Remote Diagnosis of Retinal Detachment in an Emergency Department Using Nonmydriatic Hybrid Ocular Imaging

Abstract

Background:

Ocular emergencies are commonly evaluated in general emergency departments (ED) where ophthalmologists are rarely available. Nonmydriatic ocular imaging combining color fundus photographs and optical coherence tomography (NMFP-OCT) can help with rapid remote triage by ophthalmologists. We evaluated the rate at which retinal detachments (RDs) can be diagnosed with NMFP-OCT in the ED.

Methods:

Quality improvement project with prospective collection of data on RD patients who had NMFP-OCT obtained by ED staff over 1 year. Photographs were interpreted remotely by ophthalmologists and all patients underwent an in-person ophthalmologic examination in the ED to confirm the presence of a RD.

Results:

A total of 63 eyes (58 patients) had a RD, among which 53 (84.1%) had strong suggestion of RD on ocular imaging (34 [54%] were seen on both color and OCT nerve/macula; 11 [17.5%] were seen on color but missed on OCT; 8 [12.7%] were missed on color but seen on OCT). Ten RDs (15.9%) were missed on both color and OCT because of peripheral location of the RD (4, 40%), vitreous hemorrhage (4, 40%), or poor image quality (2, 20%). A total of 40 out of 58 patients were not seen by an eye care provider prior to reaching our ED and 10 had an inappropriate stroke workup for acute vision loss of presumed vascular origin.

Conclusion:

NMFP-OCT of the posterior pole obtained by ED staff revealed the RD in 84.1% of eyes, allowing for rapid remote triage of patients with visual symptoms and avoiding unnecessary testing when the diagnosis of RD is confirmed.

Introduction

Ocular emergencies comprise about 1.5–3% of all emergency department (ED) diagnoses, with an estimated mean of about 2 million ED visits for ophthalmic conditions per year.^1,2 However, except for rare institutions equipped with a dedicated ophthalmology ED or major academic hospitals with ophthalmology trainee coverage, most EDs do not have easy access to ophthalmology.^3,4 The accurate evaluation of ocular emergencies is difficult in busy EDs because of lack of ophthalmic equipment and trained personnel. Most ED providers are not comfortable with the ocular examination, and examination of the ocular fundus with ophthalmoscopy is rarely performed in the ED.⁵ This leads to diagnostic errors, delayed care, and often inappropriate investigations.^5,6

The burden of ocular emergencies is compounded by heterogenous and disproportionate access to eye care, especially within rural areas and across low income and minority populations. This, in turn, may force patients with acute vision loss to seek care through the closest ED.^7,8 Almost 60% of rural hospitals report the absence of ophthalmology services, requiring transfer of a patient to a distant larger tertiary hospital where an ophthalmologist is available.⁸

This increase in ED visits has resulted in the implementation of strategies to compensate for the dearth of ophthalmic consultants in most EDs. One such method is ocular point-of-care ultrasound (POCUS) to rule-out retinal detachments (RD) or vitreous hemorrhage in patients with acute vision loss.^9

–12 However, POCUS is operator dependent and not reliable for other urgent causes of acute vision loss, such as retinal artery occlusions and optic neuropathies.^9

–12 Ophthalmic telemedicine in ED settings is an alternative to remotely manage emergencies and avoid systematic transfers.^13

–19 However, to be most effective, ophthalmic telemedicine consultations should include remote interpretation of ocular fundus imaging studies such as color fundus photographs and optical coherence tomography (OCT) of the macula and optic nerves.^13,17
–19 Although very few EDs are currently equipped with retinal cameras, studies have shown that implementation of nonmydriatic cameras in the ED is feasible and very useful for the diagnosis of ocular emergencies such as acute vision loss from acute retinal ischemia and optic neuropathies.^5,18,20,21

Since the majority of visually compromising conditions involve the posterior pole of the eye, imaging studies should ideally provide high resolution images of the optic nerves, macula, and vascular arcades. However, RDs are also a common cause of acute vision loss, but often involve the peripheral retina with obscured view when there is associated vitreous hemorrhage. Although the gold standard for the diagnosis of RD is dilated ocular fundus examination by a trained eye care specialist preferably aided by ultra-widefield fundus photography, and possibly complemented by OCT,^22,23 that scenario is very rarely immediately available in EDs. It is unclear how many RDs would be missed by remote interpretation of ED staff-performed ocular imaging studies without ultrawide field ocular imaging and obtained without pharmacologic dilation of the pupils.

