Description
Objective
Septic knee arthritis is an orthopedic emergency. Ideally, it should be addressed as fast as possible and, when left untreated, it leads to irreversible damages, besides an associated mortality rate of around 10%. The treatment gold standard is the surgical irrigation and debridement, followed by a course of antibiotics. Although open arthrotomy is currently used, some authors advocate that arthroscopy is preferable because it is less invasive and enables direct visualization of the joint. Also, it may improve the immediate range of motion and functional recovery. However, there is still no available literature stating which is the best approach for septic knee arthritis.
Materials
We designed a retrospective study including all patients hospitalized for septic knee arthritis during 2022, with at least 1 year of follow-up. We aimed to compare patients submitted to arthroscopic (@DI) versus open arthrotomy (ODI) debridement and irrigation. The main outcome was functional recovery, reported with the KOOS score and subjective knee value (classified from 0 to 100%). Secondary outcomes were considered for C-reactive protein (CRP) evolution during hospital stay (2 in 2 days), length of stay, post-op fever development, recurrence, and other reported complications. Additional demographic and clinical data were collected, namely age, onset of symptomatology and duration of surgery. Statistical analysis was performed using IBM SPSS® 27.
Results
We included 12 patients meeting the required criteria. There were no differences between groups regarding mean age (62 ± 17 vs 69 ± 16 years, p=0,870) and duration of symptoms (7 ± 6 vs 5,5 ± 2 days, p=0,155). The mean KOOS score and the subjective perception of knee function was higher in the @DI group, although not statistically significant (74,5 IQR 34,55-81,85 VS 44,55 IQR 30,9-50, p=0,394); 60% IQR 38-83 VS 45% IQR 20-50, p=0,484). No statistically significant differences were found between groups for median surgery duration (58 IQR 41-90 vs 51 IQR 32-72 minutes, p=0,699) and length of stay (15 IQR 9-38 vs 12 IQR 7-45 days, p=0,699).
There were no differences regarding PCR on admission (151,25 IQR 73,5-203,75 vs 131,5 IQR 69,75-231,13 mg/dL, p=0,472). Relatively to CRP monitoring, concerning the @DI group, there was a substantial decrease between the value of CRP on admission and the first assessment (52,1 IQR -34,20-122,92 VS 21,55 IQR -4,32-41,65 mg/dL, p=0,589), that was sustained over time. Concerning ODI group, the higher difference decrease was observed by the 2nd measurement (23,85 IQR -4,12-44,03 VS 71,8 IQR 5,08-330,08, p=0,310) and again by the 4th (21,4 IQR -1,63-74,2 vs 56,65 IQR 0-121, p=0,686). At discharge, @DI had a slightly higher mean CRP value (16 ± IQR 14,47-20,75 vs 5,8 ± IQR 3,15-10,6 mg/dL, p=0,04).
Fever spikes occurred slightly more frequently in the ODI group, most in the first 48h post operatively (p=0,558). There was only one case requiring re-operation, that occurred in the @DI group. This case was further complicated by a concomitant bacterial pneumonia and viral superinfection.
Conclusions
Although duration of surgery and hospitalization were slightly higher in the @DI group, our findings indicate that arthroscopy approach can lead to a quicker recovery in terms of clinical and analytical aspects. Although limited by a small sample, this paper indicates that, at least, the clinical results after an arthroscopic debridement are not inferior, when compared with open arthrotomy.
