25

presented by regions that represent relatively self-contained markets for neonatal intensive care. In this section, national neonatal intensive care regions (NICRs; n=246) are used, which were initially defined in the late 1990s using Vital Records data8 to reflect travel patterns of mothers of low birth weight newborns, and continue to be valid regional markets for neonatal intensive care.7,9 They are also reasonable markets for Anthem-insured newborns except where there is low Anthem market penetration. For the Texas Medicaid analyses, we defined regions (n=21) using recent Medicaid utilization data. Norway defined their own regions of neonatal care. Similar to the U.S. regions, there is limited border crossing from Norwegian region of birth or maternal residence to hospital of intensive care. Further information on the definition of regions in the U.S. and Norway can be found at http://www. dartmouthatlas.org/. A REPORT OF THE DARTMOUTH ATLAS PROJECT 19 Map 2.2. Admission Rates to Level III/IV NICUs among Very Low Birth Weight Newborns by NICR (2013) Map 2.3. Admission Rates to Level III/IV NICUs among Moderately Low Birth Weight Newborns by NICR (2013) Map 2.4. Admission Rates to Level III/IV NICUs among Normal Birth Weight Newborns by NICR (2013) A Report of the Dartmouth Atlas Project 20 DARTMOUTH ATLAS OF NEONATAL INTENSIVE CARE Is Neonatal Intensive Care Bed Supply Associated with Newborn Medical Needs?7 The regional supply of NICU beds and neonatologists is known to vary widely. In the late 1990s, Dartmouth researchers demonstrated that there was virtually no association between regional supply and a number of measures of perinatal risk.10 In other words, regions with a high proportion of premature newborns, or other factors related to newborn illness, were not the regions with a higher number of NICU beds or neonatologists per newborn. Fifteen years later (2013), the alignment between NICU bed supply in relation to need had not improved. As seen in Figures 2.3-2.5, the supply of NICU beds per newborn was not associated with the regional percent of low birth weight births (< 2,500 grams) (Figure 2.3), maternal education level (Figure 2.4), or the rate of cesarean sections (Figure 2.5). It is very troubling that such an important and expensive health care resource is not found in greater supply in the places where it is most needed by newborns. It raises two important questions: How are the extra beds used? And is greater supply of NICU resources linked to better outcomes for newborns? A REPORT OF THE DARTMOUTH ATLAS PROJECT 21 Figure 2.5. Relationship Between Cesarean Section Rate and NICU Bed Supply by NICR (2013) The figures show the correlation between several newborn risk factors and the regional bed supply among the 208 neonatal intensive care regions with sufficient data to report. There was no relationship between the supply of beds and measures of newborn risk.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25