Risk Factors for Composite Failure of Hip Dysplasia Treated With Periacetabular Osteotomy: A Minimum 10-Year Follow-up

Introduction: Periacetabular osteotomy (PAO) is a common surgical treatment of prearthritic hip dysplasia in young adults, but there are few long-term studies of clinical outcomes. The purpose of this investigation was to report a minimum 10-year clinical follow-up of hip dysplasia treated with PAO and identify risk factors for composite failure. Methods: We identified 151 patients (198 hips) who underwent PAO to treat hip dysplasia at a single institution. Enrolled subjects completed a series of six patient-reported outcome instruments and provided information about subsequent surgeries. We defined composite failure as conversion to total hip arthroplasty or modified Harris Hip Score <= 70. Logistic regression with generalized estimating equations was used to evaluate the relationships between odds of failure and potential predictor variables in univariate and multivariate analyses. Results: A total of 124 subjects (167 hips) with a minimum 10-year follow-up were enrolled. The median time from PAO to the final follow-up was 13 years (range 10-18 years). There were 71 hips that met criteria for failure: 32 with total hip arthroplasty and 39 with modified Harris Hip Score <= 70. Univariate logistic regression analyses revealed multiple preoperative factors that predicted composite failure: increased age and body mass index, osteoarthritis (OA), and more severe acetabular dysplasia. Postoperative factors that predicted failure included lateral undercoverage and formation of heterotopic ossification (HO). The final multivariate model identified body mass index >= 30 kg/m(2) (odds ratio [OR], 3.84 [95% confidence interval (CI), 1.68-8.78], P = 0.001), higher preoperative Tonnis grade OA (OR, 2.65 [95% CI, 1.50-4.66], P < 0.001), and HO formation (OR, 16.52 [95% CI, 2.08-135.96], P = 0.009) as independent predictors of failure. Conclusions: This study corroborates current hip dysplasia literature, identifying increasing age and presence of preoperative OA as risk factors for composite failure in univariate analyses. In addition, we found that obesity and HO formation were independent predictors of persistent hip dysfunction.