Ethnic Disparities in Care of Opioid Exposed Newborns in Colorado Birthing Hospitals

Introduction
The prevalence of opioid use disorder among pregnant people has increased 400% in the past decade in the US. Colorado’s incidence of neonatal abstinence syndrome (NAS) was 2.9 per 1,000 live births in 2013, lower than the national average of 6.0 per 1,000 live births, but significantly higher than the 1.5 per 1,000 nationally in 1999.

Objective
To investigate disparities by maternal Hispanic ethnicity in the care of opioid exposed newborns (OENs) in Colorado birthing hospitals within a statewide perinatal quality improvement (QI) collaborative.

Methods
The Colorado Hospitals Substance Exposed Newborn Quality Improvement Collaborative (CHoSEN QIC) was formed in 2017 in response to the rising rate of NAS and the great variability of care within and across hospitals caring for OENs and their families. Hospitals participating in the collaborative reported deidentified data on OENs from Q2 2017 through Q1 2020, including limited maternal sociodemographic information, polysubstance exposure, and use of postnatal opioids to treat neonatal abstinence syndrome (NAS). Infants were categorized as being born to Hispanic or Non-Hispanic mothers based on maternal self-report. We used process control charts to detect special cause variation by ethnicity for outcomes including infant length-of-stay (LOS), use of postnatal opioids, and breastfeeding eligibility and receipt. Only hospitals that delivered both Hispanic and Non-Hispanic OEN during the study period were included in this study.

Results
We analyzed 795 mother-OEN dyads, 241 (30%) were Hispanic and 554 (70%) were Non-Hispanic. Hispanic OENs experienced a 63% decrease in overall LOS and an 83% decrease in LOS for those who received postnatal opioids. The Non-Hispanic population saw 62% and 64% decreases, respectively. The percentage of OEN being treated with postnatal opioids decreased in both populations; 84% for Hispanic OEN and 72% for Non-Hispanic. Neither group experienced a process change related to eligibility to breastfeed or rates of breastfeeding.
Despite both populations meeting the SMART aims of this QI initiative, delays to process change were observed among the Hispanic population. The Hispanic group experienced sustained change in lowering overall LOS and LOS among OEN treated with opioids three quarters after sustained change was achieved in the Non-Hispanic population. A delay of one quarter was seen for change in the percentage of OENs being treated with postnatal opioids.
Conclusions
This statewide QI initiative resulted in positive outcomes for both Hispanic and Non-Hispanic OENs, however the experience of improvement was delayed in the Hispanic population. This indicates a need to explore care practices of Hispanic mothers and OENs, likely with more culturally competent approaches to care, including language inclusivity.