Twelve double-blind RCTs (overall number of participants not reported).
Active Crohn's disease (n=6 studies):
Budesonide was less effective than conventional corticosteroids for inducing remission of active Crohn's disease (pooled RD=-8.5%, 95% CI: -16.4, -0.7, p=0.02, n=4 studies and 621 participants), corticosteroid-related adverse events were reduced (RD=-22.4%; 95% CI: -32, -12.8%, P<0.001). No significant heterogeneity was detected chi-squared=1.89, P=0.595. Budesonide was significantly superior to placebo for inducing remission after 8wks of treatment (RD=22.5%; 95% CI: 10.6, 34.3); and significantly superior to mesalazine for inducting remission within 16wks (RD=27.1%; 95% CI: 13.3, -40.8%).
Quiescent Crohn's disease (n=6 studies):
Budesonide was as effective as placebo for preventing relapse in medically induced remission (RD=-0.8%, 95% CI: -9.9, 8.3, P=0.42, n=4 studies and 449 participants). No significant heterogeneity was detected chi-squared=1.01, P=0.798. Similarly budesonide was as effective as placebo for preventing endoscopic recurrence in surgically induced remission (RD=-3.5%, 95% CI: -16.9, 9.8, P=0.30, n=2 studies and 212 participants). The pooled risk difference for preventing clinical recurrence was -3.0 (95% CI: -15.0, 8.8, P=0.33). No significant heterogeneity was detected.
Corticosteroid-related adverse effects (n=6):
In the long-term treatment, budesonide had an occurrence rate of corticosteroid-related adverse effects similar to placebo (RD=5.3%, 95% CI: -3.9, 14.5, P=0.30, n=6 studies and 661 participants). However, there was significant heterogeneity chi-squared=9.96, P=0.076. If two trials with particularly low occurrences of adverse events are removed from the analysis, the remaining trials are homogeneous, chi-squared=2.66, P=0.447 (RD was not stated).