Peroneal nerve palsy after total knee arthroplasty. Assessment of predisposing and prognostic factors.

Thirty-two postoperative peroneal-nerve palsies in thirty patients were documented in a retrospective review of 10,361 consecutive total knee arthroplasties performed at one institution from 1979 through 1992. The mean age of the thirty patients was sixty-five years (range, twenty-eight to seventy-eight years). Four of these patients had had a previous proximal tibial osteotomy and five had had a previous lumbar laminectomy. Ten knees (ten patients) had preoperative valgus alignment of 12 degrees or more. A control group of 100 patients who had had total knee arthroplasty during the same period was computer-matched to the patients by age, sex, and operating surgeon. Comparison of this control group with the thirty patients showed that epidural anesthesia for postoperative control of pain (p < 0.03), previous laminectomy (p < 0.04), and preoperative valgus deformity (p < 0.0001) were significantly associated with peroneal nerve palsy. The relative risk for patients who had had a previous proximal tibial osteotomy was doubled, but this was not significant (p < 0.4). To determine risk factors associated with anesthesia better, a subgroup of 4388 total knee arthroplasties performed during a five-year period (1988 through 1992) was also studied. In this sample, twenty-five peroneal-nerve palsies were documented. Eighteen occurred after epidural anesthesia (p < 0.03); five, after general anesthesia; and two, after spinal anesthesia. An important finding of this study is the high frequency of delayed presentation of peroneal nerve palsy. We believe that epidural anesthesia for postoperative control of pain leads to decreased proprioception and sensation postoperatively. It is postulated that positioning of the limb in this unprotected state may be a factor in the late development of palsy. The concept of the so-called double-crush phenomenon may partially explain the palsy seen in the patients who had had a lumbar laminectomy and asymptomatic peripheral neuropathy.