Evaluating the Ability of Hand Surgeons to Reliably Classify the Percentage of Tearing Based on MRI for Patients With a Partial Distal Biceps Tendon Tear

Abstract

Background: Patients with partial distal biceps tendon tears (PDBT) have both surgical and nonsurgical treatment options. The percentage of the tendon that has torn plays an important role in decision making, although this can be difficult to reliably measure, even with advanced imaging. The purpose of this study was to assess the reliability of hand surgeons within our practice in determining the percentage of the distal biceps tendon that has torn based on magnetic resonance imaging (MRI) in patients with PDBT. Methods: We retrospectively identified 50 patients at our institution who received an MRI for a distal biceps injury and received a diagnosis of PDBT. All of these 50 patients’ initial elbow MRI images were reviewed by 4 fellowship-trained hand and upper extremity surgeons within our practice. Each reviewer classified the partial tear into 1 of 3 categories: less than 1/3 torn, between 1/3 and 2/3 torn, and greater than 2/3 torn. In addition, the presence or absence of an FABS view (elbow flexed, shoulder abducted, forearm supinated) was identified. Interobserver agreement and reliability were calculated with Cohen’s kappa test. Results: One patient’s images were not available in our electronic medical record, resulting in 49 elbow MRIs for PDBT reviewed by 4 reviewers. The flexion, abduction, supination view was available for 11 patients. The 4 reviewers had an overall agreement of 51.4%, with a kappa statistic of 0.27, indicating “fair” agreement. Notably, the reviewers only fully agreed on their PDBT grading for 12 patients (24%), and 10 patients received scores across all 3 categories. Among patients with FABS view available, interobserver agreement increased to 62.1%, with a kappa statistic of 0.43, indicating “moderate agreement.” Conclusions: Reliable identification of the degree of tearing in partial distal biceps injuries is difficult based on MRI alone. The presence of an FABS view increased both agreement and reliability between the reviewing surgeons.