Abstract
Purpose :
Previous studies reported the sensitivity of CT scan for detecting open globe injuries to range from 51% to 75%. We performed a retrospective chart review to determine if any characteristics of the patient or injury affect the likelihood for the open globe injury to be detected on CT scan.
Methods :
Retrospective chart review of 77 eyes of 75 patients ages 18 to 96 with confirmed open globe injuries at UMC New Orleans and Shreveport from 2011 to 2016 who had CT maxillofacial, orbits, or head performed on presentation. Exclusion criteria included open globe secondary to chemical injury or spontaneous PKP dehiscence, age less than 18, or imprisonment at the time of injury. A backwards logistic regression was used to determine whether age, gender, mechanism of injury, type of imaging performed, size of defect, location of defect, uveal loss, vitreous prolapse, hyphema, lens involvement, or presence of a intraocular foreign body significantly impacted the ability of the CT scan to detect the open globe injury.
Results :
The overall rate of CT scans detecting open globe injuries was 74.03%. 16 CT Orbits w/o contrast (20.8%), 52 CT Maxillofacial w/o contrast (67.5%), and 9 CT Head w/o contrast (11.7%) were performed. CT Maxillofacial had a detection rate of 78.4%, CT Orbits had a detection rate of 62.5%, and CT head had a detection rate of 66.7%. The presence of a corneal laceration significantly increased the ability of a CT scan to detect an open globe injury (p = 0. 003). There was a trend towards significance for the presence of a hyphema increasing the likelihood of open globe detection (p= 0.0975). The ability of the CT scan to detect an open globe injury was not dependent on the size of the defect, age of the patient, mechanism of injury, presence of intraocular foreign body, vitreous hemorrhage, uveal loss, or vitreous prolapse.
Conclusions :
CT Maxillofacial without contrast was the most sensitive at detecting open globe injuries, while CT Head without contrast was the least sensitive. The presence of a corneal laceration significantly increased the likelihood of a CT scan detecting the open globe injury. The size of the defect, location of defect, uveal loss, vitreous prolapse, hyphema, lens involvement, and presence of a intraocular foreign body were not found to have a significant impact on the ability of CT scan to detect open globe injuries.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.