Aim: To determine the incidence of missed femoral head avascular necrosis (AVN) on pelvic computed tomography (CT) performed for clinical indications other than assessment for AVN.
Materials and methods: The study was a Health Insurance Portability and Accountability Act (HIPAA)-compliant retrospective study. The picture archiving and communication system (PACS) database was queried for patients with diagnosis of femoral head AVN on magnetic resonance imaging (MRI), who also underwent pelvic multidetector CT after or <30 days before the MRI examination. The MRI and CT images of 144 hips of 72 patients (39 patients with reported AVN on MRI and 33 age-matched controls; mean age = 60 years, 28 male, 44 female), were reviewed in consensus by two readers in a randomized, blinded manner. Using MRI-proven CT-visible AVN as a reference standard, the incidence of missed AVN on initial CT interpretation was determined.
Results: Readers confirmed AVN in 33 patients on the MRI images. Nine hips with AVN underwent joint replacement of the affected joint(s) prior to subsequent CT and were excluded. Forty-three MRI-proven AVN cases in 28 patients (15 bilateral, 13 unilateral) were available for analysis. The study readers diagnosed 35/43 (81%) MRI-proven AVN cases in 22/28 (79%) patients. Four of the 35 (11%) cases of MRI-proven, CT-visible AVN were prospectively reported in 3/22 (14%) patients at initial clinical interpretation, with a miss rate of 89% per hip and 86% per patient.
Conclusion: Multidetector CT has high accuracy for detection of AVN; however, this is frequently missed as an incidental finding (89% missed in the present study). Assessment for signs of femoral AVN should be part of routine search pattern in interpretation of pelvic CT.
Copyright © 2013 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.