Septic arthritis is a therapeutic emergency with a high mortality rate (about 11%)(1). Inadequate treatment can cause permanent joint damage. Management of the septic arthritis includes prompt antibiotic treatment as well as joint-decompression and removal of purulent material(2). It is still discussed controversially and there is little evidence which surgical concept is preferable: arthroscopy with lavage and debridement or open arthrotomy with synovectomy(3,4). The aim of the study was to compare efficacy of arthroscopy and arthrotomy in patients with septic gonarthritis.

We evaluated 70 consecutive patients who underwent arthroscopy or arthrotomy at our clinic, because of a bacterial monarthritis of the knee between 2002 and 2010. Our primary outcome was the early recurrence of infection (> 3 months after surgery), which made a second surgery necessary. We compared patients who suffered reinfection and those who did not, in regard to the surgery type as well as potential confounders like comorbidity (measured by Charlson comorbidity index), age, body mass index (BMI), Gächter's -, Kellgren and Lawrence - and Outerbridge classification, duration of symptoms and inflammatory parameters. Furthermore we evaluated differences of the confounders between the surgery groups.

From the 70 patients 41 were treated arthroscopic and 29 with arthrotomy. In total eight patients (11.4%) had to undergo a second surgery because of early reinfection. The rate was significantly higher in patients treated with arthrotomy (n=6; 20.7%) compared to those treated with arthroscopy (n=2; 4.9%) (p=0.041). Whereas we found no significant influence of potential confounders between the reinfection group and the group where primary eradication was achieved. Patients who underwent arthrotomy were significantly older, had more comorbidities (both p<0.001) and higher grades of osteoarthritis according to Kellgren and Lawrence classification (p=0.023). In order to adjust the study population towards confounders we performed a subgroup analysis on patients of the second and third age percentile. When we repeated our analysis we still found a significant higher reinfection rate in the arthrotomy group (p=0.036). At the same time there were no differences in prevalence of confounders, neither between the two surgery groups, nor between the reinfection and the primary eradication group.

Patients with bacterial monarthritis of the knee who were treated with arthroscopy had a significantly lower reinfection rate than those treated with arthrotomy. As arthroscopy is the less invasive and more sufficient method it should be considered the routine treatment according to our data.