We retrospectively reviewed 791 consecutive revision TKAs performed at our institution between years 2004 and 2011. Following exclusion for infection, periprosthetic fracture, prior revision TKA, positive cultures, incomplete records, and patients with less than 24-month follow-up, a total of 228 aseptic revisions (89 knees with 1 abnormal serologic marker) were available for final analysis. No patients met the current established criteria for infection. All knees underwent revision TKA using antibiotic-impregnated cement. The 2 groups were compared in terms of overall survivorship, infection rate, and rate and causes of subsequent aseptic revision.
The average follow-up was 60 months (24-110). There were no significant differences between the 2 groups in terms of age, sex, American Society of Anesthesiologists class, and Charlson comorbidity index. A preoperative abnormality of either ESR or CRP significantly increased the risk for reoperation for all reasons (odds ratio [OR], 3.2; P = .0028), infection (OR, 4.0; P = .034), and revision for aseptic loosening (OR, 3.69; P = .044). There were no differences in reoperations for any other reason. The average time to revision in the study group was 28.3 months compared to 40.0 months in the control group (P = .213).
A single abnormality in either the ESR or CRP increased the likelihood of both infection and reoperation following revision TKA. Conventional methods and criteria for infection detection may not be sufficiently sensitive or specific in these cases. Further work-up with additional modalities may help increase the confidence of aseptic failure before revision TKA.