Page 123 - Physiological based CPAP for preterm infants at birth Tessa Martherus

Feasibility and effect of physiological based CPAP in preterm infants at birth
The primary outcome SpO2 was compared over time using a linear mixed-effect regression model, accounting for the relation between multiple measurements of the same infants. Physiological effects of the CPAP strategies were examined in a per-protocol analysis that excluded infants who were randomized to PB-CPAP but rather received 5-8 cmH2O CPAP. Additional intention-to-treat and a sensitivity analysis were performed to test the robustness of the study outcomes. Fixed effects in the regression models were group, time and the interaction group*time. P-values of the group variable were used to determine the results of the primary outcome while graphical representations have been used to illustrate the effects of randomization groups over time.
Demographic and secondary outcomes were analysed as per-protocol. Area under the curve (AUC) was calculated for outcomes over time wherein missing values were replaced by the mean, previous or following measurement. Numerical outcomes were analysed using a Mann-Whitney U test, whereas categorical outcomes were analysed using a Fisher’s exact test or a Logrank test.
Results
127 eligible infants were born in the LUMC during the study enrolment period from October
2019 until March 2021, with the study being halted from March 2020 until May 2020 due to COVID-19 restrictions. 112 infants were not included as they met exclusion criteria (n=23),
parents declined to participate (n=37), there was insufficient time to ask for study participation (n=34) or there was insufficient time to complete the randomization procedure 5 (n=2) in an emergency setting. Thirty-one infants were randomized. One infant was excluded
from the analysis due to withdrawal of parental consent, leaving thirty infants for inclusion in the intention-to-treat analysis (PB-CPAP n=10, 5-8 cmH2O n=20). As one twin pair was randomized to PB-CPAP but received 5-8 cmH2O in an emergency setting, his pair was excluded from the per-protocol analysis due to protocol violation. Therefore, twenty-eight infants were included in the per-protocol analysis (PB-CPAP n=8, 5-8 cmH2O n=20, Figure 2). In the PB-CPAP group, seven infants were supported with the initial 15 cmH2O CPAP that was gradually decreased to 8 cmH2O, while one infant received the escape strategy of continuous 8 cmH2O CPAP.
P
119
r