A new study is challenging the idea that all Level III neonatal intensive care units are interchangeable. Although hospitals grouped under the Level III designation are expected to provide advanced care for critically ill newborns, researchers report substantial variation in the babies treated, the services available, and the outcomes recorded across these units. The findings, published in the Journal of Perinatology, suggest that a hospital’s NICU level alone may be too broad a measure to describe the complexity, capabilities, or performance of the care it delivers.
Level III NICUs occupy a crucial position in the neonatal care system. They generally care for infants who need intensive monitoring, respiratory support, surgery, or other highly specialized treatment, while Level IV centers typically provide the most comprehensive services, including complex surgical programs and extracorporeal life support. But the Level III category encompasses a wide range of hospitals, from large regional referral centers to community-based facilities. The new analysis by Goldstein Novick, Lorch, Radack, and colleagues indicates that this shared label can conceal major differences in clinical populations and institutional resources.
The researchers examined heterogeneity—variation between units—in three connected areas: patient characteristics, services, and outcomes. Patient heterogeneity refers to differences in the infants admitted, such as gestational age, birth weight, medical complications, congenital conditions, and the need for advanced respiratory or surgical care. These factors are not merely descriptive. They strongly influence the probability of survival, complications, length of hospitalization, and discharge needs. A unit caring predominantly for extremely premature infants may face a very different clinical risk profile from one treating mostly full-term newborns with short-term respiratory problems, even if both units carry the same formal designation.
The study also focuses attention on the services that make intensive neonatal care possible. A NICU’s capability is determined not only by the number of incubators or ventilators it contains, but also by staffing, specialist availability, diagnostic technology, surgical access, transport systems, and protocols for managing rapidly changing illness. Services such as high-frequency ventilation, continuous renal replacement therapy, therapeutic hypothermia, neonatal surgery, advanced imaging, and dedicated developmental care can differ considerably between hospitals. Differences may also exist in whether specialists are physically present around the clock or available through consultation and transfer arrangements. Such distinctions can affect how long infants remain in a hospital and whether they must be moved to another center when their condition becomes more complex.
Outcome comparisons are especially difficult when hospitals treat different kinds of patients. In clinical research, this problem is known as confounding by case mix. An unadjusted mortality rate, for example, may appear higher at a hospital that accepts the sickest infants from a wide geographic region, even if its care is highly effective. Conversely, a hospital treating lower-risk newborns may report better crude outcomes without necessarily providing more advanced or higher-quality care. For this reason, meaningful comparisons require risk adjustment: statistical methods that account for differences in baseline illness severity and other patient characteristics before examining how outcomes vary across institutions.
The findings raise questions about how neonatal quality is measured and communicated. Hospital rankings and performance reports often rely on broad categories, administrative data, or a small number of outcomes. But a single label may not capture whether a unit has the capacity to provide prolonged mechanical ventilation, perform specialized procedures, stabilize a newborn before transfer, or support families through months of hospitalization. The researchers’ analysis suggests that evaluations should include a more detailed description of what each NICU actually does and which infants it serves. Without that context, families, clinicians, policymakers, and referring hospitals may draw misleading conclusions from apparently simple comparisons.
The study has implications for regionalized neonatal care, a system in which hospitals are organized according to their capabilities and infants are transferred to the most appropriate level of care. Regionalization depends on accurate information. If two Level III units differ sharply in staffing, technology, case mix, or outcomes, referring clinicians need more than the shared designation to determine where a newborn should be treated. The same information may help health systems identify gaps in local capacity, improve transfer networks, and decide where additional specialized services are needed. It could also support more precise planning for neonatal transport, which is particularly important when minutes or hours can influence an unstable infant’s condition.
For families, the results underscore the complexity of choosing or reaching a NICU. Parents may reasonably assume that a Level III designation provides a uniform package of care, but the designation is better understood as a broad category than a guarantee that every hospital offers identical services. The appropriate unit depends on the infant’s gestational age, diagnosis, expected treatment needs, and the local referral network. In many cases, newborns can receive excellent initial stabilization close to home and then be transferred if advanced care is required. The study does not imply that one category of hospital is universally superior; instead, it shows why the match between a baby’s needs and a unit’s capabilities matters.
The research also points toward a more nuanced future for neonatal benchmarking. Better comparisons could combine clinical data with information on staffing, technology, transfer patterns, family support, neurodevelopmental follow-up, and long-term outcomes. Researchers may need to distinguish between the ability to rescue a critically ill infant, the capacity to provide prolonged intensive care, and success in supporting healthy development after discharge. As neonatal medicine advances, classifications created to organize hospitals may need to evolve as well. The central message of the study is clear: Level III is an important designation, but it is not a complete description. Understanding the real variation among neonatal intensive care units may lead to fairer evaluations, smarter resource allocation, and more individualized care for the smallest and most vulnerable patients.
Subject of Research: Variation in patient populations, clinical services, and outcomes among Level III neonatal intensive care units.
Article Title: Level III neonatal intensive care unit heterogeneity in patients, services, and outcomes
Article References: Goldstein Novick, N.P., Lorch, S.A., Radack, J.K. et al. Level III neonatal intensive care unit heterogeneity in patients, services, and outcomes. J Perinatol (2026). https://doi.org/10.1038/s41372-026-02846-z
Image Credits: AI Generated
DOI: https://doi.org/10.1038/s41372-026-02846-z
Keywords: neonatal intensive care, NICU, Level III NICU, premature infants, newborn health, neonatal outcomes, hospital services, risk adjustment, patient heterogeneity, perinatal medicine
Tags: Level III NICU patient outcomesneonatal care quality and outcomesneonatal care service differencesneonatal critical illness managementneonatal intensive care unit classificationneonatal intensive care unit heterogeneityneonatal intensive care unit regional disparitiesNeonatal intensive care unit variabilityneonatal patient demographic variationsneonatal surgical and respiratory supportneonatal unit performance measuresNICU resource and capability assessment
