Using Neonatal Intensive Care Units More Wisely for At-Risk Newborns and Their Families | Critical Care Medicine | JAMA Network Open | JAMA Network
[Skip to Navigation]
Invited Commentary
June 18, 2020

Using Neonatal Intensive Care Units More Wisely for At-Risk Newborns and Their Families

Author Affiliations
  • 1Department of Pediatrics, Harvard Medical School, Boston, Massachusetts
  • 2Department of Neonatology, Beth Israel Deaconess Medical Center, Boston, Massachusetts
JAMA Netw Open. 2020;3(6):e205693. doi:10.1001/jamanetworkopen.2020.5693

Escalating US health care expenditures, including estimates that 20% to almost 50% of these costs involve processes, products, and services that do not improve outcomes, have brought renewed attention to the need to improve value in health care.1 Among the 6 waste categories outlined by Berwick and Hackbarth,1 there has been considerable focus on opportunities to reduce overtreatment, “the waste that comes from subjecting patients to care that… cannot possibly help them… rooted in outmoded habits, supply-driven behaviors, and ignoring science.”

Neonatal intensive care unit (NICU) services are at particularly high risk of overuse. Hospital and professional services reimbursements, reflecting the acute and highly technical nature of intensive care, are favorable and remain closely linked to admission volume and patient days in most regions. Both a legacy of intervention and a fear of litigation in caring for an at-risk population can also contribute to ineffective testing and treatments. The neonatology community is, however, starting to recognize the potential for improving care and controlling resource utilization. A 2015 study describing a systematic process to identify ineffective or harmful neonatal tests and treatments yielded a “Choosing Wisely Top Five” list in part to guide these efforts.2 In recent years, the neonatal care value literature has evolved to also focus more broadly on trends relating to NICU utilization—specifically, increasing admission rates and longer lengths of stay.

In the study by Braun et al,3 investigators from Kaiser Permanente share a population-based study describing a decline in NICU utilization—both admission rates and patient days—during a 9-year period. This is an important study, as it describes a trend that is counter to several recent reports of unexplained increasing NICU utilization, particularly for more mature and higher birth-weight infants, using a clinical rather than administrative data set. It is also important because the results may have been associated with several intercurrent performance improvement initiatives. Kaiser Permanente is a large, integrated health care system with a diverse population and a population-based financial payment structure and is in many ways uniquely suited (and motivated) to undertake a project to identify and characterize potential approaches to safely reduce neonatal care that is costly, may be ineffective, separates families, and is potentially harmful. The authors used a risk-adjustment model to ensure that the improvements were associated with postnatal care practices and not with changes in case mix reflecting patients less in need of acute neonatal care. They were also careful to include balancing measures, such as readmission and mortality, among the outcomes. Also important is the residual practice variation, which may hint at future opportunities for reduction in NICU utilization.

In the study by Braun et al,3 12% of more than 300 000 liveborn infants were admitted to the NICU. Contrary to public perceptions of NICUs as prematurity colonies, more than two-thirds of these admissions were infants born after 34 weeks gestational age with birth weights more than 2000 g. The risk-adjusted NICU admission rate, accounting for socioeconomic, prenatal, and delivery room variables to control for independent factors that might affect admission or length of stay, decreased 25% over the study period to 10.9% of births, with 92% of the decline represented by infants with greater gestational age and higher birth weights. Importantly, these changes occurred without evidence of higher 30-day readmission or mortality rates.

There are compelling reasons that these results might not have been a random occurrence, as the health care system’s clinical leadership had implemented several concurrent performance improvement initiatives associated with decreased NICU admissions. A revised policy raised the threshold for NICU admission by lowering the gestational age (<35 weeks) and birth weight (<2000 g) for which well-appearing preterm infants were routinely admitted. A decision support tool based on individual infant estimates of early onset sepsis risk was introduced to guide laboratory testing and empirical antibiotic treatment. Finally, obstetric policies to decrease the rate of nonmedically indicated deliveries before 39 weeks of gestation and to reduce nonmedically indicated nulliparous, term, singleton, and vertex cesarean births were introduced.

The findings by Braun et al3 stand in contrast to a national trend documented in a 2015 population-based study.4 In that study using a public data set, birth-weight–specific NICU admission rates of US neonates were examined over a 6-year period (2007-2012). During this time, despite adjustment for maternal and neonatal characteristics, NICU admissions increased by 23%. These increases were generally represented by larger and less premature infants, such that by the end of the study period, most NICU admissions were for infants with birth weight more than 2500 g.

