Intrawound vancomycin application after spinal surgery: a propensity score-matched cohort analysis.

OBJECTIVE

Surgical site infection (SSI) following spine surgery is associated with increased morbidity and healthcare costs. In an effort to reduce SSI rates, the application of intrawound vancomycin powder has gained popularity. However, there is limited high-quality evidence to support the safety and efficacy of this practice. The authors sought to determine if intrawound application of vancomycin powder improves 90-day overall SSI rates.

METHODS

The authors performed a retrospective, vancomycin exposure-matched cohort study at a single tertiary care hospital over 21 months. They included all patients undergoing elective spinal surgery and stratified the patients into two groups: those who received intrawound vancomycin powder application and those who received no application of vancomycin powder. The primary outcome of interest was the 90-day overall SSI rate. Secondary outcomes included rates of superficial SSI, deep SSI, wound disruption, and a post hoc analysis of the microbiology and minimum inhibitory concentrations. Baseline patient demographics, clinical presentation, comorbidities, perioperative factors, and 90-day postoperative outcomes were manually abstracted from patient charts. To mitigate bias, we performed 1:1 matching after calculating propensity scores and identified 1 patient from the no-vancomycin cohort for each patient in the vancomycin cohort.

RESULTS

A total of 997 patients met our inclusion criteria (473 patients receiving vancomycin and 524 patients not receiving vancomycin). Propensity score matching produced 221 matched pairs. Risk-adjusted analysis demonstrated similar overall SSI rates between the groups (OR 1.9, p = 0.329). On unadjusted analysis, the overall 90-day SSI rate was greater in the vancomycin group (n = 10 [4.5%]) than in the no-vancomycin group (n = 5 [2.3%]) (p < 0.001), as were the superficial SSI rate (7 [3.2%] vs 4 [1.8%], p < 0.001), deep SSI rate (3 [1.4%] vs 1 [0.5%], p < 0.001), and wound disruption rate (5 [2.3%] vs 1 [0.5%], p < 0.001). No cultured isolate demonstrated vancomycin resistance.

CONCLUSIONS

The authors observed no difference in SSI rates after the intrawound application of vancomycin powder during spine surgery. Vancomycin use did not contribute to antimicrobial resistance; however, it may select out gram-negative bacteria and increase rates of wound disruption.