High failure rate of 2-stage revision for the infected total elbow arthroplasty: a single institution’s experience

Despite 2-stage revision being a common treatment for elbow prosthetic joint infection (PJI), failure rates are high. The purpose of this study was to report on a single institution’s experience with 2-stage revisions for elbow PJI and determine risk factors for failed eradication of infection. The secondary purpose was to determine risk factors for needing allograft bone at the second stage of revision in the setting of compromised bone stock. We retrospectively analyzed all 2-stage revision total elbow arthroplasties (TEAs) performed for infection at a single institution between 2006 and 2020. Data collected included demographics and treatment course prior to, during, and after 2-stage revision. Radiographs obtained after explantation and operative reports were reviewed to evaluate for partial component retention and incomplete cement removal. The primary outcome was failed eradication of infection, defined as the need for repeat surgery to treat infection after the second-stage revision. The secondary outcome was the use of allograft for compromised bone stock during the second-stage revision. Risk factors for both outcomes were determined. Nineteen patients were included. Seven patients (37%) had either the humeral or ulnar component retained during the first stage, and 10 (53%) had incomplete removal of cement in either the humerus or ulna. Nine patients (47%) had allograft strut used during reimplantation and reconstruction. Nine patients (47%) failed to eradicate the infection after 2-stage revision. Demographic data were similar between the repeat-infection and nonrepeat-infection groups. Six patients (60%) with retained cement failed compared with 3 patients (33%) with full cement removal (P = .370). Two patients (29%) with a retained component failed compared to 7 patients (58%) with full component removal (P = .350). Allograft was used less frequently when a well-fixed component or cement was retained, with no patients with a retained component needing allograft compared to 9 with complete component removal (P = .003). Three patients (30%) with retained cement needed allograft, compared with 6 patients (67%) who had complete cement removal (P = .179). Nearly half of the patients failed to eradicate infection after 2-stage revision. The data did not demonstrate a clear association between retained cement or implants and risk of recurrent infection. Allograft was used less frequently when a component and cement were retained, possibly serving as a