No. of knees
Duration of symptoms
Flood et al. 
Gram +ve organisms with sensitivity
Waldman et al. 
Dixon et al. 
Mixture of organisms, mostly Staph sp.
Better results in cementless prostheses
Ilahi et al. 
12 l of irrigant
Drain left in situ
Liu et al. 
Use of closed, continuous irrigation system
Vancomycin added to irrigant
Chung et al. 
Remaining 37.5% underwent subsequent open procedure achieving 100% eradication
Arthroscopic debridement was first reported in two successful cases by Flood in 1988 . However, larger series were less promising. Waldman et al.  reported 16 cases in 2000 with only 38% success. This contrasted with 83% success rate in a series of open debridements, performed by the same group. The group recommended it only in frail or anticoagulated patients.
Penicillin-sensitive streptococcal infection is reported to have the highest success rates. In a series of 19 patients from the Mayo Clinic, three were treated with arthroscopic DAIR with 100% success . When looking at open DAIR, Zurcher-Pfund et al. demonstrated a much higher eradication rate with streptococcal infections over MSSA (up to 89%) . This is likely due to a poorer rate and quality of biofilm production when compared to MSSA . This is an area where arthroscopic debridement could be considered.
Time from onset of symptoms seems to be a crucial variable in the ultimate success of arthroscopic debridement, irrigation and implant retention. If the symptoms are less than 12 h old and the arthroscopy is performed within 12 h of admission, the success rate may be 100% . In previously well-functioning TKRs, arthroscopic debridement up to 72 h from the onset of symptoms can result in a 62.5% success rate . Most literature on the subject involves patients with late-onset infections due to acute haematogenous spread. Ilahi et al. demonstrated 100% survivorship of implants in five patients treated within 7 days of clinical onset . Liu et al. included 17 patients in their study with, again, less than 7 days of acute symptoms, reporting an 88% prosthesis survival rate .
Implant type may have a bearing on the outcome. Dixon et al. noted a trend suggesting a more favourable outcome following arthroscopic debridement of cementless TKRs over cemented ones . Some authors have theorised that infection eradication may be more effective in cementless TKR due to the absence of an “avascular” zone, i.e. bone-cement-prosthesis interface . It is harder to perform arthroscopy on more constrained prostheses, and there does not appear to be a place for arthroscopic treatment of revision implants. Arthroscopic DAIR of compartmental knee replacements is not, as yet, reported.
There are papers reporting particularly poor results. Byren et al. evaluated the use of debridement and implant retention in 112 cases (13% arthroscopic debridement), finding success rates of 47% for arthroscopic washout and 88% for open washout. Arthroscopic surgery was also a significant predictor of failure of subsequent revision surgery . Dixon et al. reported implant retention in 9 of 15 patients . The 2013 International Consensus on Periprosthetic Infection had 90% agreement that there is no role for arthroscopic debridement of infected knee replacement .
Failure of DAIR is associated with failure to exchange meniscal bearings and failure to perform extensive debridement of the whole soft tissue envelope. As neither is possible in arthroscopic debridement, many authors dismiss the technique as inadequate and do not routinely practise it.
There has not been a prospective trial comparing arthroscopic with open debridement, and it seems unlikely that there ever will. Similarly, there are no studies alluding to arthroscopy followed by lifelong suppression treatments.
The aim of treatment is elimination of all microorganisms with maintenance of a stable, pain-free joint. The patient will benefit most from the least invasive technique which eradicates infection. Avoiding revision surgery reduces the risks of bone loss or aseptic failure mechanisms. Avoiding an arthrotomy means less damage to the extensor mechanism and lower wound failure rates. The choice of treatment of acute PJI is multifactorial, and the doubts as to the efficacy of arthroscopic debridement mean that it is often overlooked. It is, undoubtedly, the least invasive surgical option and can be considered in a small number of patients who meet several criteria. It is only suited to acute infective presentations of well-fixed implants. Success is far from guaranteed and is only likely in streptococcal infection. The arthroscopic surgery is only the beginning of treatment, and prolonged antibiotic therapy is mandated. It should never be used in chronic infected cases.