Eight RCTs were included in the review (n=806 patients; range: 51 to 184). All but one trial used adequate randomisation. All trials were double blind, five trials described use of a power calculation, but only one trial reported analysing intention-to-treat data.
Adults (five RCTs, n=413 patients): In four RCTs, steroid treatment resulted in a statistically significant earlier reduction of pain (ranging from four to 11.8 hours earlier) and complete pain relief (ranging from 13.8 to 28.2 hours earlier). One trial reported significantly better reduction of pain after 12 and 24 hours.
Children (three RCTs, n=393): All trials showed a statistically significantly earlier reduction in pain, ranging from 5.1 hours to one day. In one trial, dexamethasone for three days was not superior to a single dose for any outcome.
No serious adverse effects were reported, and no significant differences in time off work or school were found.
Subgroup results were also reported.