Friday, September 26, 2014

Intra-articular antibiotics as prophylaxis in total shoulder arthroplasty


Intraoperative intra-articular injection of gentamicin: will it decrease the risk of infection in total shoulder arthroplasty?

These authors conducted a retrospective study of 507 shoulder arthroplasties (433 primary and 71 revisions) performed between 2005 and 2011. All patients received Cefazolin (or Clindamycin or Vancomycin) intravenous prophylaxis. In 343 patients operated after June 2007, 160 mg of gentamicin in 20 mL of saline was injected into the joint just before closure.

These cases were reviewed for evidence of deep postoperative infection within 6 month of surgery as evidenced by increasing pain, elevated erythrocyte sedimentation rate and Creactive protein level, clinical appearance of infection at the time of surgery, possible positive culture, and more than 10 white blood cells per high-power field.

Six shoulders were thought to have developed infection, one with Staph aureus, one with Methacillin resistant Staph aureus and one with Staph epidermidis.  In three cases the cultures were negative. Five of these cases were operated before June 2007 and one after.

Comment: The use of topical antibiotics to reduce the risk of infection in shoulder arthroplasty is receiving increased attention as shown here. This approach makes sense in that high local concentrations can be achieved without the risk of systemic complications, at least initially (we don't know how long injected or sprinkled antibiotics will hang around in the wound before reabsorption). In this study half of the cases showing signs of infection were culture negative, so we can't be sure that bugs were really there. It is also of note that although the demographics of the patients operated before and after June of 2007 were similar, the surgeon and the team caring for the patients later in the series were more experienced than for the patients earlier in the series. For example, it is possible that open wound times became shorter over the duration of the study, a factor that could also lower the risk of infecton. Finally, the culture protocols for this study were not specified. We know that many apparently 'aseptic' revisions are actually culture positive. The organism accounting for the preponderance of positive cultures at revision arthroplasty is Propionibacterium - this organism requires special means for detecting its presence.  Surprisingly, none of these cases were culture positive for this most common arthroplasty infector.