Geospatial

Geospatial mapping of SDoH is instrumental to investigating and mapping potential variations in disruption, mitigation and clinical outcomes by vulnerable population and community that is difficult to discern through other techniques. These data sets provide valuable local/regional and even real-time information that is impossible to be obtained from medical records. However, they also introduce a new challenge which is how to use them effectively. There are several problems associated with the use of these diverse data sets such as the different levels of spatial resolution (state, county, zip code, census block), sparsity (missing spatial and time information), and noise. The most significant challenges are how to address the high dimensionality which may lead to overfitting and difficulties with model interpretation, and how to incorporate the neighboring influence into the model.

Suicide and Overdose Prevention

Suicide is a major public health problem affecting US Veterans and the US in general. Veterans suicide increased by 57% from 2001 to 2019, compared with a 20% increase among civilians.

Many variables (e.g., demographic, clinical, biological, geographic) have been associated with risk for suicide and suicidal behavior, including altitude, rurality, social and environmental determinants of health (SDoH and EDoH);

Previously, our group used both geospatial data (county and zip codes) and individual-level data to comprehensively assess the association between altitude and suicide mortality, suicide attempts, and suicidal ideation among US Veterans between 2000 and 2018, we demonstrated that there was a strong correlation between altitude and suicide rates at all the levels investigated and using different statistical analyses and even after controlling for significant covariates such as percent of age >50yr, percent male, percent white, percent non-Hispanic, median household income, and population density by using multi-linear regression model (Fig. A).

We also used geographic information and temporal effects to integrate SDoH (e.g., employment, commute time, crime) and EDoH features(air quality, climatology) with EHR to identify geospatial clusters with high suicide-specific rates (SSR) which refers to suicide deaths, attempts, and ideation (Fig. B)..

We aim to develop prioritized and facilitate county-level areas by using community-based rural Veteran suicide maps that identify the vulnerability area with top selected SDoH and EDoH features, percent veterans health administration enrollees residing in rural communities, percent population that are veterans, crude suicide mortality rate, the distance between enrollees with nearby VA facilities/ VISN, waiting time for nearby VA MH facilities, psychiatric morbidities.  We had the preliminary results shown in Fig C of the Vulnerability area in several states. We will create our geospatial and temporal model to improve our prediction.