“…Subgroup analyses revealed possible trends toward different adverse event rates by publication period, maximum dilator size, and pediatric versus adult studies, although none of these differences reached statistical significance (Supplemental Tables 5–7). For instance, the point estimates of frequency of perforation (0.0%; 95% CI, 0%–1.88%), clinically significant chest pain (2.19%; 95% CI, 0%–5.65%), and any chest discomfort (10.2%; 95% CI, 0%–23.2%) were lower for pediatric than for adult studies, although only 2 to 3 studies contributed to each of these pediatric estimates 28, 29, 36 Perforation rate with larger dilators (>17 mm) was 1.35% (95% CI, 0%–8.43%) versus 0.03% (95% CI, 0%–0.226%) for smaller dilators, hospitalization 4.02% (95% CI, 0%–16.6%) versus 0.511% (95% CI, 0%–1.08%), and clinically significant chest pain 7.14% (95% CI, 0%–19.8%) versus 4.46% (95% CI, 1.41%–7.50%). Finally, the meta-analysis confirmed the direction of association between studies published before 2009 with higher perforation rates, estimating a 0.41% (95% CI, 0%–2.75%) rate before 2009 compared with 0.030% (95% CI, 0%–0.225%) in more recent studies.…”