ORIGINAL ARTICLE
The Distal Humerus Axial View: Assessment of Displacement in Medial Epicondyle Fractures Christopher D. Souder, MD,* Christine L. Farnsworth, MS,* Natalie P. McNeil, RT, RMSK,* James D. Bomar, MPH,* and Eric W. Edmonds, MD*w
Background: The assessment and treatment of childhood medial epicondyle humerus fractures continues to be associated with significant debate. Several studies demonstrate that standard radiographic views are unable to accurately portray the true displacement. Without reliable ways to assess the amount of displacement, how can we debate treatment and outcomes? This study introduces a novel imaging technique for the evaluation of medial epicondyle fractures. Methods: An osteotomy of a cadaveric humerus was performed to simulate a medial epicondyle fracture. Plain radiographs were obtained with the fracture fragment displaced anteriorly in 2-mm increments between 0 and 18 mm. Anteroposterior (AP), internal oblique (IR), lateral (LAT), and distal humerus axial (AXIAL) views were performed. Axial images were obtained by positioning the central ray above the shoulder at 15 to 20 degrees from the long axis of the humerus, centered on the distal humerus. Displacement (mm) was measured by 7 orthopaedic surgeons on digital radiographs. Results: At 10 mm displacement, AP views underestimated displacement by 5.5 ± 0.6 mm and IR views underestimated by 3.8 ± 2.1 mm. On LAT views, readers were not able to visualize fragments with 10 mm.13 Gottschalk et al14 proposed obtaining internal oblique (IR) radiographs in an effort to approximate true fracture displacement while avoiding the expense and radiation exposure associated with CT imaging. They found an improved ability to quantify displacement by using a trigonometric calculation involving the displacement as measured on the internal
Volume 35, Number 5, July/August 2015 Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.pedorthopaedics.com |
449
J Pediatr Orthop
Souder et al
rotation view. This study describes and evaluates a novel radiographic technique used to obtain an axial view of the distal humerus to more accurately assess the displacement of medial epicondyle fractures.
METHODS An osteotomy of the medial epicondyle of a skeletally mature humerus was performed by an experienced pediatric orthopaedic trauma surgeon using an osteotome at the site of the presumed medial epicondylar physis to simulate a medial epicondyle fracture. The fracture fragment was then repositioned onto the humerus and held in place using a radiolucent adhesive tape. Plain radiographs were obtained with the fracture fragment displaced anteriorly in 2-mm increments between 0 and 18 mm. Standard anteroposterior (AP), internal oblique (IR), and lateral (LAT) radiographic views were obtained in addition to a newly described distal humerus axial (AXIAL) view. Axial images were obtained by positioning the cadaveric humerus at a 45-degree angle from vertical (to represent a child resting their arm on the examination table). The x-ray tube was then positioned with the central ray above the shoulder at about 25 degrees from the long axis of the humerus, centered on the distal humerus (Figs. 1, 2). When applying the technique to an actual clinical patient, it was discovered that propping the forearm up off of the table so that the elbow is at a 90-degree angle provided a superior image of the distal humerus. Displacement (mm) was measured by 6 fellowshiptrained pediatric orthopaedic surgeons and 1 orthopaedic resident on digital radiographs (MergePACS 6.0.4; Merge Healthcare, Chicago, IL). Evaluators were asked to determine the displacement in the manner that they would for a clinical case. Correlation between AP and AXIAL measurements were tested by calculating a Pearson product-moment correlation coefficient (Pearson r) between the actual fragment displacement and the average of the measured values. The intraclass correlation coefficient (ICC) was used to evaluate measurement consistancy for the 7 readers. ICC was calculated for the AXIAL measurements alone and for the AP and IR combined and AP, IR, and LAT combined measurements (Pearson r and ICC were evaluated using SPSS version 12; SPSS Inc., Chicago, IL).
