Surgeons often rely on intra-operative histology (frozen sections [FS]) to determine the next step in surgical management during the second stage (re-implantation surgery) of 2-stage revision arthroplasty. The purpose of the study is to assess the accuracy of permanent sections (PS) and FS in the diagnosis of persistent infection during re-implantation in patients with an inflammatory arthritis.
From 2001 to 2016, 47 planned second-stage revision total hip arthroplasty and total knee arthroplasty in patients with inflammatory arthritis were identified. Revisions were classified as having persistent infection if they were Musculoskeletal Infection Society positive at the time of second stage. PS or FS was considered to be positive for infection when at least one of the specimens demonstrated an acute inflammation. Receiver operating characteristic analysis was performed to obtain the diagnostic parameters.
There were 9 (19%) persistent infections. Both PS and FS had very high specificity (PS = FS = 94.7%). Sensitivity of PS was higher than FS, although not statistically significant (PS = 88.9%, FS = 55.6%, P = .083). Overall, PS had a better diagnostic utility than FS (area under the curve: PS vs FS = 0.92 vs 0.75, P = .045). Four specimens had discrepancies between PS and FS histology. In all 4 instances, the specimens were read as positive (infected) by PS, but negative by FS.
Histological analysis is recommended at the time of re-implantation surgery even in patients with inflammatory arthritis. PS had a better diagnostic utility than FS suggesting that areas of acute inflammation may be scattered and may not always be captured in the specimens taken for FS.