First Place: The utility of frozen sections in hip and knee reimplantation surgery

Background: The intraoperative decision to proceed with reimplantation total joint arthroplasty (rTJA) during the management of a periprosthetic infection can be challenging. Frozen section histopathology analysis can be employed to determine a wound's suitability for rTJA; however, its utility in predicting infection eradication remains controversial. Methods: All revision total hip and knee arthroplasty procedures performed at a single institution from 2002-2012 were reviewed retrospectively. Inclusion criteria were prior resection arthroplasty for infection and the availability of both intraoperative frozen section and culture results. Exclusion criteria consisted of procedures with missing data, a negative history of deep infection, subtotal revision arthroplasty, and prisoner status. Negative frozen sections (<10 polymorphonuclear leukocytes per high power field [PMN/HPF]) were correlated with the subsequent intraoperative culture outcomes and the well-established confounding factors of body mass index, number of previous surgeries, diabetes mellitus, and current smoking status. Results: A total of 331 rTJA procedures were identified, 43 of which were eligible for analysis. Negative frozen sections (<10 PMN/HPF) occurred in 13 rTJA procedures (30.2%) that subsequently had positive intraoperative cultures (negative predictive value=69.8%). Among these 13 procedures, 12 had <4 PMN/HPF. Patient confounding factors did not correlate with these outcomes. Conclusions: Negative intraoperative frozen sections during rTJA do not guarantee negative intraoperative cultures. In these patients, clinicians should consider staged rTJA and await a negative culture prior to revision arthroplasty.