Association Between Social Drivers of Health and Analgosedatives in Ventilated PICU Patients

Introduction: Social drivers of health (SDOH) are associated with patient care and outcomes in the pediatric ICU (PICU), but limited data exist on SODH associations with analgo-sedative drug use in PICU patients on mechanical ventilation (MV). We used an administrative multicenter database to evaluate associations between analgo-sedative drugs and various measures of SDOH.


Methods: With IRB approval, we identified PICU patients < 18yo who received MV in a Pediatric Health Information System (PHIS) hospital between 2016-2024. Demographics, Childhood Opportunity Index (COI), insurance, length of stay, length of MV, ICD diagnosis codes, and medication billing data were obtained. Drug exposure was defined as days of receipt divided by duration of MV. Linear regression models for each drug included age, post-operative status, illness severity, primary diagnosis category, and all SDOH variables; p< 0.05 defined significance.


Results: Of 146,095 subjects at 48 centers, 17% had primary respiratory illness, 57% were male, 59% publicly insured, 44% non-Hispanic white, and 27% with very low COI. In the multivariate models, very low COI subjects were more likely to be prescribed fentanyl, methadone, dilaudid, and ketamine than very high COI subjects, but with less exposure to diazepam and midazolam (all p< 0.001). Non-Hispanic Black patients had lower exposure to morphine, dilaudid, and methadone (all p< 0.001) than their non-Hispanic White counterparts. Hispanic ethnicity was associated with increased exposure to most opioids, propofol, and dexmedetomidine, but decreased exposure to ketamine (all p< 0.001). Public insurance status was associated with lower use of fentanyl, dilaudid, and midazolam, but greater use of methadone, chloral hydrate, and antipsychotic medications, compared to commercial insurance status (all p< 0.05).


Conclusions: In this administrative dataset, analgo-sedative medication prescription was associated with several facets of SDOH after adjusting for clinical variables, including COI, race/ethnicity, and insurer. These differences may reflect regional, institutional, and provider practice variations, though underlying biases in medicine cannot be excluded as contributing factors. Further research is needed to evaluate these associations.