Referral Patterns of Central Retinal Artery Occlusion to an Academic Center Affiliated With a Stroke Center

Central retinal artery occlusion (CRAO) is a medical emergency, and patients who present acutely should be immediately referred to the nearest stroke center. We evaluated practice patterns for CRAO management at one academic center over the last decade.

Original Contribution Section Editors: Clare Fraser, MD Susan Mollan, MD Referral Patterns of Central Retinal Artery Occlusion to an Academic Center Affiliated With a Stroke Center Alexis M. Flowers, MD, Wesley Chan, MD, MSc, Benjamin I. Meyer, MD, Beau B. Bruce, MD, PhD, Nancy J. Newman, MD, Valérie Biousse, MD Background: Central retinal artery occlusion (CRAO) is a medical emergency, and patients who present acutely should be immediately referred to the nearest stroke center. We evaluated practice patterns for CRAO management at one academic center over the last decade. Methods: This was a retrospective study on all adult patients diagnosed with a CRAO seen at one tertiary hospital and outpatient clinic affiliated with a comprehensive stroke center (“our institution”) from 2010 to 2020. Our electronic medical records were searched for CRAO diagnoses, and patient medical records were reviewed. The exclusion criteria were incorrect diagnosis, unclear diagnosis, historical CRAO, or satellite clinic location. Demographics, distance and time to presentation to our institution, number and type of prior providers seen, diagnostic tests performed, and treatments provided were collected. Summary statistics of median, mean, and frequency were calculated and reported with measures of variance (interquartile range [IQR], ranges). F, Tukey, and Fisher exact tests were used for comparisons. Results: We included 181 patients with a diagnosis of CRAO (80 [44.2%] women; median age 69 years [range 20– 101]). The median distance from patient’s home to our institution was 27.8 miles (IQR 15.5–57.4; range 2.4– 930). The median time from visual loss to presentation at our institution was 144 hours (IQR 23–442 hours, range 0.5–2,920) from 2010 to 2013, 72 hours (IQR 10.5–372 hours, range 0–13,140) from 2014 to 2016, and 48 hours (IQR 7–180 hours, range 0–8,030) from 2017 to 2020 (P = 0.07). 91/181 (50%) patients presented to an outpatient provider. 73/181 (40%) presented to an emergency Departments of Ophthalmology (AMF, WC, BIM, BBB, NJN, VB), Neurology (BBB, NJN, VB), Epidemiology (BBB), and Neurological Surgery (NJN), Emory University, Atlanta, Georgia. V. Biousse and N. J. Newman are supported in part by NIH/NEI core grant P30-EY06360 (Department of Ophthalmology, Emory University School of Medicine) and by NIH/NINDS (RO1NSO89694). Presented at the North American Neuro-Ophthalmology Society (NANOS) meeting, February 20, 2021, virtual. N. J. Newman is a consultant for GenSight, Santhera, Neurophoenix, and Stealth. V. Biousse is a consultant for GenSight and Neurophoenix. The remaining authors report no conflicts of interest. Address correspondence to Valérie Biousse, MD, Department of Ophthalmology, Emory University School of Medicine, 1365-B Clifton Road, Atlanta, GA 30322; E-mail: [e-mail] 480 department. Eighty-six percent presented within 1 week of visual loss onset, and rates of comprehensive inpatient evaluation for acute CRAO improved from 44% in 2010– 2013 to 82% in 2017–2020 (P , 0.01). Conclusions: Patients with CRAO often present late and only after evaluation by multiple outpatient providers. Improvement has occurred over the past decade, but delays underscore the barriers to performing clinical trials evaluating very acute treatments for CRAO. Educational interventions for healthcare providers and patients are necessary. Journal of Neuro-Ophthalmology 2021;41:480–487 doi: 10.1097/WNO.0000000000001409 © 2021 by North American Neuro-Ophthalmology Society A cute central retinal artery occlusion (CRAO) is a medical emergency defined as a central nervous system infarction (stroke) per the American Heart Association (AHA) and American Stroke