This study was performed in line with the principles of the Declaration of Helsinki. This study was approved by our institutional review board and a waiver of the requirement for written informed consent was granted (Asan Institute for Life Science, approval no.: 2021–3165-0003). Data collection was performed in accordance with relevant guidelines and regulations issued by the committee.
Patient selection
This retrospective comparative study was conducted in two university teaching hospitals. The inclusion criteria were as follows: patients aged ≥ 60 years, patients with acute acetabular fractures, and patients who had received surgical treatment. Initially, 127 patients were enrolled from the pelvis and acetabular fracture cohort database from 2003 to 2020. Then, we excluded 73 patients with fracture patterns unrelated with dome impaction (anterior wall, posterior wall, posterior column, and posterior column with posterior wall), who were treated conservatively or with acute THA, had concomitant pelvic ring injury, had periprosthetic fractures, and were followed up for < 12 months. Fracture patterns which can contain dome impaction include anterior column, transverse, transverse and posterior wall, T-type, anterior column and posterior hemitransverse, and both column fractures. Finally, 54 patients were enrolled in the present study (Fig. 2).
Surgical technique for reduction of dome impaction
The surgeries were performed by three operators (H.-C.S., J.J.K, and J.W.K.) with more than 10 years of experience performing pelvic and acetabular fracture surgeries. The patient was placed in a supine position with silicone padding under the knee to flex the hip and knee joints so that the iliopsoas muscle, external iliac vessel, and femoral nerve could relax. Skeletal traction at the distal femur was prepared to facilitate reduction of central dislocation of femoral head and exposure of acetabular joint surface.
After identifying the obturator nerve through the intrapelvic approach, the fracture site between the anterior column and quadrilateral plate was exposed. Hematoma and soft tissue in the fracture site were removed carefully, and an elevator was inserted through the fracture site to confirm the position of the dome fragment in the C-arm. Then, the impacted dome fragment was reduced using an elevator and temporarily fixed using a K-wire (Fig. 3).
Surgical procedure for reduction of dome impaction. (a) The impacted dome fragment was reduced using an elevator. (b) The reduced dome fragment was temporarily fixed using a K-wire. (c) A 3.5-mm cannulated screw was inserted along the K-wire as a guide pin.
Through a lateral window, the anterior column was reduced and fixed using a five-hole reconstruction plate or a 3.5-mm locking compression plate. The K-wire for a cannulated screw was started 1.5 cm downward and 1.5 cm medially from the anterior superior iliac spine in the lateral window to fix the dome fragment directly or provide support with a rafting screw in tibial plateau surgery. The K-wire was inserted only to cross the dome fragment without penetrating the quadrilateral surface because the K-wire could interfere with the reduction of the quadrilateral surface if it was fully inserted before reduction.
The quadrilateral surface was reduced using a ball and spike or collinear clamp and was then penetrated by the K-wire. A 3.5-mm cannulated screw was inserted along the K-wire which was used as a guide pin for the cannulated screw. The anterior column was stabilized with a long reconstruction plate, and the posterior column was fixed using a posterior column screw if necessary.
After postoperative recovery, active and passive knee and hip range of motion exercises were commenced. Wheelchair ambulation was encouraged as soon as possible if the pain was tolerable. Afterward, weight-bearing was gradually permitted according to the patient’s condition and fracture pattern.
Data collection and reduction evaluation
Data on demographics, including age; sex; body mass index (BMI); age-adjusted Charlson comorbidity index (CCI)14; injury mechanism; the lowest T-score for the lumbar spine, neck, and total femur of the bone mineral density; preoperative osteoarthritis (OA); and follow-up period, were assessed. A low-energy mechanism of injury was defined as falls from heights of ≤ 1 m15. Preoperative OA was evaluated in the fractured hip whenever possible using Kellgren and Lawrence (K–L) grade16 by two authors (E.J.L. and J.A.). When the evaluation was difficult due to intra-articular fracture, the OA grade of the contralateral hip joint was used. Fracture pattern was evaluated based on the Judet and Letournel classification16. Approaches used for the acetabular fracture surgery were evaluated using surgical records. The presence of negative predictive factors, such as posterior wall involvement, dome impaction, femoral head impaction, posterior dislocation, central dislocation, and separation of the quadrilateral plate was evaluated through a preoperative computed tomography (CT) scan.
Quality of reduction was evaluated according to the Matta criteria17. We evaluated gap and step displacement in the weight-bearing dome based on the pelvis CT scan with axial, coronal, and sagittal planes18. Patients with anatomic and imperfect displacement of ≤ 3 mm according to the Matta criteria were classified as the ‘good reduction group’ and those with poor displacement of > 3 mm as the ‘poor reduction group’ (Fig. 4).
Evaluation of quality of reduction. (a) Gap and (b and c) step displacement in the weight-bearing dome was evaluated based on postoperative computed tomography with (a) axial, (b) coronal, and (c) sagittal planes.
Outcome assessment and comparison
Postoperative radiographs at the last follow up or immediately before THA were used to assess postoperative OA using the K–L grade. Radiographic OA was defined as K–L grades 3 and 4. Functional score was assessed using the Harris hip score (HHS)19 at the final outpatient visit or before THA conversion. Patients without functional scores at the outpatient visit were interviewed by telephone. Conversion to THA was assessed using final radiographs and telephone interview.
The outcomes were categorized dichotomously using radiographic grade and functional score. Poor outcomes were defined as one of the following: conversion to THA or the presence of radiographic OA, combined with poor or fair results in the HHS (HHS < 80). The HHS value is based on the 15th percentile of HHS score following acetabular fracture, which represents the THA conversion after acetabular fracture21,22,23.
Therefore, a good outcome was defined as requiring no additional surgical treatment, good or excellent results in the HHS (HHS ≥ 80), or no evidence of radiographic OA. Deceased patients were evaluated by assessing the radiographs and medical records at the final outpatient visit. Radiographic OA and severe pain or inability to perform activities of daily living were considered poor outcomes.
Medical complications such as pneumonia, urinary tract infection, and pulmonary embolism and surgical complications such as postoperative foot drop that resulted from sciatic nerve injury, hardware breakage, and reoperation for any reason were assessed.
Radiographic OA, HHS, THA conversion, good/poor outcomes, and complications were compared between the good and poor reduction groups. A subgroup analysis was performed in patients with dome impaction according to the postoperative reduction state, and in patients in the good reduction group according to presence of dome impaction. A multivariate analysis was performed to assess which factors were associated with good/poor outcomes.
Statistical analysis
Statistical analyses were performed using the SPSS Statistics version 23.0 (IBM Corp., Armonk, NY, USA). We used the chi-square test or Fisher’s exact test for categorical variables, and the independent t-test or the Mann–Whitney test for continuous variables. A logistic regression analysis was conducted for the multivariate analysis. First, a bivariate analysis was performed for the variables which could be clinically significant. Then, the variables with P < 0.2 in the bivariate analysis were included in the multivariate analysis. All continuous data are expressed as means and standard deviations. Significance was set at P < 0.05, and Bonferroni correction was applied for multiple comparisons.