Who Needs the NICU? Trends and Opportunities for Improvement Free

The American Academy of Pediatrics, the Academy of Breastfeeding Medicine, and the American Congress of Obstetrics and Gynecology all recommend rooming in for newborns to promote breastfeeding and general family wellbeing.^1–3  Although admission to the NICU is often necessary and even life-saving for many newborns with complex medical issues, there is growing concern that NICU care is overutilized, increasing costs and potential iatrogenicity and undermining many goals of the birth hospitalization.^4–9  Yet many recent publications have revealed a stable or increasing rate of NICU admissions without a concurrent increase in high-acuity or very preterm newborns requiring this level of care.^10,11  There remains a need to study recent trends in NICU admissions to better understand opportunities for improving value and health outcomes in this fragile population.

In their article entitled, “Evaluating Epidemiologic Trends and Variations in NICU Admissions in California, 2008–2018,” Pang and colleagues conducted a retrospective cohort study to evaluate trends in NICU admissions in California from 2008 to 2018.¹²  In their cohort of nearly 4 million live births in the state, they identified an overall stable rate of NICU admissions (mean 11.9% of all live births), with a stable percentage (4.1%) considered high acuity (defined as meeting 1 or more of the following criteria: 4 or more hours of intubated or non-intubated assisted ventilation, early-onset sepsis, major surgery requiring anesthesia, requiring acute transport to a NICU, suspected encephalopathy or perinatal asphyxia, acute therapeutic hypothermia, or death). However, among infants weighing ≥2500 g at birth or born at 37 weeks of gestation or greater, the proportion of high-acuity newborns increased from 1.8% to 2.1% and 1.5% to 1.8% of live births, respectively. Rates of high-acuity newborns <2500 g and/or <37 weeks remained stable. The authors also found significant differences when comparing newborns born within and outside of the Kaiser Permanente health care system; although the proportion of high-acuity births remained stable at non-Kaiser hospitals, NICU admission rates increased from 11.3% to 12.3%. Meanwhile, although the rate of high-acuity admissions increased significantly (from 3.1% to 4.4%) within the Kaiser network, the network’s overall NICU admission rate decreased from 13.9% to 10.1% over the study period.

Although this study highlights many important trends and opportunities for improvement, there are some key limitations. The authors analyzed overall trends in high acuity births at both Kaiser and non-Kaiser hospitals, but they did not stratify data on preterm births by these 2 hospital types. As such, it is unclear whether changes in admission patterns of late preterm newborns (born between 34 and 36 6/7 weeks gestation), a substantial proportion of NICU admissions,^4,13,14  were responsible for the notable difference in admission trends between Kaiser and non-Kaiser hospitals. Understanding trends in this late preterm subset may help explain the broader trends observed in this study because previous work has revealed great variation in admission practices by gestational age,^4,15,16  and even this study notes a change in admission practices at Kaiser hospitals during the study period. A recent analysis from the Vermont Oxford Network also reveals a significant shift over the last decade in place of birth and NICU admissions of premature infants <30 weeks of gestation; volume decreased at high-volume/high-acuity centers but increased among low-acuity/low-volume centers (ie, community NICUs).¹⁷  Given that >70% of units included in this study were low-volume/low-acuity NICUs, a similar shift may have occurred here and would help explain some of the trends observed. Finally, the authors did not study the long-term outcomes of newborns in this cohort; although the underlying assumption of many conclusions made here is that fewer NICU admissions represents improvement, long-term neurodevelopmental and other health outcomes may represent an important balancing measure to ensure the provision of high-value care in the long term.

Pang and colleagues’ study offers many opportunities for future efforts in both implementation and research. Perhaps one of the most notable findings is the Kaiser hospital network’s ability to decrease NICU admissions despite an increase in high-acuity admissions, which suggests a successful reduction in nonacute NICU admissions. Assuming equal or superior long-term outcomes of newborns falling into the latter category, the pediatric community stands to learn many valuable lessons from the Kaiser Permanente network. Possible drivers of this improvement include the establishment of a large health care network with the ability to compile and study large datasets, draw conclusions, and develop and implement clinical practice on the basis of these observations; one famous example is the Kaiser Early Onset Sepsis calculator.^18–20  As the authors point out, the role of Kaiser as both a payor and provider is also meaningful; hospitals like Kaiser may be uniquely incentivized to safely decrease health care costs through evidence-based quality improvement initiatives. The implementation of similar models in other parts of the country may substantially improve our ability to deliver patient-centered, high-value care.

Another important finding was the increase in high-acuity newborns of ≥2500 g or 37 weeks of gestation at both Kaiser and non-Kaiser hospitals. Whether this represents a change in epidemiology (ie, an increased incidence of higher acuity newborns), enhanced recognition of newborns requiring ICU level of care, or changing admission patterns of high-risk obstetrical patients remains unclear. Future investigations exploring this increase and possible explanations for this trend could help to highlight additional opportunities for enhanced obstetrical and/or pediatric care to further decrease NICU admissions and improve outcomes. Also, as previously mentioned, long-term outcomes on newborns previously requiring NICU admission but now managed in lower acuity settings (eg, late preterm newborns, newborns receiving dextrose gel instead of intravenous infusions, and newborns managed using the eat, sleep, console approach) is needed to ensure we are providing not only short-term value to newborns and their families but also the best long-term outcomes and quality of life. Finally, Pang and colleagues’ work should inspire investigators in other regions and health care networks to conduct similar work, to establish whether the trends observed here are consistent across the country.

Pang and colleagues’ ambitious effort to observe and analyze NICU admission trends of such a large cohort represents promising work, both in establishing a need for ongoing improvement and investigation to ensure stewardship of NICU admissions, and for offering a model for high-value neonatal care in the trends observed within the Kaiser network. Clinicians and investigators alike may find in this work inspiration and opportunities to continually reappraise and analyze drivers of NICU utilization and prompt future efforts to improve the newborn family experience and value in neonatal care.