Accuracy Of Computed Tomography For Detection Of Wooden Foreign Bodies In Dogs

in Conference Proceedings. American College of Veterinary Radiology 2016.

Introduction/Purpose: Although imaging is unnecessary for successful management of many dogs with acute stick injuries, imaging is indicated in dogs with persistent or recurrent signs of localized infection, unexplained pleuritis or peritonitis, suspected abscess or a draining sinus because of the possibility of retained foreign material. The aims of the present study were to describe computed tomographic (CT) signs associated with wooden foreign bodies in a series of dogs, and to estimate the accuracy of CT for detection of wooden foreign bodies.

Methods: For this retrospective cross-sectional study, medical records in the period 2005-2015 were searched for dogs that had been referred with a history of suspected stick injury and CT of the affected body part, or possible wooden foreign object reported on CT, and surgical exploration during the same period of hospitalization.

Results: Records were found of 72 dogs that satisfied the inclusion criteria. Duration of clinical signs was acute (<7 days) in 48 (67%) dogs and chronic (7 or more days) in 24 (33%) dogs. Fragments of wood were removed surgically from 43 (60%) dogs, including 9 dogs from which a piece of wood had been removed prior to referral. Wood was removed surgically from the cervical region in 18 (42%) dogs, abdomen in 10 (23%), pharynx in 7 (16%), thorax in 7 (16%), and nasal cavity in 1 (2%). Description of the wood removed was recorded in 59 cases. Wood was identified as a piece of a branch of a tree or shrub in 46 (78%) instances, a kebab stick in 9 (15%), piece of bamboo garden cane in 2 (3%), and cocktail stick in 2 (3%). Sensitivity of CT for wooden foreign bodies was 79% (95% confidence interval 65-89%), specificity 93% (78-98%), positive likelihood ratio 11.5 (2.9-44.1), and negative likelihood ratio 0.23 (0.13-0.41). In CT images, wooden foreign bodies appeared predominantly rectangular or linear, with median length 48mm (range 2-270mm), median thickness 3mm (1-22mm), and median attenuation 111HU (-344-640HU). A CT finding of gas in soft tissues was significantly associated with acute cases, whereas suspected foreign material, cavitary lesions, fat stranding, and periosteal reaction on adjacent bones were significantly associated with chronic cases.

Discussion/Conclusions: CT has moderate sensitivity for wooden foreign bodies, which can have a wide range of attenuation. Fragments of wood are liable to be undetected by CT when they have attenuation similar to adjacent gas or soft tissues. Use of a lung window is recommended when searching for wood. Long thin pieces of wood, such as kebab sticks, can be difficult to recognize unless the image plane is oriented parallel to their long axis.