Management of device-associated bone and joint infections combines surgical treatment and antibiotic therapy. Immediate postoperative broad-spectrum empiric treatment is often recommended pending complete microbiology results. The probabilistic antibiotics often used are not harmless, and expose the patient to the selection of resistant bacteria and changes in bacterial ecology within the healthcare units. The aim of this study was to answer the following questions: (1) Have the pathogenic bacteria isolated in osteoarticular infections and/or their antibiotic susceptibility profile changed over a 9-year period at our center, with possible implications for antibiotic therapy protocols? (2) Is there a way to optimize the incubation time for microbiological cultures while still being acceptable for slow growing pathogens?

Our hypothesis was that no difference in microbial epidemiology and antibiotic resistance rates would be found over this 9-year period, confirming antibiotic protocols.

We conducted a retrospective study describing a nine-year bacterial epidemiology in four traumatology and orthopedic surgery departments. Bacteria identified, site of infection, and antibiotic resistance were evaluated. Antibiotics for use in probabilistic protocols and after documentation of infection were selected for follow-up. Bacterial species were isolated after a maximum incubation period of 15 days up to 2020, and 10 days from 2021 with the addition of a solid culture medium that promotes anaerobic bacterial growth.

Of the 648 clinical situations, 824 bacteria were identified. Gram-positive cocci accounted for 61.4% (506/824) of the strains, including 291/824 (35.3%) Staphylococcus aureus and 136/824 (16.5%) coagulase-negative staphylococci. For the latter, fluoroquinolone resistance remained stable at 7.3% (18/291 (p = 0.086)) and 22.8% (34/136 (p = 0.432)), respectively. Resistance to rifampicin was also stable for Staphylococcus aureus (1.2%,4/291 (p = 0.486)) and for coagulase-negative staphylococci (10.7%, 15/136 (p = 0.596)). Enterobacteriaceae accounted for 148/824 (18.0%) of identifications and were more resistant to piperacillin/tazobactam (15.1%, 24/148) than to cefepime (5.6%,9/148 (p = 0.006)). There was no significant difference between the 15-day and 10-day culture protocols in the proportion of anaerobic bacteria isolated from shoulder (p = 0.721), hip and knee (p = 0.530) prosthesis infections and spinal device-associated infections (p = 0.373). A microbiological diagnosis was made within five days using the 10-day culture protocol in 333/344 (96.8%) cases.

The increase in Cutibacterium acnes isolates is mainly explained by the increased number of patients recruited for shoulder arthroplasty. No change in antibiotic resistance was observed in our retrospective study. This is probably related to the use of some precautions when prescribing antibiotic therapy. For gram-negative bacilli, cefepime is the beta-lactam of choice for probabilistic treatment when combined with an antibiotic that is particularly effective against methicillin-resistant staphylococci. An incubation period limited to 10 days seems to be sufficient for the culture of slow-growing organisms, as epidemiological analysis has not shown any negative effects.

Level IV. Retrospective cohort study.