Association (ASA) (1). The 2016 and 2020 AAO Preferred Practice Pattern on Retinal and Ophthalmic Artery Occlusions emphasized that patients who present acutely with this diagnosis should be referred immediately to the nearest stroke center (2,3). These important publications highlight the paradigm shift in the combined fields of stroke neurology and ophthalmology in which acute retinal ischemia is now defined as a stroke equivalent and should be managed as such. Several surveys previously assessed the practice patterns of acute CRAO by neurologists and ophthalmologists. In 2009, Atkins et al (4) reported that only 35% of ophthalmologists referred patients with an acute CRAO to a local emergency department (ED) for immediate evaluation. In 2017, Abel et al (5) found that only 46% of neurologists and 8% of retinologists pursued inpatient evaluation for CRAOs occurring within 48 hours. Youn et al (6) reported in 2018 that, again, only 35% of ophthalmologists referred patients to an ED, highlighting the lack of improvement compared with the 2009 survey. However, early diagnosis and timely presentation inherently depends on both public Flowers et al: J Neuro-Ophthalmol 2021; 41: 480-487 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution awareness (7,8) and appropriate infrastructure (9), in addition to the practice management decisions of medical providers. The Balance, Eyes, Face, Arm, Speech, and Time (BE-FAST) stroke campaign (with “E” for eyes) was recently publicized with the goal of educating the general public on immediate recognition of a potential stroke (CRAO) and the need to call for emergency services (10). Hospitals proficient in stroke management in the United States can be certified by the joint commission and other governing bodies as stroke centers (11). There are analogous certification programs in other regions. Studies have shown that patient outcomes at stroke centers are better compared with hospitals that are not certified (11). At stroke centers, a complex algorithm of interconnected services provides the most efficient, effective, and advanced care. Stroke medicine, including the evaluation and treatment of patients with CRAO, is a constantly evolving science to which stroke centers adapt and avoid prevalent misconceptions such as that “normal” carotid ultrasounds and echocardiograms rule out embolic sources (12). The objective of this study is to analyze referral patterns over the past decade for patients with CRAO at one tertiary care center affiliated with a comprehensive stroke center. METHODS This was a retrospective study of all patients with CRAO who had presented to our tertiary hospital, which includes a comprehensive stroke center and an affiliated outpatient clinic (“our institution”) over the past decade. This study was approved by the Emory Institutional Review Board and adhered to the Declaration of Helsinki. Patient Selection Our 2 electronic medical records (one inpatient and one outpatient system) were searched for the CRAO diagnosis in the medical records of all patients seen between January 1, 2010, and July 31, 2020. We included all adult patients who were referred to our institution for the diagnosis of CRAO or were diagnosed with a CRAO by us, regardless of the time elapsed since visual loss. Exclusion criteria included a historical diagnosis of CRAO (when the patient had a history of CRAO in their medical record but was not evaluated at our institution specifically for a CRAO), age less than 18 years old, or satellite clinic location. Data Collection We collected demographic information (age, sex, ethnicity, and body mass index), distance traveled to our institution using the patient’s home zip code, number and type of providers seen, examination findings, all diagnostic tests performed, and any treatments provided. The disposition of the patient was classified as 1) inpatient, comprehensive stroke evaluation, 2) outpatient comprehensive stroke evaluation by a neurologist, or 3) no evaluation, incomplete Flowers et al: J Neuro-Ophthalmol 2021; 