Method

A retrospective monocentric study was conducted at the 40-bed orthopedics department of a tertiary centre. Medical files of patients with proven prosthetic knee or hip infection with PJACs from maximum 6 months prior to the first stage of a two-stage revision admitted between March 2010 and December 2014 were evaluated. A proven PJI was defined as at least two positive preoperative or intraoperative cultures, the presence of purulent synovial fluid or purulence at the implant site or surrounding the prosthesis without other identifiable causes, the presence of acute inflammation upon histopathological examination of the periprosthetic tissue at the time of surgery or the presence of a sinus tract communicating with the prosthesis. In order to identify the causative pathogen(s) per patient, a multidisciplinary team, consisting out of a microbiologist, a septic orthopedic surgeon, two infectious diseases specialists and two clinical pharmacists, assessed the relevance of pathogens cultured in the PJACs and intraoperative deep samples based on the current 2012 IDSA guidelines. Per patient, agreement of PJACs corresponding to the retained causative pathogen(s) was investigated in two ways: 1) on species level and 2) on Gram stain or fungi level.