Cutting it Clos(er): Non-inferior Comparative Study Comparing One- or Two-day Antimicrobial Prophylaxis after Clean Orthopedic Surgery (NOCOTA study)

Reviewed by Rebekah Moehring, MD, MPH, FIDSA, FSHEA; Duke University, Durham, NC

Investigators from Japan reported on a 5-hospital, open label, cluster-randomized non-inferiority trial of 1211 patients which showed no difference in hospital acquired infection (HAI) outcomes from 24 versus 48 hours of antibiotic prophylaxis following clean orthopedic procedures. The five study sites were cluster-randomized to either a <24 hour or 24 to 48-hour prophylaxis protocol with each study site switching to the other arm after 2-4 months until sample size was reached. Twenty-nine (4.6%) HAIs requiring antibiotic therapy occurred in the 24-hour group and 38 HAIs (6.6%) occurred in the 48-hour group with zero loss of subjects within 30 days (risk difference -1.99% (95% CI -5.05 to 1.06%)). The study met the non-inferiority margin of 4% in all 3 planned analyses and showed a numerically lower surgical site infection rate in the short duration arm (14/633 vs. 19/578). However, the most interesting part of this read was in the discussion. Participating surgeons did not think it was ethical to do a study of prophylaxis only until wound closure despite this recommendation from CDC and WHO guidelines due to subgroup assessments and conflicting observational results cited. This probably echoes what many stewards encounter, as intra-op only antibiotics have historically been a hard sell. So, we’re still waiting for definitive RCTs of the even closer cut at intra-operative dosing only – but this is strong citable literature and a good example for pragmatic trial designs of the future.

Reference:

Nagata K, Yamada K, Shinozaki T, et al. Effect of Antimicrobial Prophylaxis Duration on Health Care–Associated Infections After Clean Orthopedic Surgery: A Cluster Randomized Trial. JAMA Netw Open. 2022;5(4):e226095. doi:10.1001/jamanetworkopen.2022.6095https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2790992

We use cookies to help improve your experience
Ok