41: 480-487 evaluation, or evaluation with a provider other than a neurologist (e.g., primary care provider). The time from visual loss onset to presentation to any provider, to the ED, and to our institution was recorded. Statistical Analysis Summary statistics of median, mean, and frequency were calculated and reported with measures of variance (interquartile range [IQR], ranges). F, Tukey, and Fisher exact tests were used for comparisons. RESULTS One hundred and eighty-one patients met our inclusion criteria. Of the 181 patients, 80 (44.2%) were female and the median age was 69 years (mean 67, range 20–101). For data available on 173 patients, 87 (50.3%) were identified as White, 78 (45.1%) as Black or African American, 6 (3.5%) as Asian, and 2 (1.2%) as Hispanic. For data available for 162 patients, the median Body Mass Index was 28 kg/m2 (mean 29.1, range 17.6–67.5). Distance Traveled Seven patients (7/181, 3.9%) were already admitted to our institution at the time of visual loss; thus, their distance traveled was 0 miles. The median distance from patients’ home zip code to our institution for the remaining 174 patients was 27.8 miles (IQR: 15.5–57.4; range 2.4–930), 30 (17.2%) lived within 10 miles, 123 (70.7%) within 50 miles, and 157 (90.2%) within 100 miles. Seventeen patients (9.8%) traveled 100 miles or more to our institution. Although patients who first presented to a local ED vs our institution’s ED lived further away, this was not statistically significant (P = 0.43) (Table 1). This was also true for patients who presented to a local outpatient eye care provider compared with an outpatient eye care provider at our institution (P= 0.63) (Table 1). Patients who lived within 12 miles of our institution presented faster to their first provider (any provider) compared with those that lived further away (median 5 hours within 12 miles [IQR 1–40 hours] vs median 24 hours outside of 12 miles [IQR 10– 72 hours], P , 0.001). Time to Presentation Data were available for the time to presentation from visual loss to first provider seen for 176 patients. The median time from visual loss to presentation to any provider was 24 hours (IQR: 4–51; range 0 hours–1.5 years). The time to presentation from visual loss to the first provider seen for 2010–2013 compared with 2014–2016 and 2017–2020 was not statistically significant (P = 0.46) (Table 2). The median time from visual loss to presentation to our institution for all 181 patients was 72 hours (IQR: 10.5–336, range 0 hours–1.5 years). The time to presentation to our institution from 2010 to 2013 compared with 2014–2016 481 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution TABLE 1. Distances traveled by patients with CRAO Location of the First Provider Seen Local ED Our institution’s ED Outpatient eye care provider Our institution’s outpatient eye care provider Median Distance From Home Zip Code to Our Institution (IQR1–IQR3) [Range] Number (%) 37/181 36/181 63/181 28/181 (20.4) (19.8) (34.8) (15.4) 38.9 18.5 40.5 21.1 miles miles miles miles (24.8–72.9) (5.9–25.32) (18.0–69.8) (13.0–30.7) [5.7–221] [2.7–129] [2.8–352] [2.4–199] CRAO, central retinal artery occlusion; ED, emergency department; IQR, interquartile range. and 2017–2020 was not significantly different but did trend toward faster presentation times (P = 0.066) (Table 2). Among the 132 patients who presented to an ED, the time to presentation to the ED was available for 125 patients. The time to presentation to the ED did not improve over time (P = 0.2) (Table 2). The time to presentation to a local ED (median 24 hours, IQR: 6.25–36 hours, range 0.5–504 hours) compared with our institution’s ED (median 22.5 hours, IQR: 4.85–96 hours, range 0.33–840 hours) did not differ statistically (P = 0.6). Neither the time to presentation to an outside ED or to our institution’s ED changed over time from 2010 to 2013 compared with 2014–2016 and 2017–2020 (all P . 