Our goal was to assess the utility of a 45-degree hybrid (color photographs and OCT) nonmydriatic camera to detect RDs on ocular imaging obtained by ED personnel in our general ED and interpreted remotely by ophthalmologists.

Methods

This study was part of a large prospective quality-improvement project approved by our institutional review board; patients’ data were deidentified, and consent was not required by the IRB. We prospectively collected data and reviewed all imaging studies of patients diagnosed with a RD in our general ED. Only patients who underwent nonmydriatic ocular imaging studies by ED staff in our general ED were included.

OCULAR FUNDUS IMAGING

In June 2023, we installed in our general ED a nonmydriatic hybrid true color 45-degree field ocular fundus camera combined with OCT (NMFP-OCT) of the peripapillary retinal nerve fiber layer and of the macula (Topcon Maestro2—Japan). This fully automated camera is directly connected to our electronic medical record, allowing for immediate remote review of medical records and ocular imaging studies by ophthalmologists. Patients presenting to our ED with any eye complaints received standard ocular imaging studies (both 45-degree color photographs and OCT) without pharmacologic dilation of the pupils, obtained by trained ED staff. Although this camera is capable of taking photographs of the retinal periphery in the manual mode, the ED staff only used the standard protocol which is faster, entirely automated, and images the posterior pole, including the optic disc, the macula and the vascular arcades, but does not show the peripheral retina.

PATIENT POPULATION AND DIAGNOSIS OF RD

All patients with visual complaints or a known diagnosis of RD prior to arrival to our ED also underwent an in-person ocular examination with dilated fundus examination by the ophthalmologist on call (in addition to NMFP-OCT); all patients with confirmed or suspected RD were then seen in-person by a retina specialist to further characterize the RD and offer treatment options. Retinal detachments were classified as “macula-on” (involving the peripheral retina with normal visual acuity) or “macula-off” (involving the posterior pole of the eye with decreased visual acuity). Only patients with confirmed RD by a retina specialist were included in this project. Basic demographic information, including age, self-reported race/ethnicity, and eye affected by the RD, were collected.

REVIEW OF OCULAR IMAGING STUDIES

All imaging studies of included patients were reviewed by three ophthalmologists (E.B.S., G.A., V.B.) to determine whether the RD could be visualized on (1) both 45-degree true color fundus photographs and OCT; (2) color photographs alone (not seen on OCT because of peripheral detachment outside of the posterior pole or poor quality of OCT); (3) OCT alone (not seen on color fundus photographs because of peripheral detachment outside of the posterior pole, vitreous hemorrhage obscuring view, or poor quality of color photographs); and (4) not seen on either modality of imaging (missed by imaging due to location of RD or poor quality of both imaging studies).

QUALITY OF OCULAR IMAGING STUDIES

Each color fundus photograph was also assigned a grade based on the quality of the photographs (on a scale of 1 to 5, where 1 = inadequate to diagnose any pathology and 5 = adequate to diagnose a retinal detachment). The presence of light, dust, and eyelash artifacts was noted.

STATISTICAL ANALYSIS

Descriptive statistics were used to summarize the clinical features of included patients. We calculated the sensitivity with 95% confidence intervals (CIs) for color photography and OCT compared with the ophthalmologists’ final diagnosis of RD. These measures were calculated in the standard manner with continuity correction.

