Can delivery room management impact the length of hospital stay in premature infants?



To evaluate the impact of initiating early nasal continuous positive airway pressure (ENCPAP) on the length of hospital stay (LOS) for the very low birth weight (VLBW) infants.

Study design:

LOS at the George Washington University Hospital (GW) after the institution of ENCPAP policy was compared to benchmark values using two-tail t-tests. The incidence of neonatal morbidity was calculated using Bonferroni corrected 95% confidence interval as compared to benchmark rates (α=0.001). Comparisons were repeated after stratification of the population into four birth weight subcategories: group A (GrpA) (501 to 750 g), GrpB (751 to 1000 g), GrpC (1001 to 1250 g) and GrpD (1251 to 1500 g).


We studied 228 consecutive VLBW infants (birth weight: 995±294 g and gestational age: 27.7±2.7 weeks). Compared to benchmark values, the GW experience was associated with a significant reduction of 5.1 days in LOS (55.9±25.2 vs 61±32 days; P=0.04). The decrease in LOS was consistent in all subgroups, but was most noticeable in infants of the smallest weight subcategory (LOS in GrpA=86±21 vs 104±32, P=0.004; in GrpB=69.9±16.7 vs 79±27, P=0.018; in GrpC=48.2±13 vs 56±22, P<0.001 and in GrpD=31.7±12.5 vs 40±19, P=0.003).

In the overall population, a lower incidence of chronic lung disease (CLD) (17.8 vs 29%, P<0.001) was also noted. There were no differences in mortality rates (9 vs 14%), or the incidence of necrotizing enterocolitis (NEC) (8 vs 6%) or intraventricular hemorrhage (6.2 vs 9%) between GW and the established benchmark rates.


ENCPAP may reduce LOS in VLBW infants in our study population. This relatively shorter LOS was associated with a lower incidence of CLD, which may be a contributing factor.


Several neonatal units advocated early use of nasal continuous positive airway pressure (ENCPAP) as an alternative strategy to initial endotracheal intubation.1, 2, 3 Close evaluation of short-term success, long-term benefits and more importantly possible risks associated with this alternative strategy have not been explored sufficiently. In a recent observational study, infants supported with ENCPAP required shorter length of hospital stay (LOS) when compared to infants intubated in the delivery room (DR) at the same institution. Infants supported on ENCPAP experienced a significantly shorter LOS after controlling for possible confounders.1 In this communication we report the average LOS for premature infants managed in a single neonatal unit that emphasized the use of ENCPAP as the primary mode of respiratory support in the DR as it compares to known benchmark values. Furthermore, we compare respiratory and nonrespiratory outcomes to available benchmark rates in order to evaluate any underlying mechanisms that could influence LOS in these premature infants.


This study included all very low birth weight (VLBW) infants (birth weight <1500 g) admitted to the George Washington University Hospital (GW) after the institution of ENCPAP as the primary mode of respiratory support in the DR. Two-tailed t-test was used to compare LOS with benchmark values. Analysis was repeated after stratifying the population by birth weight into four subcategories. Group A (GrpA) (501 to 750 g), GrpB (751 to 1000 g), GrpC (1001 to 1250 g) and GrpD (1251 to 1500 g). Bonferroni corrected 95% confidence interval (CI) was used to plot different individual outcome variables at GW in relation to benchmark values. A difference was considered significant if the CI for the variable of interest did not cross the benchmark value. Given the retrospective nature of this study, we opted to use the most conservative test ‘Bonferroni test’ to compare these categorical variables because of its high accuracy level (significance is detected only if α=0.001). Outcome variables followed were chronic lung disease (CLD), intraventricular hemorrhage (IVH) grades 3 to 4, retinopathy of prematurity (ROP) stages 3 to 4, periventricular leukomalacia (PLV) and necrotizing enterocolitis (NEC). Clinical definitions and benchmark values for these variables were derived from the Vermont Oxford network (VON).4


In total, 228 consecutive VLBW infants were included in the study. Mean birth weight was 995±294 g and gestational age was 27.7±2.7 weeks. Forty-four percent of infants were males and 65% were black. The demographic data of the two populations did not differ except for a higher prevalence of black infants in the GW population (65 vs 24%).

ENCPAP was applied to 161 (71%) of VLBW infants, 130 infants received CPAP in the DR and 31 infants within 24 h from admission to the neonatal unit. Of all infants, 73.1% were discharged home, 9% died in the hospital and 17.9% required transfer for surgical procedures. These rates did not differ from benchmark values (Figure 1).

