The diagnostic accuracy of smartphones in the radiographic analysis of ankle fractures

Abstract

Background: Orthopaedic surgeons rely on radiographs to assist in the evaluation and treatment of patients. The technological advances of smartphones have allowed data to be transferred quickly and viewed from remote locations. The diagnostic accuracy of viewing radiographs on smartphones remains unproven. As such, this study aimed to evaluate the ability of smartphones to allow surgeons to correctly diagnose and interpret radiographs of ankle fractures on one of the most popular smartphones, an Apple iPhone 4S (Apple Inc, Cupertino, CA, USA). Methods: Using the smartphone, individual pictures were taken of anteroposterior, mortise, and lateral radiographs of 20 ankle fractures. These images were then evaluated through 10 questions that included acceptable quality, presence and displacement of fibula, medial malleolus, and posterior malleolar fractures, ankle subluxation, syndesmotic disruption, Lauge-Hansen classification, and the need for open reduction and internal fixation. Results were compared to the same radiographs viewed on a picture archiving and communication system (PACS) workstation. Results: Ten orthopaedic surgery residents and five orthopaedic surgery attendings evaluated the radiographs and answered a total of 6000 questions. When comparing the responses between those images viewed on the smartphone versus those viewed on the PACS workstation, 826 differences were found. After analysis of both interobserver and intraobserver reliability, a significant difference was found between viewing modalities, with the smartphone found to be not as reliable as the PACS workstation. However, when looking at the single question of whether or not the fracture would require operative fixation, no significant difference was found between the use of a smartphone or PACS workstation. Conclusions: The smartphone images demonstrated a reliable prediction of the proper treatment for the 20 ankle fractures despite noted differences between the two viewing modalities.