The role of postoperative antibiotic duration on surgical site infection after lumbar surgery.

ABSTRACT

OBJECTIVE: Despite a general consensus regarding the administration of preoperative antibiotics, poorly defined comparison groups and underpowered studies prevent clear guidelines for postoperative antibiotics. Utilizing a data set tailored specifically to spine surgery outcomes, in this clinical study the authors aimed to determine whether there is a role for postoperative antibiotics in the prevention of surgical site infection (SSI). METHODS: The Michigan Spine Surgery Improvement Collaborative registry was queried for all lumbar operations performed for degenerative spinal pathologies over a 5-year period from 2014 to 2019. Preoperative prophylactic antibiotics were administered for all surgical procedures. The study population was divided into three cohorts no postoperative antibiotics, postoperative antibiotics ≤ 24 hours, and postoperative antibiotics > 24 hours. This categorization was intended to determine 1) whether postoperative antibiotics are helpful and 2) the appropriate duration of postoperative antibiotics. First, multivariable analysis with generalized estimating equations (GEEs) was used to determine the association between antibiotic duration and all-type SSI with adjusted odds ratios; second, a three-tiered outcome-no SSI, superficial SSI, and deep SSI-was calculated with multivariable multinomial logistical GEE analysis. RESULTS: Among 37,161 patients, the postoperative antibiotics > 24 hours cohort had more men with older average age, greater body mass index, and greater comorbidity burden. The postoperative antibiotics > 24 hours cohort had a 3% rate of SSI, which was significantly higher than the 2% rate of SSI of the other two cohorts (p = 0.004). On multivariable GEE analysis, neither postoperative antibiotics > 24 hours nor postoperative antibiotics ≤ 24 hours, as compared with no postoperative antibiotics, was associated with a lower rate of all-type postoperative SSIs. On multivariable multinomial logistical GEE analysis, neither postoperative antibiotics ≤ 24 hours nor postoperative antibiotics > 24 hours was associated with rate of superficial SSI, as compared with no antibiotic use at all. The odds of deep SSI decreased by 45% with postoperative antibiotics ≤ 24 hours (p = 0.002) and by 40% with postoperative antibiotics > 24 hours (p = 0.008). CONCLUSIONS: Although the incidence of all-type SSI was highest in the antibiotics > 24 hours cohort, which also had the highest proportions of risk factors, duration of antibiotics failed to predict all-type SSI. On multinomial subanalysis, administration of postoperative antibiotics for both ≤ 24 hours and > 24 hours was associated with decreased risk of only deep SSI but not superficial SSI. Spine surgeons can safely consider antibiotics for 24 hours, which is equally as effective as long-term administration for prophylaxis against deep SSI.