Background: No controlled data exists regarding the risk factors for redislocation after a technically proficient open reduction for developmental dysplasia of the hip (DDH). The purposes of this study were to examine predictors of redislocation and to evaluate the long-term outcomes after revision surgery.
Methods: We performed a retrospective match-controlled study comparing 22 patients who had successful open reduction for DDH with 22 who required revision open reduction. Radiographs were compared in terms of acetabular index, pelvic width, triradiate cartilage width, height of dislocation, size of ossific nucleus, abduction angle in the spica cast, Tönnis grade, and Severin grade. At final follow-up, Sharp's angle, center-edge angle, migration index, and continuity of Shenton's line were compared between the 2 cohorts. Hips were reclassified according to the Tönnis and Severin criteria, and graded for avascular necrosis. Univariate t tests, multivariate logistic regression, and Fisher exact tests were used to compare the statistical data.
Results: Twenty-five of 421 patients (5.9%) developed a redislocation at a mean of 124 days after the initial open reduction. Patients with right or bilateral DDH were significantly more likely to fail (P=0.01). Compared with matched controls, the 22 study patients had significantly larger pelvic width and lower abduction angle (mean 39 degrees vs. 51 degrees) in the postoperative spica cast (P=0.003 and 0.037). According to the surgeon's findings at revision surgery, the most common reasons for failure were a dysmorphic femoral head and abnormal femoral version. At final follow-up, subluxation rate and Severin grade were significantly higher in the revision group versus controls but the incidence of avascular necrosis was comparable (revision group=55% and control group=41%; P=0.55).
Conclusions: On the basis of this study, right-sided (or bilateral) involvement, greater pelvic width, and decreased abduction in the spica cast were risk factors for redislocation after open reduction. We believe that abnormal femoral version and femoral head dysplasia are also important factors although they were not addressed in this study.
Level of Evidence: Level III (case-controlled study).