Is Assessment of Femoral Head Perfusion During Modified Dunn for Unstable Slipped Capital Femoral Epiphysis an Accurate Indicator of Osteonecrosis?

BACKGROUND: The modified Dunn procedure, which is an open subcapital realignment through a surgical dislocation approach, has gained popularity for the treatment of unstable slipped capital femoral epiphysis (SCFE). Intraoperative monitoring of the femoral head perfusion has been recommended as a method of predicting osteonecrosis; however, the accuracy of this assessment has not been well documented. QUESTIONS/PURPOSES: We asked (1) whether intraoperative assessment of femoral head perfusion would help identify hips at risk of developing osteonecrosis; (2) whether one of the four methods of assessment of femoral head perfusion is more accurate (highest area under the curve) at identifying hips at risk of osteonecrosis; and (3) whether specific clinical features would be associated with osteonecrosis occurrence after a modified Dunn procedure for unstable SCFE. METHODS: Between 2007 and 2014, we performed 29 modified Dunn procedures for unstable SCFE (16 boys, 11 girls; median age, 13 years; range, 8-17 years); two were lost to followup before 1 year. During this period, six patients with unstable SCFE were treated by other procedures. All patients undergoing modified Dunn underwent assessment of epiphyseal perfusion by the presence of active bleeding and/or by intracranial pressure (ICP) monitoring. In the initial five patients perfusion was recorded once, either before dissection of the retinacular flap or after fixation by one of the two methods. In the remaining 22 patients (81%), perfusion was systematically assessed before dissection of the retinacular flap and after fixation by both methods. Minimum followup was 1 year (median, 2.5 years; range, 1-8 years) because osteonecrosis typically develops within the first year after surgery. Patients were assessed for osteonecrosis by the presence of femoral head collapse at radiographs obtained every 3 months during the first year after surgery. Seven (26%) of the 27 patients developed osteonecrosis. Measures of diagnostic accuracy including sensitivity, specificity, and the area under the receiver operating curve (AUC) were estimated. Multiple variable logistic regression analyses were used to test whether the test options were better than random chance (AUC > 0.50) at differentiating between patients who did versus did not develop osteonecrosis. Nonparametric methods were used to test for a difference in AUC across the four methods. A secondary analysis was performed to identify risk factors associated with osteonecrosis. RESULTS: After adjusting for body mass index, which was found to be a confounding variable, assessment of femoral head perfusion with ICP monitoring before retinaculum dissection (adjusted AUC: 0.79; 95% confidence interval [CI], 0.58-0.99; p = 0.006), femoral head perfusion with ICP monitoring after definitive fixation (adjusted AUC: 0.82; 95% CI, 0.65-1.0; p < 0.001), bleeding before retinaculum dissection (adjusted AUC: 0.77; 95% CI, 0.58-0.96; p = 0.006), and bleeding after definitive fixation (adjusted AUC: 0.81; 95% CI, 0.63-0.99; p = 0.001) were found to be helpful at identifying osteonecrosis. We were not able to identify a specific test that had performed best because there was no difference (p = 0.8226) in AUC across the four methods. With the numbers available, we were unable to identify clinical factors predictive of osteonecrosis in our cohort. CONCLUSIONS: Assessments of femoral head blood perfusion by ICP monitoring or by the presence of active bleeding in combination with the patient's body mass index are effective at differentiating between patients who do versus do not develop osteonecrosis after a modified Dunn procedure for unstable SCFE. Additional research is needed to determine whether information gained from assessment of femoral head perfusion during surgery should be used to guide targeted treatment recommendations that may reduce the development of femoral head deformity secondary to osteonecrosis. LEVEL OF EVIDENCE: Level III, diagnostic study.