2.2 | Data analysis
Statistical analysis was carried out using SAS version 9.4 software (Cary, NC). Continuous variables were reported as medians and interquartile ranges. Categorical variables were reported as frequencies and percentages. For Table 1 and Table 2, Wilcoxon rank sum test was used to test differences in distribution of continuous variables between the two modes of ventilation. Chi-square test and Fisher’s exact test were used for the comparison of proportions of categorical variables between the two modes. P values of <0.05 were considered statistically significant.
A multilevel linear mixed model was used to test for association between gastric emptying rates and ventilation modes (MD-nCPAP, bCPAP) accounting for emptying period (early, late) and gestational age (25+0 to 27+6 weeks, 28+0 to 34+6 weeks). All possible interactions, as well as main effects, were included. Multivariable linear regression with backward elimination was used to test for association between total gastric emptying rates and ventilation mode accounting for GA. The interaction between ventilation mode and GA group, as well as the main effects were included in the model. Manual backward elimination was used to identify main effects and interactions that were statistically significant. In the case of a significant interaction term, post-hoc paired comparisons were performed, and a Bonferroni adjustment was used to account for multiple comparisons.