Results

We included 63 eyes of 58 patients who received NMFP-OCT in the ED and were diagnosed with a RD. Men comprised the majority of patients (33, 56.9%); 30 patients (51.7%) were White, 18 (31%) Black, 3 (5.2%) Asian, and 7 (12.1%) were categorized other. The age at presentation ranged from 21 to 94 years, with a median of 56 years.

Of the 63 eyes’ RDs, 51 (81.0%) were rhegmatogenous, 11 (17.5%) were tractional, and 3 (4.8%) were serous. The RDs were categorized as “macula off” in 41 eyes (65.1%). The remaining 22 eyes (34.9%) were limited to the peripheral retina (“macula-on RD”).

Of 58 patients, 40 were not seen by an eye care provider prior to reaching our ED, and 10 had a stroke workup at the outside ED (including brain MRI and MRA in 1, and head CT and CT/CTA head and neck in 9 for acute vision loss of presumed vascular origin).

RESULTS OF OCULAR IMAGING (Table 1)

In 53 eyes (84.1%), ocular imaging strongly suggested the presence of a RD. Among all 63 eyes, 34 RDs (54%) were seen on both color photographs and OCT; 11 RDs (17.5%) were seen on color photographs, but not on OCT; 8 RDs (12.7%) were seen on OCT, but not on color photographs; 10 RDs (15.9%) were not seen on both color photographs and OCT because of lack of involvement of the posterior pole (4, 40%), vitreous hemorrhage leading to poor quality of photographs and lack of signal on OCT (4, 40%), or poor quality of all imaging (2, 20%), resulting in poor quality color photographs and lack of signal on OCT. Of the 10 cases of RD missed on both color photographs and OCT, 4 were macula-off and 6 were macula-on. Similar as with color photographs, macula-on RDs were more difficult to visualize on OCT, with separation of the retina typically existing at the edge of the OCT and requiring careful scrolling through the entire OCT to be seen.

Overall, color fundus photography alone had a sensitivity of 71.4% (95% CI: 60.3 − 82.6%) for detecting a RD, while OCT alone had a sensitivity of 66.7% (95% CI: 55.0 − 78.3%) (p = 0.56). Used together, color fundus photography and OCT had a sensitivity of 84.1% (95% CI: 75.1 − 93.2%).

Table 1.

Results of Ocular Imaging Studies (45-Degree Color Fundus Photographs and OCT Optic Nerve and Macula) Obtained in Our Emergency Department without Pharmacologic Dilation of the Pupils in 63 Eyes with a Retinal Detachment

SEEN ON COLOR PHOTOGRAPH	SEEN ON OCT	# EYES	%	MACULA OFF (n, %)		MACULA ON (n, %)		MEDIAN COLOR PHOTOGRAPH QUALITY
YES	YES	34	54.0	25	73.5	9	26.5	5.0
Yes	No	11	17.5	8	72.7	3	27.3	3.0
No	Yes	8	12.7	4	50.0	4	50.0	2.0
No	No	10	15.9	4	40.0	6	60.0	1.0
Total:		63		41		22

Photography quality grading scale: 1 = inadequate to diagnose any pathology and 5 = adequate to diagnose a retinal detachment.

OCT, optical coherence tomography; %, percentage.

QUALITY OF OCULAR IMAGING

The median quality of all color fundus photographs was 4. For RDs seen on color photographs, the median grading was 5, while the quality decreased to 1.5 for images where a RD was not seen. Although the RD itself was not always clearly visualized on color photographs, signs suggesting a RD on nonmydriatic fundus photographs included loss of choroidal vascularization, wrinkles, or haziness of the retina and vessels. In patients with a macula-on RD, abnormalities were often subtle and seen only at the edge of the picture (Fig. 1). On OCT, a RD was confirmed by separation of the retina from the retinal pigment epithelial layer. The missed RDs on OCT were difficult to visualize due to focal loss of OCT signal and “static” appearance on OCT which occurred more commonly in patients with associated vitreous opacities from blood or pigment which can be present in RDs.