Figure 1

Survival, transfers and death rates of the study population. Grp A, infants with birth weight=501–750 g; Grp B, infants with birth weight=751–1000 g; Grp C, infants with birth weight=1001–1250 g and Grp D, infants with birth weight=1251–1500 g.

LOS at GW was significantly shorter than reported in the benchmark population (55.9±25.2 vs 61±32 days, P=0.04) with an estimated mean difference of 5.1 days. In the subgroup analyses, shorter LOS was most noticeable in the smallest subgroup of infants (a difference of 18 days in GrpA). When compared to benchmark values LOS at GW was significantly shorter for all subgroups: GrpA (86±21 vs 104±32, P=0.004), GrpB (69.9±16.7 vs 79±27, P=0.018), GrpC (48.2±13 vs 56±22, P<0.001) and GrpD (31.7±12.5 vs 40±19, P=0.003) (Figure 2).

Figure 2

Length of stay in the hospital of the study population. *P<0.05; **P<0.01; ***P<0.001; LOS, length of hospital stay; Grp A, infants with birth weight=501–750 g; Grp B, infants with birth weight=751–1000 g; Grp C, infants with birth weight=1001–1250 g and Grp D, infants with birth weight=1251–1500 g.

Incidence of CLD was significantly lower than reported in the benchmark population (17.8 vs 29%, P=0.001). Subgroup analyses of the GW cohort demonstrated a significant advantage for infants in Grps A, C and D (P=0.001) but not Grp B (P>0.001). The incidence of IVH and NEC did not differ in the two populations. Incidence of ROP and PVL at GW were significantly lower (Table 1) in the overall population when compared to benchmark values but no subgroup comparisons could be conducted due to limitation of sample size.

Table 1 Incidence of neonatal outcomes by birth weight subgroups


We observed an average of 5.1 days decrease in LOS in this single institutional experience. The decrease in LOS was significant in all weight subcategories. The significantly shortened hospitalization may be a reflection of several favorable outcomes related to the protocol of respiratory care, namely institution of ENCPAP. As previously reported, an observed trend to increased average daily weight gain in these neonates1 is probably secondary to decreased work of breathing in infants free of CLD. CLD itself may play a direct role, independent of growth in prolonging hospitalization, suggested mechanisms may include oxygen dependency, secondary infections and inability to nipple independently. The average shortened LOS in our analysis was not attributable to a shortened life span of infants at GW. Mortality rates at GW in the overall population as well as in all weight subgroups were not dissimilar from benchmark.

Several mechanisms can explain the observed improvement in respiratory outcomes at our institution. CPAP prongs provide a significantly lower resistance and therefore, less work of breathing for premature infants when compared to conventional endotracheal tubes. Airway and alveolar inflammation may be minimized with the use of less invasive forms of respiratory support, yet still provides adequate distending alveolar pressure. The continuous distending pressure of CPAP has been shown to enhance growth of the premature lung,5 and is proposed to stent airways. Consequently, collapse of air spaces is prevented, and the functional residual capacity is increased. These factors singularly or in combination may have contributed to improved lung outcomes in infants supported by ENCPAP.

We find it reassuring to report that ENCPAP use is not associated with increased IVH in any of the weight categories, including the smallest group of infants. Suspected risks of permissive hypercarbia in association with ENCPAP are not substantiated by our previous report,6 and the low incidence of IVH further dispels any suspicion in this regard. No decline in the incidence of IVH grades III and IV was observed which is in agreement with previously published data.2 It is likely that other associated risk factors, independent of mode of respiratory support may weigh-in on the incidence of IVH significantly, particularly in the smallest weight categories. Such associated risk factors include metabolic, hemostatic, hemodynamic and homeostatic fluctuations.7, 8, 9 On the other hand, several mechanisms have been suggested for a possible increased incidence of IVH in association with mechanical ventilation. Stress associated with tracheal intubation and suctioning, as well as the relatively increased intra-thoracic pressure and the resulting compromise in cerebral venous return are plausible. It is worthy of note that the incidence of IVH grades III and IV in both the GW and benchmark cohorts is low, especially infants >1000 g. Similarly the data regarding PVL, in the face of the very small numbers, can only be used as a reassurance that ENCPAP does not strikingly present additional risk for this complication.

