Session VII - Pelvic Trauma

The Correlation between Early Radiographic Results and Long-Term Clinical Outcome following Fracture of the Acetabulum

Mark C. Reilly, MD, Joel M. Matta, MD

Good Samaritan Hospital, Los Angeles, California, USA

Purpose: To determine whether long-term clinical outcome after fracture of the acetabulum can be predicted from early post-operative radiographic and clinical examination.

Methods: Between 1980 and 1991, the senior author utilized a prospective protocol for the post-operative evaluation of 559 patients after operatively treated acetabulum fractures. Patients were examined at 6 months, 12 months, 24 months and biannually thereafter. Separate clinical and radiographic grades were determined. Clinical evaluation was performed using the modified D'Aubigne-Postel grading system (pain, gait, range of motion). Radiographic grading was based on the evaluation of an AP pelvis radiograph for osteophyte formation, joint space narrowing and subchondral sclerosis. Two hundred seventy-one patients were followed for a minimum of 5 years (average 8.4 years). These patients were retrospectively reviewed to determine to what extent the early radiographic and clinical results were predictive of long-term outcome. The percent of poor long-term outcomes which were correctly predicted by fair or poor early results were determined for each of the follow-up periods of 6, 12 and 24 months. Poor outcomes due solely to restriction of motion by heterotopic bone were excluded. In addition, a survivorship analysis was independently performed for each of these 3 periods. Failure was defined as either hip arthrodesis or total hip arthroplasty.

Results: Radiographic grading at 6 months postoperatively was correctly predictive of a poor clinical outcome in only 45% of cases. Radiographs from 12 and 24 months postoperatively were predictive of poor results in 70% and 83% respectively. Clinical evaluation at both 12 and 24 months was predictive of a poor long-term outcome in 71%. Survivorship analysis revealed a 63% survival at 10 years for the group as a whole. For those patients with good or excellent radiographic grades at 6 months, the 9-year survival rate was 78%. Good or excellent radiographic grades and 12 and 24 months corresponded to 10-year survival rates of 85% and 93%.

Conclusion: These results give the surgeon important information to use in the discussion of prognosis following acetabulum fracture. Early radiographic evaluation was more accurate than early clinical results in predicting eventual post-traumatic arthritis and poor long-term clinical outcome. Radiographs taken at 2 years following fracture treatment were much more likely to be accurate predictors of future outcome than those taken at earlier intervals. This reinforces the need for a minimum of 2 years of radiographic and clinical follow-up for any investigation of outcomes after fracture of the acetabulum.