In a randomized, placebo-controlled trial, the corticosteroid was non-inferior to saline in the risk of bleeding that required inpatient care, according to Christopher Hartnick, MD, of the Massachusetts Eye and Ear Infirmary in Boston, and colleagues.
On the other hand, the researchers could not rule out an effect on minor, self-reported bleeding, Hartnick and colleagues reported in the Sept. 26 issue of the Journal of the American Medical Association.
The use of dexamethasone -- long a standard practice that was regarded as safe -- was challenged in 2009 by Swiss researchers who reported postoperative bleeding in 24% of children given the drug in a clinical trial of tonsillectomy for recurrent tonsillitis or sleep apnea.
That was seven-fold higher than the rate seen among patients who got a placebo, and led to the trial being stopped early.
To look into the issue, Hartnick and colleagues enrolled 314 children, ages 3 through 17, who were scheduled for adenotonsillectomy and randomly assigned them to get 0.5 mg/kg of dexamethasone or the equivalent volume of 0.9% saline.
The study drugs were given intravenously at the start of the procedure. Postoperative bleeding was monitored for 14 days after surgery.
The study was designed as a non-inferiority trial with a non-inferiority margin of 5%. The researchers defined 3 levels of postoperative bleeding: Level I was self-reported or parent-reported; level II required inpatient admission to control bleeding; and level III required re-operation to control bleeding.
All told, 157 children and adolescents were assigned to each group, with 17 dexamethasone patients (or 10.8%) and 13 placebo patients (or 8.2%) reporting bleeding events.
In an intention-to-treat analysis, Hartnick and colleagues reported:
- Level I bleeding was reported by 11 dexamethasone patients (7%) and 7 placebo patients (4.5%). The upper limit of the 97.5% confidence interval was 7.7%, which exceeded the 5% non-inferiority margin.
- Level II bleeding occurred in three patients on dexamethasone (1.9%) and five who got placebo (3.2%). The upper limit of the 97.5% confidence interval was 2.2%, within the non-inferiority margin.
- Level III bleeding occurred in three dexamethasone patients (1.9%) and one placebo patient (0.6%). The upper limit of the 97.5% confidence interval was 3.8%, within the non-inferiority margin.
Results were similar in a per-protocol analysis, Hartnick and colleagues reported.
Four of the bleeding events -- two in each group -- occurred within the first 24 hours; if those events were excluded, the rates of serious bleeding (levels II and III) remained non-inferior between the groups, while non-inferiority could not be established for level I bleeding.
The researchers noted that the subjective nature of the level I bleeds might account for the difference, especially since many of the reported events were sufficiently minor that parents only reported them when prompted by a questionnaire at the follow-up visit.
"The level II and III postoperative bleeding events are a more reliable indicator for complications because they are documented by treating physicians," Hartnick and colleagues argued.