Session VII Pelvis and Acetabulum

Predictors of Total Hip Arthroplasty in Patients following Open Reduction and Internal Fixation of Acetabular Fractures

Philip R. Wolinsky, MD, B. L. Davison, MD, Yu Shyr, PhD, V. Talwalkar, MD, K. D. Johnson, MD

Vanderbilt University Medical Center, Nashville, TN

Purpose: To determine the incidence of total hip arthroplasty in patients following ORIF of acetabular fractures and investigate which, if any preoperatively available factors correlate with the eventual need for THR.

Materials and Methods: A chart review was performed on all patients who were treated for an acetabular fracture between January 1, 1986 and January 1, 1995. Patients who were treated non-operatively were excluded. All patients included in this study were treated by one of four orthopedic trauma surgeons. Other criteria for inclusion in the study group included follow-up of at least 1 year or a documented total hip arthroplasty. For all patients in the study group the operative note was reviewed to assess the operative surgeon's comments with respect to fracture pattern, the presence of intra-articular free osteochondral fragments, impacted osteochondral articular fragments, or damage to the acetabular or femoral head articular cartilage. All fractures were classified into one of the 10 fracture types as described by Letournel using the operative note and pre-operative radiographs. Pre-operative and post-operative plain radiographs and CT scans were reviewed to assess fracture pattern and quality of operative reduction. The reduction was graded based on the largest intra-articular step-off identified by reviewing the available post-operative AP pelvis, Judet Oblique, and CT scan of the pelvis. Each fracture was assigned to one of the following categories: anatomic reduction (< 1mm step-off), acceptable reduction (1 - 3mm step-off), or unacceptable reduction (> 3mm step-off). Statistical analysis was performed to determine factors predictive of need for total hip arthroplasty.

Results: 129 acetabular fractures satisfied the inclusion criteria. Seventy-nine patients were male and 49 were female. One patient had bilateral acetabular fractures treated by ORIF. The age at injury ranged from 15 - 75 years old with a mean of 34.7 years. Ninety-nine fractures had not undergone total hip arthroplasty at the time of follow-up. Follow-up in this group ranged from 12 to 108 months with an average of 37 months. The 99 fractures were classified as follows: 2 unknown, 1 anterior column, 1 anterior wall; 27 posterior walls; 1 posterior column; 15 transverse; 6 posterior column-posterior wall; 9 T-type; 22 transverse-posterior wall; 14 associated both column and 1 T/PW. A total hip arthroplasty had been performed on 30 hips in 29 patients, or 23% of the study group. The 30 fractures were classified as follows: 2 T-type; 7 posterior wall; 2 transverse; 6 posterior column-posterior wall; 11 transverse-posterior wall; and 2 associated both column. The time from injury to total hip arthroplasty ranged from 3 months to 90 months with an average of 24 months. Twenty-two of the 30 total hip arthroplasties were performed within 2 years of the patients' acetabular fracture. Statistical analysis of the data revealed a correlation between the presence of articular impaction, a free intra-articular fragment, post wall fracture component, older age and the eventual need for total hip. The presence or absence of hip dislocation fracture pattern and the quality of operative reduction did not have a statistically significant correlation with the need for total hip arthroplasty. Power analysis suggests that with a larger sample size, operative reduction may become significant. Due to the limited number of fractures in each classification, statistical analysis was performed after grouping the 5 fracture types termed simple (posterior wall, posterior column, anterior wall, anterior column, and transverse) and the 5 types termed associated (posterior column-posterior wall, transverse- posterior wall, anterior column- posterior hemitransverse, T -shaped, and associated both-column) together.

Discussion: In this study the presence of articular impaction, free intrarticular fragments, a posterior wall fracture component and older age correlated with the failure of treatment as defined by need for a total hip arthroplasty. The incidence of total hip arthroplasty was not significantly different when comparing the simple and associated fracture types. No significant difference was demonstrated in the incidence of total hip arthroplasty in patients who had anatomic, acceptable, or unacceptable operative reductions. 23% of patients in this study required total hip arthroplasty at the time of follow-up. In patients who had a failure of treatment for their acetabular fracture, the majority had total hip arthroplasty within 2 years of their injury.

Conclusion: Preoperatively available factors that correlate with treatment failure of ORIF of acetabular fractures include articular impaction, a free osteochondral intra-articular fragment, presence of a posterior wall fracture component, and older age. Multiple logistic regression analysis revealed that presence of intrarticular free fragments is associated with fourfold increase in the risk for need of THR.