Good results in postoperative and hematogenous deep infections of 89 stable total hip and knee replacements with retention of prosthesis and local antibiotics

Abstract

Background Deep postoperative and hematogenous prosthesis infections may be treated with retention of the prosthesis, if the prosthesis is stable. How long the infection may be present to preclude a good result is unclear.

Patients and methods We retrospectively studied 89 deep-infected stable prostheses from 69 total hip replacements and 20 total knee replacements. There were 83 early or delayed postoperative infections and 6 hematogenous. In the postoperative infections, treatment had started 12 days to 2 years after implantation. In the hematogenous infections, symptoms had been present for 6 to 9 days. The patients had been treated with debridement, prosthesis retention, systemic antibiotics, and local antibiotics: gentamicin-PMMA beads or gentamicin collagen fleeces. The minimum follow-up time was 1.5 years. We investigated how the result of the treatment had been influenced by the length of the period the infection was present, and by other variables such as host characteristics, infection stage, and type of bacteria.

Results In postoperative infections, the risk of failure increased with a longer postoperative interval: from 0.2 (95% CI: 0.1–0.3) if the treatment had started ≥ 4 weeks postoperatively to 0.5 (CI: 0.2–0.8) if it had started at ≥ 8 weeks. The relative risk for success was 0.6 (CI: 0.3–0.95) if the treatment had started ≥ 8 weeks. In the hematogenous group, 5 of 6 infections had been treated successfully.

Interpretation A longer delay before the start of the treatment caused an increased failure rate, but this must be weighed against the advantage of keeping the prosthesis. We consider a failure rate of < 50% to be acceptable, and we therefore advocate keeping the prosthesis for up to 8 weeks postoperatively, and in hematogenous infections with a short duration of symptoms.

JG and GW treated the patients, designed the study and wrote the manuscript. DJ collected the data of the medical records, completed the follow-up, and performed statistical analyses. AK supervised and helped in the statistical analysis. All authors contributed to interpretation of the data and to the revisions of the manuscript.

No competing interests declared.