Although not population-based, observations by NICU member collaboratives, such as the California Perinatal Quality Care Collaborative and the Vermont-Oxford Network, have documented substantial variations in NICU admission and length-of-stay profiles. One California Perinatal Quality Care Collaborative study from 20185 observed that 79% of NICU admissions in 2015 were among infants born at or after 34 weeks gestation, while 10% of infants with 34 or more weeks gestation were admitted to the NICU. Schulman et al5 documented a 40-fold variation among member hospitals in the proportion of NICU admissions meeting high acuity definitions. In a Vermont-Oxford Network6 study involving approximately 500 000 infants hospitalized for nearly 10 million days in 381 NICUs from 2014 to 2016, 74% of NICU admissions were infants at 34 or more weeks gestation and only 15% of admissions met high acuity criteria. The proportion of admissions, patient days, high acuity, and short stays varied significantly both within and between different NICU types.6

The origins of NICUs go back a half century, and NICUs have contributed substantially to reductions in US infant mortality during this time, a period during which rates of prematurity and low birth weight have actually increased.7 In 1967, the infant mortality rate was 22.4 per 1000 live births.8 Fifty years later, in 2017, the rate had declined to 5.8 per 1000 live births,7 a remarkable 74% reduction. Neonatal intensive care is highly effective and has achieved these outcomes and corresponding reductions in morbidity by mitigating the effects of prematurity, congenital anomalies, and pregnancy and perinatal complications. In the early days, NICUs were in short supply and public health entities mobilized to develop regionalized perinatal systems to ensure that obstetric and neonatal patients at high risk had access to specialized services when indicated. As the neonatology workforce and NICU bed capacity increased, hospitals and hospital systems, seeking to become full-service systems, contributed to deregionalization, and there was increasing reliance on economic forces to regulate growth and distribution. In some areas, infants at high risk were distributed more broadly, including to smaller, lower-level units, resulting in less favorable outcomes. Because NICUs are high-margin services, there are significant pressures to expand capacity and maintain volume. This can lead to overuse, including more frequent admission of infants at low risk or a failure to focus sufficiently on care practices that could potentially reduce demand.

There is a cost to these practices. Although NICUs are effective, they are also expensive. Health care system costs are largely borne by government and business, and unwarranted increases may potentially compromise funding of other essential services. Importantly, there may be hidden financial harms for families as well, including costs associated with transportation or lost work days. There are also risks. Short NICU stays by infants at low risk may interfere with breastfeeding, expose them to infection, or increase antibiotic exposure. Additionally, family-infant separation may contribute to emotional risk.

It is not clear that increases in short term, low acuity, and high gestational age and birth-weight NICU admissions have benefited these infants and their families. In fact, the study by Braun et al3 suggests that it may be possible to reverse these trends without compromising and even potentially enhancing care. Rigorous adoption of evidence-based clinical practices, such as use of early onset sepsis decision support and obstetric policies to reduce nonmedically indicated early deliveries and low-risk cesarean delivery rates is a start. There is also a need to examine the opportunities demonstrated by the enormous variation in NICU utilization and in specific NICU practices. These include gestational age thresholds for NICU admission; preferred sites of clinical evaluation, intravenous placement, and antibiotic administration for well-appearing infants with sepsis risk; preferred sites for monitoring and treatment and guidance for length of treatment for opiate withdrawal; and duration of apnea monitoring of preterm infants nearing discharge.

Neonatal intensive care is one of the major achievements of the last half century, and it has resulted in substantial reductions in mortality and long-term morbidity that benefit infants at high risk, including those born to mothers at substantial social risk. If the neonatology community is to successfully achieve the Triple Aim goal for neonatal intensive care—improved neonatal health, better family experience, and reduced cost—we must intensify efforts to learn how to use NICUs more wisely.

Back to top
Article Information

Published: June 18, 2020. doi:10.1001/jamanetworkopen.2020.5693

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2020 Pursley DM et al. JAMA Network Open.

Corresponding Author: DeWayne M. Pursley, MD, MPH, Department of Neonatology, Beth Israel Deaconess Medical Center, 330 Brookline Ave, Boston, MA 02215 (

Conflict of Interest Disclosures: None reported.

Berwick  DM, Hackbarth  AD.  Eliminating waste in US health care.   JAMA. 2012;307(14):1513-1516. doi:10.1001/jama.2012.362PubMedGoogle ScholarCrossref
Ho  T, Dukhovny  D, Zupancic  JA, Goldmann  DA, Horbar  JD, Pursley  DM.  Choosing wisely in newborn medicine: five opportunities to increase value.   Pediatrics. 2015;136(2):e482-e489. doi:10.1542/peds.2015-0737PubMedGoogle ScholarCrossref
Braun  D, Braun  E, Chiu  V,  et al.  Trends in neonatal intensive care unit utilization in a large integrated health care system.   JAMA Netw Open. 2020;3(6):e205239. doi:10.1001/jamanetworkopen.2020.5239Google Scholar
Harrison  W, Goodman  D.  Epidemiologic trends in neonatal intensive care, 2007-2012.   JAMA Pediatr. 2015;169(9):855-862. doi:10.1001/jamapediatrics.2015.1305PubMedGoogle ScholarCrossref
Schulman  J, Braun  D, Lee  HC,  et al.  Association between neonatal intensive care unit admission rates and illness acuity.   JAMA Pediatr. 2018;172(1):17-23. doi:10.1001/jamapediatrics.2017.3913PubMedGoogle ScholarCrossref
Edwards  EM, Horbar  JD.  Variation in use by NICU types in the United States.   Pediatrics. 2018;142(5):e20180457. doi:10.1542/peds.2018-0457PubMedGoogle Scholar
Ely  DM, Driscoll  AK.  Infant mortality in the United States, 2017: data from the period linked birth/infant death file.   Natl Vital Stat Rep. 2019;68(10):1-15. Accessed May 13, 2020. Scholar
March of Dimes. Peristats. Accessed May 13, 2020.
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words