Abstract

To examine the association between social determinant of health (SDH) factors and types of emergency department (ED) visits controlling for illness severity. We estimated the relative rate of primary care treatable (PCT) and emergency department care needed (EDCN) visits during a two-year period among Medicaid beneficiaries with different SDH risk profiles and medical conditions adjusted for the Chronic Disability Payment Score (CDPS), age and sex. We prospectively enrolled 6,719 subjects enrolled in the District of Columbia Medicaid program who had a medical encounter at two EDs, a primary care clinic or an obstetrics and gynecology clinic during a 15-month enrollment period. At enrollment, subjects completed a SDH survey that included questions about education and employment, health behaviors, living circumstances, financial difficulties, food insecurity, housing instability, mental health, and receipt of social services. We used latent class analysis to develop risk profiles based on their SDH data. We also obtained subjects’ Medicaid claims data for a two-year period and merged it with the SDH data. We used the claims data to count the total number of PCT and EDCN visits during the two-year period, to calculate the CDPS and to identify subjects with selected types of medical conditions. We separately modeled the total number of PCT and EDCN visits as a function of the SDH risk class, medical condition, age, sex and the CDPS using a generalized linear mixed effect model (GLIMMIX) with poisson distribution. The mean age of the study sample was 39, 67% were female and 91% black. Eighty percent of subjects were enrolled in the DC Medicaid program for the entire 2-year study period. The mean number of PCT and EDCN visits was 2.0 and 4.4 respectively. The latent class analysis revealed three distinct SDH profiles: group 1 (47%) reported few SDH adversities and receipt of few social services; group 2 (30%) reported more SDH adversities and receipt of more social services; and group 3 (23%) reported more SDH adversities but reported receiving few social services. As displayed in the table below, the higher SDH risk classes were associated with increased relative rates of PCT and EDCN visits compared to the lower SDH risk class.TableRegression Coefficients of GLIMMIX models With CDPS And Selected Medical Conditions.∗Compared to the low risk SDH risk class, adjusted for age, sex, and the CDPS. ‡ p ≤ 0.05.ED VisitCDPS OnlyCDPS, DiabetesCDPS, CardiovascularCDPS, RenalCDPS, PsychiatricCDPS, Substance AbuseA. Primary Care Treatable (PCT)Class 2: Higher SDH, more services0.12‡0.11‡0.11‡0.12‡0.060.09‡Class 3: Higher SDH, fewer services0.19‡0.19‡0.19‡0.18‡0.13‡0.14‡Medical condition type-0.010.020.17‡0.41‡0.45‡B. ED Care Needed (EDCN)Class 2: Higher SDH, more services0.09‡0.08‡0.09‡0.08‡0.070.07Class 3: Higher SDH, fewer services0.15‡0.15‡0.16‡0.14‡0.14‡0.10‡Medical condition type-0.14‡0.35‡0.32‡0.36‡0.49‡∗ Compared to the low risk SDH risk class, adjusted for age, sex, and the CDPS. ‡ p ≤ 0.05. Open table in a new tab Over and above illness severity and types of medical conditions, the relative rate of PCT and EDCN visits varied significantly by the socio-economic adversities Medicaid beneficiaries experience.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call