- Design
- Systematic review and meta-analysis (4 RCTs, 10 cohorts)
- Population
- 9,802 patients undergoing spinal surgery
- Primary outcome
- Surgical site infection
- Effect
- Overall OR 0.55 (0.44–0.70); RCTs only OR 0.73 (0.45–1.18)
This meta-analysis pooled 14 comparative studies of 9,802 spinal surgery patients, 4,516 of whom had vancomycin powder applied to the wound before closure on top of standard intravenous prophylaxis. Four studies were randomised trials and ten were non-randomised cohorts.
Overall, vancomycin powder was associated with lower odds of surgical site infection (OR 0.55, 95% CI 0.44–0.70), including deep (0.50) and superficial (0.55) infection. The effect held in sensitivity analyses. But it came from the non-randomised studies: restricted to the four randomised trials, the estimate was 0.73 (0.45–1.18) and no longer significant.
That pattern is familiar. Surgeons who adopt a new prophylactic step often adopt other infection-control changes at the same time, which flatters the cohort comparison. The randomised evidence is compatible with a modest benefit or none.
- Standard intravenous prophylaxis timed within 60 minutes before incision remains the foundation for spinal surgery.
- Intra-wound vancomycin powder is reasonable to discuss in high-risk spinal cases, but randomised evidence for benefit is not established.
- Consider the effect on local resistance surveillance if vancomycin powder becomes routine in a unit.
- Include glycaemic control and normothermia in any surgical site infection bundle before adding new agents.
Why it matters
A popular practice rests mainly on cohort data, and the randomised trials have not confirmed it.
Don't overread it
The benefit disappears when only randomised trials are counted, so this does not show the powder prevents infection.
The statistics, in plain English
An odds ratio of 0.55 suggests about 45% lower odds of infection, but this is driven by observational studies prone to bias. In the randomised trials the interval runs from 0.45 to 1.18 — crossing 1.0, so no difference cannot be excluded.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for infectious diseases, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free