​Impact of Neonatology Over Time

John V. Hartline, MD, FAAP

As noted in the historical trends, infant and neonatal mortality rates have shown steady decline over time due to many factors. (Table 1) The purpose of this short section is to look at what is known about the impact of neonatology per se and what, if any, conclusions may be drawn that are relevant to practice analysis or to career planning.

Regionalization is based in part on data relating access to intensive care to improved outcomes for patients who need intensive care. A second tenet shows regionalization to be associated with better results for infants NOT needing intensive care as well. These results follow introductions of programs regional outreach education, outcomes analysis, and quality improvement. Initially, the number and distribution of NPM specialists allowed only few of the sickest infants to receive NPM care—as there were 10,000 live births per NPM specialist when the first 375 were certified (1975). As the number of NPM specialists increased, more infants could receive specialist care. Today, neonatologists staff virtually all level III units, many of not most level II units (originally the purview of pediatricians), and an increasing number of level I nurseries. In some cases, neonatology has expanded this scope by expanding the clinical responsibilities of neonatologists; in others, neonatology practices have incorporated general pediatricians or advance-practice nurses for this clinical role. ​

Goodman and Associates (2002) analyzed the impact of increasing numbers of clinical neonatologists on the adjusted Odds Ratio (OR) for neonatal mortality. Figure 1. Using 2.7 per 10,000 live births for reference [1 neonatologist per 3700 live births], a significant reduction in Adjusted OR was associated with an increase supply to 4.3 per 10,000 LB, or approximately 2300 live births per neonatologist. As specialist numbers increased, no significant impact on NMR was found. Currently, there are about 10 NPM specialists per 10,000 live births. It is estimated that about 70% of the total work effort of NPM specialists is clinical (see "areas of practice responsibility"), placing the current availability of clinicians ( about 6 to 7 per 10,000 LB) well above the threshold associated with positive impact on mortality.

A second component of effectiveness, after training, is experience. Is there a threshold of patient numbers above which care in a neonatal intensive care unit is likely to be associated with better outcomes, and for whom? First, there are data, again from Goodman, on the impact of increasing supply of NICU beds on Adjusted OR for neonatal death.

About 2% of the birth population is estimated to need Level III infant intensive care. If the estimated length of stay were 2 weeks, perhaps a high estimate for Level III care, a birth cohort of 10,000 live births per year would result in about 200 Level III admissions, and 2800 patient days, an average Level III census of about 8 infants. Interestingly, the need for the number of beds to assure an available bed 95% of the time can be calculated by:

 [Beds Needed] = ADC + 2 ADC-2 , if ADC is average daily census.

 Applying this formula to the above estimate:

 [Beds Needed] = 8 + 2 (8) -2 = ~14 NICU beds needed to meet the intensive care bed need of a birth cohort of 10,000 live births per year with 95% bed availability. In the United States, most areas exceed this number. As the number of NICU beds per 10,000 live births increased, Goodman and Associates demonstrated no significant impact on Adjusted OR for neonatal mortality, suggesting that more intensive care beds were not needed to improve infant survival in the population studied.

The total number of beds is not the whole story, for it does not factor in the component of experience. Is there a critical mass of patients needed for a greater chance at optimal outcomes? Staiger examined the relationship between the annual volume of very low birth weight infants (less than 1500 grams) among units in the Vermont-Oxford group. Using risk-adjusted mortality, mortality risk increased as volume fell below 50 patients per year. Although there was considerable scatter in the data points, these data suggest a positive relationship may exist between experience (as reflected by volume of patients) and outcomes.

Outcomes for VLBW infants in California were analyzed based on the Level (self-reported) of the nursery of the hospital of birth and average daily census of VLBW infants per year. [Cifuentes & assoc, Pediatrics2002].​

These data also support a positive impact of experience, although that benefit was only statistically significant for infants less than 1250 grams birth weight.

Phibbs and assoc reported on the Odds Ratio for Mortality by Level of hospital and volume of patients less than 1500 grams birth weight. When compared to the reference group (Level IIIB+, greater than 100 vlbw infants/yr—in red), both lower volume and lower levels of care were associated with higher risk. In all levels of care, increased volume is associated with lower mortality risk.​​

Although the jury isn't totally in on the issues of requisite ongoing experience for the individual or volume for the facility and its multidisciplinary team, data from neonatal care are consistent with data from other complex endeavors requiring multidisciplinary care within a system: experience does matter, or "what you don't do, you lose (proficiency)." As maintaining proficiency is a tenet of professionalism; selection of a position and long-term career planning require melding "what do I want to do" with "where do I have to be to be able to do it with optimal effectiveness."

​Interviewing With Skill »​​​​