RESULTS Fragment displacement was underappreciated by every observer on all AP views (Fig. 3). At 10 mm actual displacement, AP views underestimated displacement by a mean of 5.5 ± 0.6 mm (range, 4.9 to 6.6 mm) and by a mean of 8.3 ± 0.7 mm (range, 7.3 to 9.4 mm) at 14 mm of actual displacement. Measurements of displacement from the IR view of the 10 mm actual displacement was both underappreciated (by 5.0 and 1.3 mm) and overappreciated (by 1.8 to 7.7 mm). On LAT views, readers were not able to visualize fragments with
The Distal Humerus Axial View: Assessment of Displacement in Medial Epicondyle Fractures Christopher D. Souder, MD,* Christine L. Farnsworth, MS,* Natalie P. McNeil, RT, RMSK,* James D. Bomar, MPH,* and Eric W. Edmonds, MD*w
Background: The assessment and treatment of childhood medial epicondyle humerus fractures continues to be associated with significant debate. Several studies demonstrate that standard radiographic views are unable to accurately portray the true displacement. Without reliable ways to assess the amount of displacement, how can we debate treatment and outcomes? This study introduces a novel imaging technique for the evaluation of medial epicondyle fractures. Methods: An osteotomy of a cadaveric humerus was performed to simulate a medial epicondyle fracture. Plain radiographs were obtained with the fracture fragment displaced anteriorly in 2-mm increments between 0 and 18 mm. Anteroposterior (AP), internal oblique (IR), lateral (LAT), and distal humerus axial (AXIAL) views were performed. Axial images were obtained by positioning the central ray above the shoulder at 15 to 20 degrees from the long axis of the humerus, centered on the distal humerus. Displacement (mm) was measured by 7 orthopaedic surgeons on digital radiographs. Results: At 10 mm displacement, AP views underestimated displacement by 5.5 ± 0.6 mm and IR views underestimated by 3.8 ± 2.1 mm. On LAT views, readers were not able to visualize fragments with 10 mm.13 Gottschalk et al14 proposed obtaining internal oblique (IR) radiographs in an effort to approximate true fracture displacement while avoiding the expense and radiation exposure associated with CT imaging. They found an improved ability to quantify displacement by using a trigonometric calculation involving the displacement as measured on the internal
Volume 35, Number 5, July/August 2015 Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.pedorthopaedics.com |
449
J Pediatr Orthop
Souder et al
rotation view. This study describes and evaluates a novel radiographic technique used to obtain an axial view of the distal humerus to more accurately assess the displacement of medial epicondyle fractures.
METHODS An osteotomy of the medial epicondyle of a skeletally mature humerus was performed by an experienced pediatric orthopaedic trauma surgeon using an osteotome at the site of the presumed medial epicondylar physis to simulate a medial epicondyle fracture. The fracture fragment was then repositioned onto the humerus and held in place using a radiolucent adhesive tape. Plain radiographs were obtained with the fracture fragment displaced anteriorly in 2-mm increments between 0 and 18 mm. Standard anteroposterior (AP), internal oblique (IR), and lateral (LAT) radiographic views were obtained in addition to a newly described distal humerus axial (AXIAL) view. Axial images were obtained by positioning the cadaveric humerus at a 45-degree angle from vertical (to represent a child resting their arm on the examination table). The x-ray tube was then positioned with the central ray above the shoulder at about 25 degrees from the long axis of the humerus, centered on the distal humerus (Figs. 1, 2). When applying the technique to an actual clinical patient, it was discovered that propping the forearm up off of the table so that the elbow is at a 90-degree angle provided a superior image of the distal humerus. Displacement (mm) was measured by 6 fellowshiptrained pediatric orthopaedic surgeons and 1 orthopaedic resident on digital radiographs (MergePACS 6.0.4; Merge Healthcare, Chicago, IL). Evaluators were asked to determine the displacement in the manner that they would for a clinical case. Correlation between AP and AXIAL measurements were tested by calculating a Pearson product-moment correlation coefficient (Pearson r) between the actual fragment displacement and the average of the measured values. The intraclass correlation coefficient (ICC) was used to evaluate measurement consistancy for the 7 readers. ICC was calculated for the AXIAL measurements alone and for the AP and IR combined and AP, IR, and LAT combined measurements (Pearson r and ICC were evaluated using SPSS version 12; SPSS Inc., Chicago, IL).
RESULTS Fragment displacement was underappreciated by every observer on all AP views (Fig. 3). At 10 mm actual displacement, AP views underestimated displacement by a mean of 5.5 ± 0.6 mm (range, 4.9 to 6.6 mm) and by a mean of 8.3 ± 0.7 mm (range, 7.3 to 9.4 mm) at 14 mm of actual displacement. Measurements of displacement from the IR view of the 10 mm actual displacement was both underappreciated (by 5.0 and 1.3 mm) and overappreciated (by 1.8 to 7.7 mm). On LAT views, readers were not able to visualize fragments with