Fig. 1.

True-color 45-degree nonmydriatic fundus photographs of eyes with retinal detachments and corresponding OCT cuts obtained on the Maestro2 camera (Topcon, Japan). (a). rhegmatogenous macula-off RD seen on both fundus photograph (loss of choroidal vasculature across macula and inferotemporal to optic nerve) and OCT, (b). rhegmatogenous macula-on RD seen on both fundus photograph (superior bullous RD) and OCT, (c). rhegmatogenous macula-off RD visualized on fundus photograph, but not on OCT, (d). tractional and rhegmatogenous macula-on RD seen on OCT, but not on fundus photograph. (e). peripheral rhegmatogenous macula-on RD not seen on fundus photograph or OCT, (f). tractional macula-on RD not clearly seen on fundus photograph or OCT in the setting of a dense vitreous hemorrhage. OCT, optical coherence tomography; RD, retinal detachment.

Discussion

Ophthalmologists were able to detect a RD remotely from our general ED in the majority of posterior pole imaging obtained by ED staff without pharmacologic dilation of the pupils. These results suggest that nonmydriatic retinal cameras combining color photographs with OCT could be routinely used in general EDs to relay digital images to ophthalmologists and aid in teleophthalmology consultations and appropriate immediate triage. Of the 63 RDs, only 10 (15.9%) were missed on both color photographs and OCT due to peripheral RD not involving the posterior pole in 4 eyes, vitreous hemorrhage obscuring view in 4 eyes, and poor image quality in 2 eyes. This poor quality would have prompted an in-person ophthalmology consultation.

These results are promising. Our ED staff was instructed to always use the preprogrammed default camera protocol which is entirely automated, but has only a 45-degree field of view. These imaging studies were able to detect macula-on RDs, despite limited peripheral visualization. The standard 45-degree method of ocular imaging has a high sensitivity for the diagnosis of acute pathology involving the posterior pole, including retinal artery occlusions, maculopathies, optic neuropathies, and vitreous hemorrhage, and is ideal for emergency settings outside of eye clinics.^5,17,18 However, the gold standard for evaluating the peripheral retina is ultra-wide field imaging,²² and we did not expect to be able to diagnose this many RDs using our default protocol that images the posterior pole only. Although patients with a macula-on RD may have good visual acuity when there is no vitreous hemorrhage, they typically complain of phosphenes, floaters, or peripheral shadows, which should raise concern for peripheral pathology. These symptoms should always trigger an in-person dilated ocular funduscopic examination, or at least POCUS, in order to detect a peripheral RD when 45-degree imaging is unrevealing.¹¹

Combining OCT and color photographs increased the sensitivity of detecting a RD. In 8 eyes (12.7%) with normal or poor-quality color photographs, the RD was solely detected on the OCT; alternatively, in 11 eyes (17.5%) with unreliable OCTs, a RD was seen or suspected on color fundus photography. This supports the utilization of hybrid cameras in EDs rather than color fundus photography or OCT alone.

Importantly, 40 of the 58 patients diagnosed with a RD in our ED had not had an ophthalmic evaluation prior to having NMFP-OCT. A total of 10 of these 40 patients received an unnecessary empiric stroke workup triggered by the chief complaint of “acute vision loss,” presumed of vascular origin, prior to being transferred to our ED where the correct diagnosis of RD was made.

Our study has several limitations. This study was part of a large quality improvement project performed in one general ED at our institution, and the results may not be generalizable to all EDs. We only included eyes that were imaged while in the ED and it is likely that some patients did not receive ocular imaging, especially those who were transferred from another institution with a diagnosis of RD already established. For the purpose of this study, we reviewed all imaging studies knowing that the patient had a RD, and it is possible that this led us to detect subtle findings that would have been missed otherwise. However, this highlights the importance of clinical information when interpreting ocular imaging studies remotely, as knowing what to look for influences the strategies used to interpret ocular fundus photographs and OCT, similar to what is done by radiologists who do not have direct access to patients when they interpret radiologic studies.

