Letter to the editor on "Topical antibiotic powder reduces fracture-related infections in pelvic and lower limb fractures: a systematic review and meta-analysis": the denominator matters.

Abstract

To clarify the interpretation of the reported increase in Gram-negative organisms after prophylactic topical antibiotic powder in a recent systematic review and meta-analysis. We examined the published report and its supplementary bacterial-profile forest plot, compared the denominators used for microbiological outcomes with the full treatment-cohort denominators, and evaluated the resulting estimand in light of potential index-event bias and contemporary randomized evidence. Supplementary Material S3 used denominators of 56 powder recipients and 181 controls, although the seven contributing studies enrolled 999 and 1703 treated patients, respectively. In one contributing control group, the microbiological denominator also exceeded the number of reported FRIs. Therefore, the reported OR of 2.52 appears to describe the conditional distribution of Gram-negative organisms among selected breakthrough infections or isolates rather than patient-level Gram-negative FRI risk. Under the simplifying assumption that every reported Gram-negative event represented one patient, the corresponding crude proportions would be 2.1% and 2.3%. The largest included randomized trial similarly found no significant increase in Gram-negative-only infection. A patient-level reanalysis using all treated participants as denominators and mutually exclusive microbiological categories is warranted. Until then, the finding should be interpreted as a possible change in the composition of breakthrough infections, not evidence of increased absolute Gram-negative risk, ecological selection, or resistance.