Our primary analysis was a Cox regression
stratified according to the severity of the open
fracture2 and study center, with reoperation as
the end point in the time-to-event analysis. The
Cox regression to investigate the effect of irrigation
pressure was also stratified according to
irrigation solution. Similarly, the irrigation-solution
analysis was stratified according to irrigation
pressure. We also performed analyses, using Cox
regression, that were adjusted for age, injury
(upper extremity vs. lower extremity), fracture
gap (