Obviously, the quality of the imaging studies was not always optimal, especially when patients had vitreous opacities such as blood or pigment, which are common with RDs or in patients with severe vision loss and difficulty focusing on the camera target. Very small pupils may also contribute to the poor quality of some nonmydriatic fundus photographs and OCTs. We did not analyze these parameters in this real-life study where NMFP-OCT was obtained in the ED by ED personnel, knowing that any poor quality photographs in patients with visual complaints would prompt an in-person ophthalmologic consultation. However, we could most often detect abnormalities suggesting a RD on the color fundus photographs or the OCT even in patients with poor quality images.

The purpose of this study was not to replace ophthalmologic examination by remote interpretation of ocular imaging. Rather, our results showed that the utilization of ocular imaging can be a valuable diagnostic aid. Despite the better sensitivity of ultrawide-field cameras to detect peripheral RDs, for our ED we chose a 45-degree camera because most ocular pathology causing vision loss involves the optic nerves and macula. It is important to note that the camera we used has the capability to obtain images of the retinal periphery. This function was not used by ED staff because it requires switching the camera to manual mode and takes more time and more technical expertise.

Previous studies have shown that the best way to implement new technology that changes ED flow is to keep protocols short, simple, and accessible.⁵ The next step will be to teach ED personnel to use the peripheral retina mode in selected cases where a RD is suspected from the clinical complaint and standard automated imaging of the posterior pole is either normal or nonconclusive. POCUS was not used for these patients, and therefore we cannot compare the sensitivity of both techniques or evaluate how these imaging technologies may complement each other. It is likely that the use of POCUS would be useful in patients suspected of RD who have either normal or unreliable ocular imaging studies.

Conclusions

Our findings indicate that ophthalmologists can accurately detect RDs on the majority of ocular imaging studies obtained without pharmacological dilation of the pupils by ED staff in a general ED using a 45-degree hybrid true color/OCT camera. The two techniques (true color photographs and OCT) act synergistically, suggesting that such hybrid cameras should be implemented in EDs in order to facilitate teleophthalmology consultations. Future directions will include performing cost-analysis studies to confirm that implementation of NMFP-OCT in general EDs not only avoids expensive unnecessary workup, but may accelerate patient care and reduce length of ED stay by facilitating remote triage by ophthalmologists, especially in remote areas.

Footnotes

Authors’ Contributions

All authors contributed equally to the study and the article.

Disclosure Statement

The authors do not have any competing financial interests in relation to the work described. V. Biousse is a consultant for GenSight Biologics, Neurophoenix, and Topcon, and receives research support from GenSight Biologics. N. J. Newman is a consultant for GenSight Biologics, Santhera/Chiesi, Stoke, Neuroph, and Neurophoenix, and receives research support from GenSight Biologics and Santhera/Chiesi. Sruthi Arepalli is consultant for Eyepoint, Alimera Sciences, and AbbVie.

Funding Information

This study was not funded. V Biousse, NJ Newman and S Arepalli are supported by the National Institutes of Health’s National Eye Institute core Grant P30-EY06360 (Department of Ophthalmology, Emory University School of Medicine) and by a departmental grant from Research to Prevent Blindness (New York, NY).

References

Channa

, Zafar

, Canner

, et al. Epidemiology of eye-related emergency department visits. JAMA Ophthalmol, 2016; 134(3):312–319.

Vaziri

, Schwartz

, Flynn

, et al. Eye-related emergency department visits in the United States, 2010. Ophthalmology, 2016; 123(4):917–919.

Hall

, Jeng-Miller

, Gardiner

, et al. Utilization trends of an ophthalmology-specific emergency department: The Massachusetts eye and ear experience. Digit J Ophthalmol, 2021; 26(4):31–35.

Sridhar

, Isom

, Schiffman

, et al. Utilization of ophthalmology-specific emergency department services. Semin Ophthalmol, 2018; 33(2):185–190.

