Using Geomarkers and Sociodemographics to Inform Assessment of Caregiver Adversity and Resilience.

BACKGROUND AND OBJECTIVES: A high level of caregiver adverse childhood experiences (ACEs) and/or low resilience is associated with poor outcomes for both caregivers and their children after hospital discharge. It is unknown if sociodemographic or area-based measures (ie, "geomarkers") can inform the assessment of caregiver ACEs or resilience. Our objective was to determine if caregiver ACEs or resilience can be identified by using any combinations of sociodemographic measures, geomarkers, and/or caregiver-reported household characteristics. METHODS: Eligible participants for this cohort study were English-speaking caregivers of children hospitalized on a hospital medicine team. Caregivers completed the ACE questionnaire, Brief Resilience Scale, and strain surveys. Exposures included sociodemographic characteristics available in the electronic health record (EHR), geomarkers tied to a patient's geocoded home address, and household characteristics that are not present in the EHR (eg, income). Primary outcomes were a high caregiver ACE score (≥4) and/or a low BRS Score (<3). RESULTS: Of the 1272 included caregivers, 543 reported high ACE or low resilience, and 63 reported both. We developed the following regression models: sociodemographic variables in EHR (Model 1), EHR sociodemographics and geomarkers (Model 2), and EHR sociodemographics, geomarkers, and additional survey-reported household characteristics (Model 3). The ability of models to identify the presence of caregiver adversity was poor (all areas under receiver operating characteristics curves were <0.65). CONCLUSIONS: Models using EHR data, geomarkers, and household-level characteristics to identify caregiver adversity had limited utility. Directly asking questions to caregivers or integrating risk and strength assessments during pediatric hospitalization may be a better approach to identifying caregiver adversity.

Neighborhood Socioeconomic Deprivation and Health Care Utilization of Medically Complex Children.

OBJECTIVES: To assess the association between neighborhood socioeconomic deprivation and health care utilization in a cohort of children with medical complexity (CMC). METHODS: Cross-sectional study of children aged <18 years receiving care in our institution's patient-centered medical home (PCMH) for CMC in 2016 to 2017. Home addresses were assigned to census tracts and a tract-level measure of socioeconomic deprivation (Deprivation Index with range 0-1, higher numbers represent greater deprivation). Health care utilization outcomes included emergency department visits, hospitalizations, inpatient bed days, and missed PCMH clinic appointments. To evaluate the independent association between area-level socioeconomic deprivation and utilization outcomes, multivariable Poisson and linear regression models were used to control for demographic and clinical covariates. RESULTS: The 512 included CMC lived in neighborhoods with varying degrees of socioeconomic deprivation (median 0.32, interquartile range 0.26-0.42, full range 0.12-0.82). There was no association between area-level deprivation and emergency department visits (adjusted risk ratio [aRR] 0.98; 95% confidence interval [CI]: 0.93 to 1.04), hospitalizations (aRR 0.97; 95% CI: 0.92 to 1.01), or inpatient bed-days (aRR 1.00, 95% CI: 0.80 to 1.27). However, there was a 13% relative increase in the missed clinic visit rate for every 0.1 unit increase in Deprivation Index (95% CI: 8%-18%). CONCLUSIONS: A child's socioeconomic context is associated with their adherence to PCMH visits. Our PCMH for CMC includes children living in neighborhoods with a range of socioeconomic deprivation and may blunt effects from harmful social determinants. Incorporating knowledge of the socioeconomic context of where CMC and their families live is crucial to ensure equitable health outcomes.

Socioeconomic and Racial Disparities in Diabetic Ketoacidosis Admissions in Youth With Type 1 Diabetes.

OBJECTIVE: We sought to determine whether census tract poverty, race, and insurance status were associated with the likelihood and severity of diabetic ketoacidosis (DKA) hospitalization among youth with type 1 diabetes (T1D). METHODS: We conducted a retrospective population-based cohort study using Cincinnati Children's Hospital electronic medical record (EMR) data from January 1, 2011, to December 31, 2017, for T1D patients ≤18 years old. The primary outcome was admission for DKA. Secondary outcomes included DKA severity, defined by initial pH and bicarbonate, and length of stay. Exposures were the poverty rate for the youth's home census tract, parent-reported race, and insurance status. We used multivariable logistic regression to analyze effects on odds of admission. RESULTS: We identified 439 patients with T1D; 152 were hospitalized. The cohort was 48% female, 25% Black, and 36% publicly insured; the median age was 14 years. For every 10% increase in a youth's census tract poverty rate, the adjusted odds of admission increased by 22% (95% CI, 1.03-1.47). Public insurance status was associated with DKA admission (adjusted odds ratio [AOR], 2.71, 95% CI, 1.62-4.55) while race was not. There were no clinically meaningful differences in pH or bicarbonate by census tract poverty, race, or insurance status; however, Black patients experienced differences in care (eg, longer length of stay). CONCLUSION: Youth with T1D living in high poverty areas and on public insurance were significantly more likely to be admitted for DKA. Severity upon presentation was similar across exposures. Understanding contextual mechanisms by which disparities emerge will inform changes aimed at equitably improving care.

Neighborhood socioeconomic deprivation, racial segregation, and organ donation across 5 states.

One in 10 people die awaiting transplantation from donor shortage. Only half of Americans register as organ donors. In this cross-sectional study, we evaluated population-level associations of neighborhood socioeconomic deprivation and racial segregation on organ donor registration rates. We analyzed state identification card demographic and organ donor registration data from 5 states to estimate the association between a neighborhood socioeconomic deprivation index (range [0, 1]; higher values indicate more deprivation) and a racial index of concentration at the extreme (ICE) (range [-1, 1]; lower values indicate predominantly black neighborhoods, higher values indicate predominantly white neighborhoods) on organ donor registration rates within a specified geography (census tract or ZIP code tabulation area [ZCTA]). Among 26 720 738 registrants, 32% of the sample were registered organ donors. At the census tract level, with each 0.1 decrease in the deprivation index, the organ donor registration rate increased by 6.8% (95% confidence interval [CI]: 6.6%, 7.0%). With each 0.1 increase in the racial ICE, the rate increased by 1.5% (95% CI: 1.5%, 1.6%). These associations held true at the ZCTA level. Areas with less socioeconomic deprivation and a higher concentration of white residents have higher organ donor registration rates. Public health initiatives should consider neighborhood context and novel data sources in designing optimal intervention strategies.