.jpg "Kayla Phelps, MD, MPH photo")
Kayla Phelps, MD, MPH
C.S. Mott Children's Hospital at University of Michigan
Ann Arbor
Disclosure information not submitted.
Errin Mitchell, MD
PICU Fellow
C.S. Mott Children's Hospital at University of Michigan, United States
Disclosure information not submitted.
Erin Carlton, MD, MSc
University of Michigan C. S. Mott Children's Hospital
Ann Arbor, MI
Disclosure information not submitted.
Erica Andrist, MD
Clinical Assistant Professor
C. S. Mott Children's Hospital, University of Michigan, United States
Disclosure information not submitted.
Title: Social Determinants of Health in Pediatric Critical Care: What Do We (Not) Know?
Introduction: There is increasing awareness of the impact of social determinants of health (SDoH) on pediatric critical care medicine (PCCM) utilization and outcomes. Studies are often limited to the few SDoH-related common data elements available in national databases. We aimed to characterize the SDoH metrics currently available in large public datasets and registries. We hypothesized SDoH are rarely and inconsistently documented in sources often used in PCCM research.
Methods: A list of national, US-based databases and registries was generated through literature review and expert consultation with PCCM researchers. We extracted SDoH metrics and reported the types of variables and rates of inclusion within the studied databases.
Results: Twenty data sources were reviewed including national all-payer hospitalization databases (e.g. HCUP), insurance-based registries (e.g. Truven Marketscan), and national patient registries (e.g. ELSO). The most commonly reported SDoH metrics were patient’s sex/gender (95%), race/ethnicity (85%), and health insurance status (75%). For socioeconomic status, 2 (10%) data sources included family income, while 5 (25%) reported county or zip-code level income quartiles. Four (20%) data sources reported on education levels (patient, parent, or county level) and 4 (20%) reported on employment status (parent or county level). While sex/gender and race/ethnicity were frequently documented, there were inconsistencies in how they were reported. This included a variety of classification schemata, if/how multiracial individuals were reported, and a lack of distinction between race/ethnicity or sex/gender. No data sources reported variables related to environmental exposures (e.g. pollution or community violence), immigrant/refugee status, religion, or sexual orientation.
Conclusions: SDoH were inconsistently documented in databases and registries often used by PCCM researchers. Gender and race were commonly but inconsistently reported while all other SDoH variables were infrequently reported. Identifying and implementing common data elements for SDoH in national databases and registries may improve the ability to document and study these important variables in critically ill children, ultimately leading to prevention and mitigation of health disparities.