Elsevier

Ophthalmology

Volume 114, Issue 8, August 2007, Pages 1448-1452
Ophthalmology

Original Article
Computed Tomography in the Diagnosis of Occult Open-Globe Injuries

Presented in part at: Association for Research in Vision and Ophthalmology Annual Meeting, May 2006, Fort Lauderdale, Florida.
https://doi.org/10.1016/j.ophtha.2006.10.051Get rights and content

Purpose

To determine the radiographic signs present on computed tomography (CT) most suggestive of occult open-globe injury.

Design

Retrospective chart review.

Participants

Forty-eight eyes (of 46 patients), 34 of which were found to have an occult open-globe injury on surgical exploration.

Methods

A retrospective chart review of all eyes of patients 18 years or older undergoing surgical exploration to rule out occult open-globe injury after CT examination at Parkland Memorial Hospital between October, 1998, and September, 2003, was conducted. Patients with obvious corneal or corneoscleral lacerations or with uveal prolapse were excluded. The CT films were obtained and independently reviewed by 3 masked observers (2 neuroradiologists and 1 ophthalmologist).

Main Outcome Measures

Presence of occult open-globe injury with respect to radiographic globe and orbital findings.

Results

The sensitivity of CT for determining occult open-globe injury varied from 56% to 68% between the observers, specificity ranged from 79% to 100%, positive predictive value ranged from 86% to 100%, and negative predictive value ranged from 42% to 50%. Open-globe injuries averaged more CT findings per patient compared with intact globes (P = 0.047). Statistically significant CT findings for occult open-globe injury included any change in globe contour (P = 0.001), obvious volume loss (P = 0.003), an absent or dislocated lens (P = 0.048), vitreous hemorrhage (P = 0.003), and retinal detachment (P = 0.044). Additionally, moderate to severe change in globe contour, obvious volume loss, total vitreous hemorrhage, and absence of lens were seen only in eyes with occult rupture.

Conclusions

Although CT scanning may provide valuable information in patients in whom an occult open-globe injury is suspected, its sensitivity and specificity are inadequate to be relied on fully, and such patients generally should be taken to the operating room for formal surgical evaluation. Significant changes in globe contour or obvious volume loss are strong predictors of globe rupture, and any vitreous hemorrhage should be a concern for occult injury.

Section snippets

Patients and Methods

The study was approved by the Institutional Review Boards of the University of Texas Southwestern Medical Center and Parkland Memorial Hospital. The medical records of all patients taken to the operating room at Parkland Memorial Hospital between October 1, 1998, and September 30, 2003, for known or suspected open-globe injury were reviewed. Patients younger than 18 years and those who did not undergo preoperative CT scanning were excluded, as were patients whose CT films could not be located

Results

Fifty-two eyes of 50 patients underwent surgical exploration to rule out occult open-globe injury. The CT films could not be located for 4 unilateral patients, leaving 48 eyes of 46 patients for analysis. Axial images were obtained for all patients, whereas additional coronal imaging was available for 35 (73%) of the 48 eyes. The mean age of the patients was 36 years (range, 19–84 years), and 38 (83%) were men. Twenty-eight (58%) of the 48 injured eyes were left eyes. Latino persons accounted

Discussion

Computed tomography provides valuable information to the ophthalmologist assessing the traumatized eye. We found that preoperative CT scanning for diagnosis of occult open-globe injuries (by original radiographic interpretation) had sensitivity of 71%, specificity of 76%, and PPV of 89%. Joseph et al5 retrospectively reviewed 200 patients who underwent preoperative CT scanning for eye injuries and found sensitivity, specificity, and PPV of CT in the diagnosis of open-globe injury to be 75%,

References (9)

There are more references available in the full text version of this article.

Cited by (59)

  • High risk and low prevalence diseases: Open globe injury

    2023, American Journal of Emergency Medicine
    Citation Excerpt :

    Another retrospective study found the sensitivity and specificity of CT were 51–77% and 97%, respectively [73]. A third retrospective study of 200 patients presenting from 1989 to 1993 found the sensitivity and specificity of CT to be 75% and 93% [81], while a retrospective study of 46 patients found the sensitivity and specificity of CT were 56–68% and 79–100%, respectively [82]. Based on the data, overall CT sensitivity ranges from 51 to 77% and specificity from 79 to 100% [51,70,73,81,82].

  • Adult male with vision loss

    2017, Visual Journal of Emergency Medicine
View all citing articles on Scopus

Manuscript no. 2006-816.

Supported in part by an unrestricted research grant from Research to Prevent Blindness, Inc., New York, New York.

The authors have no proprietary or financial interest in the material discussed.

View full text