Hypothesis: Does 8mg of IV dexamethasone increase surgical site infection during elective surgery with incision >5cm during general anesthesia for noncardiac surgery compared to saline placebo?
Methods
Randomized controlled trial
Multicentered & multinational
Triple blinded
Stratified based on trial center and diagnosis of diabetes mellitus
Non inferiority study
Margin of 2%
Inclusion criteria
Admitted overnight for at least one night post op
Total incision length ≥ 5cm
Surgery length ≥ 2hrs
Diabetics with Hgb A1c < 9.0%
Exclusion criteria
Emergency & cardiac surgery
Existing infection
Total incision length < 5cm
Surgery length < 2hrs
Hgb A1c > 9%
Other indication and use of any steroids ≤ 30 days post op
Results
Dexamethasone was noninferior to placebo for the outcome of surgical-site infection (8.1% vs 9.1%; risk difference -0.9 percentage points; 95.6% confidence interval [CI], -2.1 to 0.3; P<0.001 for noninferiority).
Subjects in the dexamethasone had less PONV (42.2% vs. 53.9%; risk ratio, 0.78; 95% CI, 0.75 to 0.82).
Hyperglycemia occurred in 0.6% of subjects in the dexamethasone group and 0.2% in the placebo group.
Strengths
Study design
Randomized controlled trial
Triple-blinded
International multicenter trial
Australia (46), New Zealand (6), China (2), South Africa (1)
Large # of subjects
Clinically relevant question
Appropriate key subgroups (ie. T2DM)
Statistical strengths
Results confirmed via intention-to-treat and per-protocol analysis
Subgroup analyses confirmed noninferiority
Weaknesses
Nonadherence to protocol randomization (6.3-6.6%)
“Pragmatic study”
Telephone interview follow-up
Exclusion of key subgroups
Generalizability
Key Point: Dexamethasone is non-inferior to placebo for the outcome of surgical site infections at 30 days.