Biousse

, Bruce

, Newman

. Ophthalmoscopy in the 21st century: The 2017 H. Houston Merritt lecture. Neurology, 2018; 90(4):167–175.

Bruce

, Cockerham

. Bridging the diagnostic gap between emergency medicine and neuro-ophthalmic disorders with technology and telehealth. Med Res Arch, 2024; 12(2).

Soares

, Mokhashi

, Sharpe

, et al. Patient accessibility to eye care in the United States. Ophthalmology, 2023; 130(4):354–360.

Brant

, Kolomeyer

, Goldberg

, et al. United states population disparities in ophthalmic care: Blindness and visual impairment in the iris® registry (Intelligent Research in Sight). Ophthalmology, 2023; 130(11):1121–1137.

Chirurgi

, Duvvi

, Hussain

, et al. Monocular painless vision loss in the emergency department: Multipurpose utility of point-of-care ultrasound. J Emerg Med, 2024; 66(2):192–196.

10.

Propst

, Kirschner

, Strachan

, et al. Ocular point-of-care ultrasonography to diagnose posterior chamber abnormalities: A systematic review and meta-analysis. JAMA Netw Open, 2020; 3(2):e1921460.

11.

Lahham

, Shniter

, Thompson

, et al. Point-of-care ultrasonography in the diagnosis of retinal detachment, vitreous hemorrhage, and vitreous detachment in the emergency department. JAMA Netw Open, 2019; 2(4):e192162.

12.

Gottlieb

, Holladay

, Peksa

. Point-of-care ocular ultrasound for the diagnosis of retinal detachment: A systematic review and meta-analysis. Acad Emerg Med, 2019; 26(8):931–939.

13.

Woodward

, Bavinger

, Amin

, et al. Telemedicine for ophthalmic consultation services: Use of a portable device and layering information for graders. J Telemed Telecare, 2017; 23(2):365–370.

14.

De Arrigunaga

, Aziz

, Lorch

, et al. A review of ophthalmic telemedicine for emergency department settings. Semin Ophthalmol, 2022; 37(1):83–90.

15.

Tauber

, Ayoub

, Shah

, et al. Assessing the demand for teleophthalmology in Florida emergency departments. Telemed J E Health, 2020; 26(12):1500–1506.

16.

Shah

, Fliotsos

, Alaqeel

, et al. Use of teleophthalmology for evaluation of ophthalmic emergencies by ophthalmology residents in the emergency department. Telemed J E Health, 2022; 28(6):858–864.

17.

Dumitrascu

, English

, Alhayek

, et al. Telemedicine for acute monocular visual loss: A retrospective large telestroke network experience. Telemed J E Health, 2023; 29(11):1738–1743.

18.

Lema

GMC

, De Leacy

, Fara

, et al. A remote consult retinal artery occlusion diagnostic protocol. Ophthalmology, 2024; 131(6):724–730.

19.

Artymowicz

, Douglas

, Cockerham

. The role of ophthalmology in tele-stroke consults for triaging acute vision loss. Open Access Emerg Med, 2024; 16:45–56.

20.

Bruce

, Lamirel

, Wright

, et al. Nonmydriatic ocular fundus photography in the emergency department. N Engl J Med, 2011; 364(4):387–389.

21.

Bruce

, Lamirel

, Biousse

, et al. Feasibility of nonmydriatic ocular fundus photography in the emergency department: Phase I of the FOTO-ED study. Acad Emerg Med, 2011; 18(9):928–933.

22.

Midena

, Marchione

, Di Giorgio

, et al. Ultra-wide-field fundus photography compared to ophthalmoscopy in diagnosing and classifying major retinal diseases. Sci Rep, 2022; 12(1):19287.

23.

Nakanishi

, Hangai

, Unoki

, et al. Spectral-domain optical coherence tomography imaging of the detached macula in rhegmatogenous retinal detachment. Retina, 2009; 29(2):232–242.