to develop real-time models that provide community-level COVID-19 risk. This type of early sensing system can improve health equity by providing continued real-time data and analytics to devise needed real-time interventions.
Using similar technology, including remote sensing, we are also focused on developing tools and predictive modeling of transmission of MERS-CoV from dromedary camels to humans. These prediction models are informed by human and camel interactions, land use and physical environmental changes, and overall social economic practices. Using machine learning to analyze satellite images with human and camel virologic tests, we will identify land use and socioeconomic patterns that are predictors of MERS-CoV transmission.
2. Using Geospatial Tools to Better Understand the Impact of Social Determinants of Health in Predicting Chronic Disease Management
Our team is developing and testing prediction models that inform real-time interventions within in health care plan environments that are aimed at better understanding the impact of social determinants of health through geospatial science and technology. To better know and understand how the social and physical environment that may place patients at particular risk for interruptions in chronic disease management, we can intervene faster and more efficiently by knowing more about how the environment in which people live, work and play impacts their health care management. By exploring social policy, community mobility patterns, and physical environmental factors (air and water quality), we can advise health care plans on their delivered real-time interventions to better manage chronic conditions, such as asthma on poor air quality days and informing the healthcare system of daily challenges in management of Type 2 diabetes and prioritizing health plan recipients to receive housing remediation for individuals managing childhood asthma.
3. Creating and Implementing E-Health Interventions to Support Extension of Health Care Environments
We have been working with health care systems to develop and deliver more comprehensive approaches to respond to the health care needs of their patients. This work takes the spatial modeling of social determinants of health further by providing real-time challenges that patients may experience through biometric, behavioral, and mood assessments through app-based data collection and intervention. These data sensors leverage machine learning to trigger health-care-delivered interventions.
4. Climate change and health public health and health care preparation.
With climate change patterns affecting regions across the world in varied ways, diverse expertise is needed to identify climate change's impact on humans, animals, water, and food security. Some of our studies focus on vector-borne illnesses, mapping patterns in how the vectors may shift across geography and leave communities ill-prepared to detect and prevent infections. Prior work identified locations in the contiguous United States where sexually transmitted infections and aedes-aegypti mosquitoes were more likely to occur. These studies have been adapted to explore tick-borne and other infectious diseases. Additionally, our studies in Kenya have shown how the country has become dryer and hotter over the past 40 years, creating changes in food insecurity and water patterns that place populations at risk and increase the risk of infectious diseases being transmitted from animals to humans.
5. Connecting Workforce with Location- and Skills-Based Needs
Our team is developing disaster preparedness training as part of the tracking and linking tool that provides guidance to the needs to an emergency environment. First, we are working with the nurse workforce to link together with needs of nurses in health care settings with their particular training needs in a disaster. We are also working to develop a responsive system that tracks and maintains where health care workforce is located in the time of an emergency.