Ultrasound for Posterior Ocular Pathology

Podcast

post ocular byte


Test characteristics of point-of-care ultrasonography for the diagnosis of acute posterior ocular pathology

Canadian Journal of Ophthalmology October 2016 - Pubmed Link


Take Home Points

1. In this study, the overall accuracy of ultrasound for diagnosing posterior ocular pathology was not high enough to support use in decision making.

2. Recognize that retinal tears can be misdiagnosed as posterior vitreous detachment on ultrasound.


Background

Missing a retinal tear or retinal detachment can lead to irreversible loss of vision (that sounds bad). Usually these can be diagnosed by a good look at the back of the eye. Unfortunately, often times dilated fundoscopy or ophthalmology consultation is not an option. Ultrasound can assist in these diagnoses. It has been used by ophthalmologists for a while and more recently emergency physicians. What we want to know is how accurate is ultrasound for each of the different pathologies we can diagnose?


Questions

What are the test characteristics of POCUS in diagnosing retinal detachment, retinal tear, posterior vitreous detachment, and vitreous hemorrhage?


Population

Patients presenting to emergency ophthalmology clinic with acute visual complaints


Inclusion

  • ≥ 18 years old

  • Acute flashes/floaters OR new visual field defect of less than 7 days duration


Exclusion

  • Suspected globe rupture

  • Anterior chamber pathology

  • Previous intraocular surgery

  • Previous retinal detachment in symptomatic eye

  • Refused to consent


Who did the scans?

1 Medical student, one of several emergency medicine residents, or one of several attending emergency physicians

Training consisted of 1 hour workshop + reviewing study protocol


Design

Prospective observational convenience sample

Patient identified for potential inclusion by emergency eye clinic staff → informed study sonologist who confirmed criteria met

Sonologist blinded to ophthalmologist exam, but had access to clinical information sufficient to determine inclusion and exclusion criteria

Ultrasounds were compared to exam by ophthalmologist (“experienced vitreoretinal surgeon”) on the same day or next day

Ophthalmologist did not use ultrasound

Subsequent chart review 6 weeks after original encounter to check diagnosis


Intervention

Study participants received:

  1. Point of care ultrasound

  2. Ophthalmologist exam

  3. 6 week follow-up


The Scan

Patient Supine

Copious amount of ultrasound gel on closed eyelid

10 MHz linear transducer, with ocular settings

linear probe image

Here are the study's (paraphrased) descriptions of each Pathology

Posterior Vitreous Detachment - low echogenicity, high mobility, attached to retina

Retinal Tear - echogenic retinal flap or SUSPECTED if echogenic retinal traction

Retinal Detachment - high echogenicity, attached at optic nerve and ora serrata, lower mobility

Vitreous Hemorrhage - low to medium echogenicity, layers posteriorly when supine, highly mobile


Check out clips of these findings and learn how to perform ocular scans for posterior pathology


Results

62 patients enrolled

50 had 6 week follow up

Patients

  • Mean age 60.8

  • 43.6% hypertension

  • 11.3% diabetes

  • 42% myopia


Symptoms:

  • Floaters - 58%

  • Flashes - 33%

  • Veil/curtain - 8%

  • Decreased peripheral vision - 4%

  • Decreased visual acuity - 5%


60/62 (96.8%) had pathology


Primary Outcomes

All pathology (n=60)

88.3% sens (86.8-89.9)

50% spec (2.7-97.3)

Retinal detachment (n=6)

100% sens (53.8-100)

67.9 spec (62.9-67.9)

Retinal detachment or tear (n = 23)

47.8% sens (30.8-62.2)

82.1% spec (72-90.6)

Posterior vitreous detachment (n=47)

80.9% sens (74.7-88.0)

33.3% spec (14-55.7)

Vitreous hemorrhage (n=14)

43.0% sens (21.4-58%)

93.8% spec (87.5-98.2)


Overall - low numbers, wide confidence intervals. It is somewhat surprising that some exams were more sensitive than specific (meaning they must have incorrectly diagnosed or over-called some findings. That is usually less common than missing positive exams in most ultrasound literature.


Other Findings

Mean time to scan was 7.4 minutes (range 4-17.2 minutes)


False Negatives:

  • 10 of the 14 missed retinal tears were superior (many superior-temporal)

  • 13/14 were diagnosed as posterior vitreous detachment (with or without VH)


All 23 cases of retinal tear/detachment required intervention

For 50 patients that had 6 week follow up visits - no change in original diagnosis


Limitations

Sonologists were not fully blinded (had to know that patient met inclusion criteria)

Population was emergency eye clinic - more pathology than usual population (96.8% had pathology). Could this have contributed to more false negatives?

85.5% of exams by same operator - medical student with little prior experience

Small sample, wide confidence intervals

Does not seem consistent with previous results. For example, (mentioned in this article) a recent systematic review and metaanalysis of ED ultrasound for retinal detachment which showed sensitivity of 97-100% and specificity of 83-100% for retinal detachment


Take Home Points

1. In this study, the overall accuracy of ultrasound for diagnosing posterior ocular pathology was not high enough to support use in decision making.

2. Recognize that retinal tears can be misdiagnosed as posterior vitreous detachment on ultrasound.


Our score

3 Probes


Published on 05/08/17 02:00 AM
comments (0)