The observed decreased incidence of ROP at GW may be explained by its policy to maintain O2 saturation of VLBW infants at a range of 87 to 93%. Infants in the first few days of life may be maintained on the higher end of this range; 90 to 93%. We speculate that the relative clinical stability of these nonventilated infants prevents significant fluctuation in arterial partial pressure of O2 and consequentially minimizes the risk for the development of ROP.10, 11 We have anecdotally noticed a trend towards increased NEC in infants managed with ENCPAP. This concern is not supported by the results of the current comparative analysis. The previously described benign gaseous distension associated with ENCPAP presents a challenge that needs to be followed closely over time and examined in larger cohorts.12 Our impression is that the safety of ENCPAP with regard to NEC risk has not been established so far.

This comparative analysis is not without limitation, since the benchmark cohort, although large and well organized, does not represent a patient population completely congruous to our own. We identified two cardinal differences; first is the higher percentage of African American infants in the DC population, a difference that might place GW at a biological advantage. The second possible difference is that the environment of care at GW may have been different in unmeasured ways from the benchmark environment. This aspect is difficult to compare since the benchmark itself represents various clinical sites with an array of practices that would be difficult to summate in one collective score.

We can conclude that in our experience the use of ENCPAP in VLBW infants is associated with shorter hospitalization time, possibly due to decrease lung morbidity. Future prospective trials on ENCPAP should include LOS as an outcome variable. It is not our intention to present this report as an alternative to such trials, but to enrich the literature further on this important modality of respiratory management.


  1. 1

    Aly H, Milner JD, Patel K, El-Mohandes AE . Does the experience with the use of nasal continuous positive airway pressure improve over time in extremely low birth weight infants? Pediatrics 2004; 114: 697–702.

  2. 2

    Lindner W, Vossbeck S, Hummler H, Pohlandt F . Delivery room management of extremely low birth weight infants: spontaneous breathing or intubation? Pediatrics 1999; 103: 961–967.

  3. 3

    Van Marter LJ, Allred EN, Pagano M, Sanocka U, Parad R, Moore M et al. Do clinical markers of barotrauma and oxygen toxicity explain interhospital variation in rates of chronic lung disease? The Neonatology Committee for the Developmental Network. Pediatrics 2000; 105: 1194–1201.

  4. 4

    Vermont Oxford Network, 2001 Database Summary.

  5. 5

    Zhang S, Garbutt V, McBride JT . Strain-induced growth of the immature lung. J Appl Physiol 1996; 81: 1471–1476.

  6. 6

    Nguyen AT, Aly H, Milner J, Patel KM, El-Mohandes A . Partial pressure of carbon dioxide in extremely low birth weight infants supported by nasal prongs continuous positive airway pressure. Pediatrics 2003; 112: e208–e211.

  7. 7

    Pryds O, Christensen NJ, Friis-Hansen B . Increased cerebral blood flow and plasma epinephrine in hypoglycemic preterm neonates. Pediatrics 1990; 85: 172–176.

  8. 8

    Perlman JM, McMenamin JP, Volpe JJ . Fluctuating cerebral blood-flow velocity in respiratory-distress syndrome. Relation to the development of intraventricular hemorrhage. N Eng J Med 1983; 309: 204–209.

  9. 9

    Amato M, Fauchere JC, Hermann Jr U . Coagulation abnormalities in low birth weight infants with peri-intraventricular hemorrhage. Neuropediatrics 1988; 19: 154–157.

  10. 10

    York JR, Landers S, Kirby RS, Arbogast PG, Penn JS . Arterial oxygen fluctuation and retinopathy of prematurity in very-low-birth-weight infants. J Perinatol 2004; 24: 82–87.

  11. 11

    Chow LC, Wright KW, Sola A, CSMC Oxygen Administration Study Group. Can change in clinical practice decrease the incidence of severe retinopathy of prematurity in very low birth weight infants? Pediatrics 2003; 111: 339–345.

  12. 12

    Jaile JC, Levin T, Wung JT, Abramson SJ, Ruzal-Shapiro C, Berdon WE . Benign gaseous distension of the bowel in premature infants treat with nasal continuous airway pressure: a study of contributing factors. Am J Roentgenol 1992; 158: 125–127.

Download references


We would like to recognize the Vermont Oxford network for their effort in compiling and organizing a comprehensive neonatal data set, and for providing us the opportunity of including their outcome data in our comparative analyses.

Author information

Correspondence to H Aly.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Aly, H., Massaro, A. & El-Mohandes, A. Can delivery room management impact the length of hospital stay in premature infants?. J Perinatol 26, 593–596 (2006) doi:10.1038/

Download citation


  • CPAP
  • chronic lung diseases
  • necrotizing enterocolitis
  • VLBW infants

Further reading