Agreement between pre-operative and intra-operative bacteriological samples in 85 chronic peri-prosthetic infections.

BACKGROUND: Whether pre-operative microbiological sampling contributes to the management of chronic peri-prosthetic infection remains controversial. We assessed agreement between the results of pre-operative and intra-operative samples in patients undergoing single-stage prosthesis exchange to treat chronic peri-prosthetic infection.

HYPOTHESIS: Agreement between pre-operative and intra-operative samples exceeds 75% in patients undergoing single-stage exchange of a hip or knee prosthesis to treat chronic peri-prosthetic infection.

MATERIAL AND METHODS: This single-centre retrospective study included 85 single-stage prosthesis exchange procedures in 82 patients with chronic peri-prosthetic infection at the hip or knee. Agreement between pre-operative and intra-operative sample results was evaluated. Changes to the initial antibiotic regimen made based on the intra-operative sample results were recorded.

RESULTS: Of 149 pre-operative samples, 109 yielded positive cultures, in 75/85 cases. Of 452 intra-operative samples, 354 yielded positive cultures, in 85/85 cases. Agreement was complete in 54 (63%) cases and partial in 9 (11%) cases; there was no agreement in the remaining 22 (26%) cases. The complete agreement rate was significantly lower than 75% (P=0.01). The initial antibiotic regimen was inadequate in a single case.

DISCUSSION: Pre-operative sampling may contribute to the diagnosis of peri-prosthetic infection but is neither necessary nor sufficient to confirm the diagnosis and identify the causative agent. The spectrum of the initial antibiotic regimen cannot be safely narrowed based on the pre-operative sample results. We suggest the routine prescription of a probabilistic broad-spectrum antibiotic regimen immediately after the prosthesis exchange, even when a pathogen was identified before surgery.

LEVEL OF EVIDENCE: IV, retrospective study.