Exploring the Relation between Contextual Social Determinants of Health and COVID-19 Occurrence and Hospitalization

Aokun Chen, Yunpeng Zhao, Yi Zheng, Hui Hu, Xia Hu, Jennifer N. Fishe, William R. Hogan, Elizabeth A. Shenkman, Yi Guo, Jiang Bian

TL;DR

This study investigates how neighborhood-level social and built environment factors (contextual SDoH) are associated with COVID-19 occurrence and hospitalization. By linking electronic health records with extensive SDoH data, the research identifies specific contextual factors that predict higher or lower risks, offering insights for targeted public health interventions.

What Problem Does This Paper Address?

The COVID-19 pandemic highlighted the profound impact of social determinants of health (SDoH) on disease outcomes. However, previous research has often been limited by geographically sparse data or a narrow focus on a few SDoH factors, leaving a knowledge gap regarding the complex interplay between a wide range of contextual SDoH and COVID-19 incidence and hospitalization. This paper aims to address this by comprehensively exploring the associations between a large number of contextual SDoH variables and COVID-19 outcomes within a major clinical research network.

Methods

The study employed a two-phase generalized linear mixed-effect model (GLMM) approach, adapted from exposome-wide association studies (ExWAS).

  • Phase 1: Each contextual SDoH variable was analyzed individually against COVID-19 outcomes (occurrence and hospitalization) while adjusting for patient covariates and geolocation. Variables showing statistical significance in either a randomly split experiment or replication set (p < 4.386 × 10⁻⁴) were carried forward.
  • Phase 2: Statistically significant variables from Phase 1 were then included simultaneously as predictors in multivariable GLMMs. Variables that remained significant in both the experiment and replication sets were considered for final analysis.
  • Meta-analysis: A fixed-effect meta-analysis was used to generate pooled odds ratios (ORs) and confidence intervals (CIs) for the significant contextual SDoH variables.
  • Data Preprocessing: Continuous SDoH variables were normalized and standardized. Multicollinearity was addressed by iteratively removing variables with high variance inflation factors (VIF > 5). Missing data for SDoH factors and covariates were imputed using the chained equations method.

Datasets

  • Electronic Health Records (EHRs): Data were obtained from OneFlorida+, a large clinical research network (CRN) within the National Patient-Centered Clinical Research Network (PCORnet). This dataset includes longitudinal, linked patient-level data from EHRs, Medicaid claims, cancer registries, and vital statistics for approximately 15 million patients across all 67 Florida counties. The study cohort comprised 15,890 COVID-19 patients and 63,560 matched non-COVID-19 patients from January 2020 to May 2021.
  • Contextual Social Determinants of Health (SDoH) Data: 199 contextual SDoH variables were collected from various sources, covering 11 domains: vacant land, social vulnerability, area deprivation, social capital, crime and safety, hospital bed capacity, healthcare status, walkability, food access, food environment, and green space. These data were spatiotemporally linked to patients using their ZIP codes.

Key Insights and Ideas

  • Granular SDoH Matters: The study underscores that even at a neighborhood level, specific environmental and social factors can significantly influence COVID-19 risk and severity. This moves beyond broad socioeconomic indicators to more nuanced contextual elements.
  • "Built Environment" as a Predictor: Factors like the number of golf courses and country clubs, and vacant business properties, emerged as significant predictors, suggesting that the physical and economic landscape of a community plays a role in disease spread and outcomes.
  • Food Access and Health: The finding that areas with a higher percentage of farmers markets selling prepared foods were associated with lower hospitalization risk is particularly interesting, hinting at the potential protective effect of access to healthier food options.
  • Religious Institutions and Risk: The association between a higher number of religious establishments and increased hospitalization risk warrants further investigation, potentially related to social interaction patterns or vaccine hesitancy.

Key Results

  • COVID-19 Occurrence: Six contextual SDoH variables were significantly associated with COVID-19 occurrence.
  • Lower risk: Higher numbers of pediatric hospital beds, higher hospital bed utilization, and a higher percentage of vacant business properties were associated with a lower probability of having COVID-19.
  • Higher risk: Residing in regions with more golf courses/country clubs, higher murder rates, and a higher percentage of minorities were associated with a higher probability of having COVID-19.
  • COVID-19 Hospitalization: Three contextual SDoH variables were significantly associated with hospitalization among COVID-19 patients.
  • Lower risk: A higher percentage of farmers markets selling baked/prepared food products was associated with a lower probability of hospitalization.
  • Higher risk: A higher percentage of SNAP households and a greater number of religious establishments were associated with a higher probability of hospitalization.

Limitations and Future Work

  • Geographic Specificity: Findings are limited to Florida and may not be generalizable to other US regions due to variations in state policies and other contextual factors.
  • Observational Nature: The study design is observational, precluding the establishment of causal relationships between SDoH and COVID-19 outcomes.
  • EHR Limitations: Inaccuracies in ICD coding, incomplete recording of deaths outside hospitals, and assumptions of independence between SDoH factors and comorbidities are acknowledged limitations.
  • Unidentified SDoH: The study acknowledges that other SDoH factors not included in the analysis might also influence COVID-19 outcomes.
  • Future Work: The authors suggest future studies to explore the causal effects of SDoH on COVID-19 outcomes and to investigate the identified intuitive relationships further. They also plan to address EHR limitations and explore the impact of correlated SDoH factors and comorbidities more deeply in future research.

Summary updated

Abstract

It is prudent to take a unified approach to exploring how contextual social determinants of health (SDoH) relate to COVID-19 occurrence and outcomes. Poor geographically represented data and a small number of contextual SDoH examined in most previous research studies have left a knowledge gap in the relationships between contextual SDoH and COVID-19 outcomes. In this study, we linked 199 contextual SDoH factors covering 11 domains of social and built environments with electronic health records (EHRs) from a large clinical research network (CRN) in the National Patient-Centered Clinical Research Network (PCORnet) to explore the relation between contextual SDoH and COVID-19 occurrence and hospitalization. We identified 15,890 COVID-19 patients and 63,560 matched non-COVID-19 patients in Florida between January 2020 and May 2021. We adopted a two-phase multiple linear regression approach modified from that in the exposome-wide association (ExWAS) study. After removing the highly correlated SDoH variables, 86 contextual SDoH variables were included in the data analysis. Adjusting for race, ethnicity, and comorbidities, we found six contextual SDoH variables (i.e., hospital available beds and utilization, percent of vacant property, number of golf courses, and percent of minority) related to the occurrence of COVID-19, and three variables (i.e., farmers market, low access, and religion) related to the hospitalization of COVID-19. To our best knowledge, this is the first study to explore the relationship between contextual SDoH and COVID-19 occurrence and hospitalization using EHRs in a major PCORnet CRN. As an exploratory study, the causal effect of SDoH on COVID-19 outcomes will be evaluated in future studies.
BibTeX
@article{chen2024exploring,
  author = {Aokun Chen and Yunpeng Zhao and Yi Zheng and Hui Hu and Xia Hu and Jennifer N. Fishe and William R. Hogan and Elizabeth A. Shenkman and Yi Guo and Jiang Bian},
  title = {{Exploring the Relation between Contextual Social Determinants of Health and COVID-19 Occurrence and Hospitalization}},
  journal = {Informatics},
  year = {2024},
  doi = {10.3390/informatics11010004},
}
Figure