Intraoperative frozen section for detecting active infection in failed hip and knee arthroplasties.

The authors reported the result of 40 sessions of intraoperative frozen section on polymorph count in 33 patients undergoing exploration or revision hip and knee surgery to detect active infection. The correlation rate between frozen section and permanent section was 0.95 to 1.00. If the polymorph count more than 5 per high-power field (40 X) was chosen as positive, the sensitivity, specificity, positive predictive value, and negative predictive value would be 0.93, 0.77, 0.68, and 0.95, respectively. Taking polymorph count of more than 10 as positive, the sensitivity, specificity, positive predictive value, and negative predictive value would then be 0.86, 0.85, 0.75, and 0.92, respectively. The authors concluded that intraoperative frozen section is an inexpensive, rapid, and helpful adjunct to detect active infection. A polymorph count of lower than 5 highly suggests the absence of active infection.