Intravenous antibiotic infusion is the current standard for prosthetic joint infections (PJIs) management. antibiotic used for PJIs have a good oral bioavailability, especially rifampicin, suggesting that oral and IV route could be as efficient. Our aim was to compare the outcome of PJIs treated by oral antibiotics to those treated intravenously.
a retrospective survey was done in two reference centers between 2014 and 2016 and included all patients presenting gram-positive PJIs. In these centers, patients suffering from gram-positive PJIs could receive IV or early oral antibiotics regimen (started the day final antibiogram were received). First, we compared these two groups in terms of demographic and infection characteristics. Then, the outcome, judged by the percentage of patient with a relapse or a new infection on the same site during the follow-up, were compared.
Within this period, 87 patients were treated for gram-positive BJIs, 51 (59%) received early oral therapy and 36 were treated intravenously. Median age was 65 [IQR=56-75], 50 (57%) were female. Overall, 18 cases suffered from polymicrobial BJIs containing at least one staphylococci. S. aureus was found in 30 (34%) cases and 7 (8%) strains were resistant to ampicillin. Median time of follow-up was 467 days [IQR=218-729]. Median time of treatment was 60 days [IQR=44-84]. When comparing oral therapy group and IV group, we did not find any difference in terms of comorbidities, infection characteristics, type of surgery and infection severity. In the oral therapy group, oral antibiotic regimen was started before day 7 following surgery for 33 (76%) individuals and before day 10 for 44 (86%). Treatment failures were observed in 9/36 (25%) and 12/51(23.5%) (p=1) in IV and oral therapy group, respectively.
Oral antibiotic treatment seems to be as efficient as intravenous regimen to treat prosthetic gram-positive prosthetic joint infections.
M. Titecat, None
S. Putman, None
N. Blondiaux, None
M. Valette, None
E. Beltrand, None
H. Migaud, None
E. Senneville, None