Study inclusion and characteristics
In the initial literature search (up to 1, April 2022), 323 records were yielded. After removing duplicates, we screened the titles and abstracts of 299 records and excluded 260 that did not meet our eligibility criteria. After evaluating the full text of the remaining 39 records, we included 21 in our meta-analysis (7-9, 13-30). In an additional literature search (performed just before the article’s submission), we identified one additional eligible article (36). Finally, 22 articles were included in our study (Fig. 1).
We regarded the Yuan et al. study as two trials because it reported separate analyses of two different age groups (14). Therefore, we included 23 trials (15 RCTs and eight cohort studies) consisting of 2,924 participants, of whom 1,451 received NHFOV for respiratory support. Eleven trials compared NHFOV and NIPPV for initial respiratory support of neonates (7, 13-15, 18, 23-25, 27, 29), and 12 trials compared NHFOV and NIPPV for respiratory support after extubation (8, 9, 16, 17, 1-22, 26, 28, 30, 36); 14 trials concentrated on premature infants with RDS (7, 8, 13-15, 17-20, 22, 23, 28-30); four concentrated on neonates with respiratory failure; 1 focused on severe BPD (9) and 1 concentrated on meconium aspiration syndrome (18). The remaining 3 trials did not limit specific disease types (16, 21, 36) (Table 1).