Ronald Isida1,2, Varenka Bariatinsky3,4, Gregory Kern3,4, Gregoire Dereudre3,4, Xavier Demondion3,5, Christophe Chantelot3,4. 1. Université Lille-Nord de France, 59000, Lille, France. isida.ronald@gmail.com. 2. Département de Chirurgie d'Orthopédique et Traumatologique, Hôpital Salengro, CHRU de Lille, Rue Emile Laine, 59000, Lille, France. isida.ronald@gmail.com. 3. Université Lille-Nord de France, 59000, Lille, France. 4. Département de Chirurgie d'Orthopédique et Traumatologique, Hôpital Salengro, CHRU de Lille, Rue Emile Laine, 59000, Lille, France. 5. Service de Radiologie et d'Imagerie, Bâtiment CCIAL, CHRU de Lille, Rue Emile Laine, 59000, Lille, France.
Abstract
BACKGROUND: Trochanteric fractures are common but difficult to analyse in the elderly on plain X-rays. Fixation failures are related to the severity of the comminution, but comminution and the degree of instability are not easy to determine on standard X-rays. Use of computed tomography (CT) improves assessment of complex fractures, but this finding has not been confirmed versus intraoperative data. The primary objective of this prospective study was to determine the error rate when evaluating comminution on X-rays and CT scans. The secondary objectives were to determine whether CT data on comminution and stability were consistent with intraoperative findings and to define the fracture characteristics. HYPOTHESIS: Standard X-ray assessment underestimates the complexity of trochanteric fractures and is not very reproducible. METHODS: Between January and December 2013, all proximal femur fractures in the trochanter area of patients aged 75 years or older (mean age 85) were analysed prospectively with standard X-rays and computed tomography (CT). One hundred and ten patients (88 women and 22 men) with trochanteric fractures were included in the study. Fracture stability was evaluated using the Müller AO classification; the other fracture characteristics were evaluated independently. A senior surgeon evaluated the anonymised X-rays. A radiologist specialised in musculoskeletal imaging interpreted the CT scan images. All patients underwent total hip arthroplasty (110 patients) and 104 fixations of the greater trochanter. The X-ray and CT findings were compared to the intraoperative findings (gold standard) to evaluate their reproducibility. RESULTS: The reproducibility of the X-ray evaluation was poor for comminution, with a kappa of 0.4, sensitivity of 44 % and a negative predictive value of 29 % but a positive predictive value and specificity of 100 %. The CT evaluation had a kappa of 0.94, sensitivity of 95 % and negative predictive value of 79 %. According to the AO classification, unstable fractures were found in 65 patients based on X-rays, 80 patients based on CT scan and 83 patients based on intraoperative findings. Comminution was mainly identified during surgery; it was medial in 34 % of cases, posterior in 22 % and posteromedial in 21.8 %. DISCUSSION AND CONCLUSION: In the current study, 75.4 % of fractures were complex and unstable, with medial and/or posterior comminution present in three-quarters of these fractures. X-ray assessment underestimates posterior comminution and the complexity of the fracture line and was not very reproducible. The current results of this study suggest that comminution contributes to instability and that this finding is not taken into account in the AO classification, which is not well suited for this type of fracture and raises the question of how to best evaluate and treat these fractures. LEVEL OF EVIDENCE: Prospective diagnostic study, Level III.
BACKGROUND: Trochanteric fractures are common but difficult to analyse in the elderly on plain X-rays. Fixation failures are related to the severity of the comminution, but comminution and the degree of instability are not easy to determine on standard X-rays. Use of computed tomography (CT) improves assessment of complex fractures, but this finding has not been confirmed versus intraoperative data. The primary objective of this prospective study was to determine the error rate when evaluating comminution on X-rays and CT scans. The secondary objectives were to determine whether CT data on comminution and stability were consistent with intraoperative findings and to define the fracture characteristics. HYPOTHESIS: Standard X-ray assessment underestimates the complexity of trochanteric fractures and is not very reproducible. METHODS: Between January and December 2013, all proximal femur fractures in the trochanter area of patients aged 75 years or older (mean age 85) were analysed prospectively with standard X-rays and computed tomography (CT). One hundred and ten patients (88 women and 22 men) with trochanteric fractures were included in the study. Fracture stability was evaluated using the Müller AO classification; the other fracture characteristics were evaluated independently. A senior surgeon evaluated the anonymised X-rays. A radiologist specialised in musculoskeletal imaging interpreted the CT scan images. All patients underwent total hip arthroplasty (110 patients) and 104 fixations of the greater trochanter. The X-ray and CT findings were compared to the intraoperative findings (gold standard) to evaluate their reproducibility. RESULTS: The reproducibility of the X-ray evaluation was poor for comminution, with a kappa of 0.4, sensitivity of 44 % and a negative predictive value of 29 % but a positive predictive value and specificity of 100 %. The CT evaluation had a kappa of 0.94, sensitivity of 95 % and negative predictive value of 79 %. According to the AO classification, unstable fractures were found in 65 patients based on X-rays, 80 patients based on CT scan and 83 patients based on intraoperative findings. Comminution was mainly identified during surgery; it was medial in 34 % of cases, posterior in 22 % and posteromedial in 21.8 %. DISCUSSION AND CONCLUSION: In the current study, 75.4 % of fractures were complex and unstable, with medial and/or posterior comminution present in three-quarters of these fractures. X-ray assessment underestimates posterior comminution and the complexity of the fracture line and was not very reproducible. The current results of this study suggest that comminution contributes to instability and that this finding is not taken into account in the AO classification, which is not well suited for this type of fracture and raises the question of how to best evaluate and treat these fractures. LEVEL OF EVIDENCE: Prospective diagnostic study, Level III.
Authors: E Cavaignac; M Lecoq; A Ponsot; A Moine; N Bonnevialle; P Mansat; N Sans; P Bonnevialle Journal: Orthop Traumatol Surg Res Date: 2012-12-25 Impact factor: 2.256