Comparison of Two Respiratory Support Strategies for Stabilization of Very Preterm Infants at Birth: A Matched-Pairs Analysis

Abstract

Objective: Respiratory support for stabilizing very preterm infants at birth varies between centers. We retrospectively compared two strategies that involved either increasing continuous positive airway pressures (CPAP), or increasing oxygen supplementation. Methods: Matched-pairs of infants (<28 weeks of gestation) were born either at the Leiden University Medical Center [low-pressure: CPAP 5-8 cmH₂O and/or positive pressure ventilation (PPV) and fraction of inspired oxygen (FiO₂) 0.3-1.0; n = 27], or at the University Hospital of Cologne (high-pressure: CPAP 12-35 cmH₂O, no PPV and FiO₂ 0.3-0.4; n = 27). Respiratory support was initiated non-invasively via facemask at both units. Infants (n = 54) were matched between centers for gestational age and birth weight, to compare physiological and short-term clinical outcomes. Results: In the low-pressure group, 20/27 (74%) infants received 1-2 sustained inflations (20, 25 cm H₂O) and 22/27 (81%) received PPV (1:19-3:01 min) using pressures of 25-27 cm H₂O. Within 3 min of birth [median (IQR)], mean airway pressures [12 (6-15) vs. 19 (16-23) cmH₂O, p < 0.001] and FiO₂ [0.30 (0.28-0.31) vs. 0.22 (0.21-0.30), p < 0.001] were different in low- vs. high-pressure groups, respectively. SpO₂ and heart rates were similar. After 3 min, higher FiO₂ levels [0.62 (0.35-0.98) vs. 0.28 (0.22-0.38), p = 0.005] produced higher SpO₂ levels [77 (50-92) vs. 53 (42-69)%, p < 0.001] in the low-pressure group, but SpO₂/FiO₂ and heart rates were similar. While intubation rates during admission were significantly different (70 vs. 30%, p = 0.013), pneumothorax rates (4 vs. 19%, p = 0.125) and the occurrence of spontaneous intestinal perforations (0 vs. 15%, p = 0.125) were similar between groups. Conclusion: Infants (<28 weeks) can be supported non-invasively at birth with either higher or lower pressures and while higher-pressure support may require less oxygen, it does not eliminate the need for oxygen supplementation. Future studies need to examine the effect of high pressures and pressure titration in the delivery room.

Keywords: CPAP; birth; breathing; preterm infants; respiratory support.

Figure 1

Study population. The LUMC and the University Hospital of Cologne stored 614 recordings, not including corrupt files. Recorded infants were assessed for eligibility and excluded based on gestational age, congenital abnormalities or unidentifiable video recordings. Infants who could not be matched were excluded as well as the remaining infants from the LUMC.

Figure 2

Mean airway pressure. Data is presented as median (IQR). Mean airway pressure is the average airway pressure between two inspiratory onsets. In the low-pressure group this includes CPAP, sustained inflations and positive pressure ventilation. In the high-pressure group CPAP was exclusively used.

Figure 3

Fraction of inspired oxygen. Data is presented as median (IQR).

Figure 4

Oxygen saturation. Data is presented as median (IQR). For the high-pressure group, at 1.5 and 2 min, the mean of the high-pressure group is calculated based on n = 9 and n = 16, respectively. All other presented data is based on n ≥ 20.

Figure 5

Heart rate. Data is presented as median (IQR). At 1.5 and 2 min, the mean of the high-pressure group is calculated based on n = 9 and n = 16, respectively. All other presented data is based on n ≥ 20.

Figure 6

SF ratio. The SF ratio (SpO₂/FiO₂) represents gas exchange over the lungs and corrects the SpO₂ for the given FiO₂. Data is presented as median (IQR). At 1.5 and 2 min, the mean of the high-pressure group is calculated based on n = 9 and n = 16, respectively. All other presented data is based on n ≥ 20.