Evaluating Racial Disparities in Care in the NICU Using Digital Tools and Electronic Health Record (EHR) Data


View the Full Poster



There is significant racial variation in quality of care delivery between and within NICUs. Consistent quality of care in the NICU requires the ability to access timely and meaningful metrics to allow for active vigilance and standardization of care, especially as it relates to feeding, nutrition, and growth.  By identifying care disparities using real-world data, clinicians can improve care, reduce preterm infant co-morbidities and provide equitable, consistent care. Digital tools that harness data from the EHR for monitoring and reporting on key care performance metrics and associated clinical outcomes by race and ethnicity are urgently needed. 


This work aimed to utilize a .NET application to extract 5 years of data directly from the EHR using secure Fast Healthcare Interoperability Resources Application Programming Interfaces (FHIR APIs), a standard for exchanging private health information.  JavaScript frameworks were used to interpret and present data in an intuitive, user-friendly display. A proprietary Nutrition IQ framework was applied to incorporate complex feeding protocols into the software application and provided unit-wide tracking of key metrics by displaying a side-by-side comparison of nutrition metrics, feeding milestones, growth measures, and discharge statistics by race.


Inclusion criteria of admission before 3 days and <33 weeks estimated gestational age (EGA) identified 234 infants who were placed in 2 cohorts. Cohort 1 consisted of 110 infants EGA <30 weeks, 52 were white and 42 were black, and cohort 2 consisted of 124 infants EGA 30 0/7 to 32 6/7 weeks, 66 were white and 46 were black.  Feedings during the five-year period were analyzed for protocol adherence and correlated with metrics, milestones, and outcomes.  


Comparing enteral feeding milestones in both cohorts, days to first enteral feed, first fortification, and first full enteral feed were either consistent or faster for the black infants compared to the white infants. Compliance with enteral feeding protocols was also relatively consistent in both cohorts for white and black infants.  The transition to oral feeding showed a statistically and clinically significant decrease in days to first oral attempt and first full oral feed for the black infants in both cohorts.  Comparing growth metrics in the <30 weeks cohort, return to birthweight (RTBW) in days, growth velocity in g/kg/day, and birth to discharge (D/C) z-score delta for weight and head circumference (HC) were either consistent or better for the black infants compared to the white infants.  For the 30 0/7 to 32 6/7 cohort, growth metrics were better for the black infants, statistically significant for days to RTBW and birth to D/C z-score delta for head circumference.  Days on parenteral nutrition (PN) and length of stay in both cohorts were relatively consistent for black infants compared to white infants.  However, the rate of human milk at discharge was significantly less for black infants in both cohorts compared to their white counterparts: 45% vs. 17% for the <30 weeks cohort and 48% vs. 28% for the 30 0/7 to 32 6/7 cohort.


The current ability to assess equitable care provision to racial or ethnic peers at the unit level by gestational age cohorts is costly and time prohibitive. By utilizing the data afforded by the EHR in this case, a greater focus can now be placed on improving adherence to breast feeding at discharge. Implementing software that harnesses the data entered into the EHR allows for real-time monitoring of protocol adherence and patient outcomes, enabling clinical teams to continuously educate staff and better target programs and interventions to improve patient outcomes.



Misty Virmani1, Dave Genetti2, Aamir Nayeem2, Tammi Jantzen2, Ashley Ross1, Laura Carroll1
1Department of Pediatrics, University of Arkansas for Medical Sciences
2Astarte Medical