0.4). Patients with an ocular history (n = 89) presented slower to the first provider, to the ED, and to our institution than those without an ocular history (P ,0.001) (Table 3). Patients with a history of stroke (n = 28) presented more slowly to the first provider than those without a history of stroke (P = 0.04) (Table 3). No statistical difference for the time of presentation to our institution was observed for those with a history of stroke vs those without (P = 0.48) (Table 3). Patients on antithrombotics (n = 52) did not present faster than those patients not on these medications (P = 0.8) (Table 3). Referral Patterns Table 4 lists the number of providers seen for visual loss before presenting to our institution. The change over the past 10 years was not statistically significant, but there was a trend toward more patients presenting directly to our institution (P = 0.13). The maximum number of providers seen by any one patient before presenting to our institution was 5. Of the total 181 patients, 91 patients (50.2%) presented first to an outpatient eye care provider; 73 patients (40.3%) presented to an ED, 9 patients were already admitted to a hospital at the time of visual loss, and 8 patients (4.4%) presented to an outpatient provider other than an eye care provider. Among the 99 total patients who presented to outpatient providers, 28 (28.3%) presented to a provider at our institution (all of whom were eye care providers) and 71 (71.7%) presented to a provider outside of our institution (63 [88.7%] of whom were eye care providers) (Fig. 1). Among the 73 total patients who presented first to an ED, 37 (50.6%) presented first to a local ED and 36 (49.3%) presented first to the ED at our institution. Overall, a total of 51 of the 181 patients (28.2%) were evaluated at an ED before being seen at our institution and a total of 83 (45.9%) patients were evaluated at our institution’s ED at some point. Seven patients developed a CRAO while admitted at our institution (admitted for the following reasons: atrial fibrillation, endocarditis, myocardial infarction, sepsis, middle cerebral artery stenting procedure, encephalopathy, and aortic valve replacement). Two patients were already admitted to an outside hospital at the time of their CRAO TABLE 2. Presentation times from symptom onset to the first provider seen, to the ED, and to our institution from 2010 to 2020 Number* Time to first provider Time to ED Time to our institution 176/ 181 125/ 132 181/ 181 2010–2013 (n = 44)† 24 h (10–72) [0–672] 24 h (10–72) [0.5–840] 2014–2016 (n = 47)† 2017–2020 (n = 87)† 24 h (4–120) [0–13,140] 24 h (5–48) [0–1,344] 11.5 h (3–48) [1–416] F Test‡ P = 0.5 24.5 h (6.7–78) [0.33–504]P = 0.2 144 h (23–442) [0.5–2,920] 72 h (10.5–372) [0–13,140] 48 h (7–180) [0–8,030] P = 0.07 *Number represents the number of patients in each row for whom there was sufficient data for each category. † Results presented as: median (IQR1-IQR3) [range]. ‡ Assesing any differences among 2010–2013, 2014–2016, and 2017–2020 for each row. ED, emergency department; IQR, interquartile range. 482 Flowers et al: J Neuro-Ophthalmol 2021; 41: 480-487 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution TABLE 3. Time to presentation to various providers in the subgroups of CRAO patients with and without an ocular history, a history of stroke, or already receiving an antithrombotic treatment Category + Ocular history 2 Ocular history + History of stroke 2 History of stroke + Antithrombotics 2 Antithrombotics Number* 89 92 28 153 52 129 Time to the First provider† 48 10 32 24 24 24 h h h h h h (2†–144) [0–1,344] (2–14) [0–13,140] (15–120) [0–504] (3–48) [0–1,344] (3–72) [0–672] (4.8–48) [0–13,140] Time to ED† 48 10.7 120 23 24 24 h h h h h h (48–504) [0.5–840] (2.6–24) [0.33–504] (12–180) [0.5–504] (4.6–48) [0.33–840] (4.5–84) [0.5–840] (5–49.5) [0.33–504] Time to Our institution† 144 23.5 108 48 96 48 h h h h h h (48–504) [0–5,040] (3–144) [0–13,140] (24–240) [0–672] (10–336) [0–5,040] (12.8–546) [0–2,920] (10–240) [0–13,140] *Number represents the number of patients in each row for whom there was sufficient data for each category. † Results presented as: median [QR] (range). CRAO, central retinal artery occlusion; ED, emergency department; IQR, interquartile range. (one patient experienced visual loss immediately after cardiac catheterization for a myocardial infarction, and the other patient noted visual loss immediately after a carotid angiogram). Excluding the 9 patients already admitted at the time of visual loss, the first provider seen was most often the ED (73/172, 42.4%) followed by comprehensive ophthalmology (54/172, 31.4%), optometry (16/172, 9.3%), and then retina specialists (16/172, 9.3%). The remaining providers were cornea specialists (4), glaucoma specialists (1), primary care providers (3), urgent care centers (3), neurologists (1), and otolaryngologists (1). Referrals to our institution (excluding the 7 patients already admitted to our institution) came from multiple sources, including comprehensive ophthalmologists (41/ 174, 23.6%), retina specialists (12/174, 6.9%), optometrists (5/174, 2.9%), neurologists (5/174, 2.9%), primary care providers (7/174, 4%), and other subspecialists (6/174, 3.4%). Seventy-seven of these 174 patients (44.3%) were self-referred (41 presented to the ED and 36 to an outpatient eye care provider). Twenty-one (12%) of these patients were transferred from outside hospitals to our ED or an inpatient service. Evaluations and Treatment Before Presentation to Our Institution Below are the results for any tests, imaging studies, or treatments provided for the 110 patients who presented to TABLE 4. Number of providers seen before presenting to our institution from 2010 to 2020 Number of Providers 2010–2013 2014–2016 2017–2020 0 1 2 3 4 5 (n=72) (n = 49) (n = 38) (n = 11) (n = 7) (n = 3) 12 (17%) 19 (38%) 9 (24%) 4 (36%) 3 (37%) 0 (0%) 20 (28%) 13 (27%) 10 (26%) 0 (0%) 3 (37%) 1 (33%) Flowers et al: J Neuro-Ophthalmol 2021; 41: 480-487 40 17 19 7 2 2 (56%) (35%) (50%) (29%) (25%) (67%) another provider (outpatient or ED) before presenting to our institution. Tests Data were available for 98 of 110 patients (12 had insufficient records). Among the 46 (46.9%) patients who had laboratory tests performed before evaluation at our institution, 43 were over the age of 50 years, 26 of whom had inflammatory markers (erythrocyte sedimentation rate and/or C-reactive protein) measured and 15 of whom did not. Two patients over the age of 50 received blood tests, but no details were provided. Three patients had temporal artery biopsies performed, all of which were negative for arteritis. Imaging Studies Data were available for 106 of 110 patients (4 had insufficient records). Forty-four patients (41.5%) had head computed tomography (CT), 32 patients (30.2%) had brain MR imaging, 48 patients (45.3%) had vascular imaging (22 carotid ultrasound, 8 CT angiography, 5 MR angiography, 2 combined CT angiography/MR angiography, 4 combined MR angiography/carotid ultrasound, 5 combined CT angiography/carotid ultrasound, and 1 carotid angiogram that precipitated the CRAO), 35 patients (33%) had a transthoracic or transesophageal echocardiogram, and 1 patient had Holter monitoring. Treatments Data on 109 of 110 patients were available (1 patient had insufficient records). Twenty patients (18.3%) received some type of treatment before presentation to our institution. Seven patients (6.4%) had an intervention to lower the intraocular pressure (either ocular massage, topical eye drops, or paracentesis). Five patients (4.6%) received oral steroids for suspected giant cell arteritis. Seven patients (6.4%) were started on an antiplatelet or anticoagulant medication (4 on aspirin, 2 on dual antiplatelet therapy, and 1 on heparin). Two patients received hyperbaric oxygen, which in both cases was ordered by an outside vascular surgeon. No patient received thrombolytic treatment before 483 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution presenting to our institution. One patient had an intravitreal bevacizumab injection and panretinal photocoagulation for neovascularization. Subgroup of Patients With Acute CRAO (Presenting Within 1 Week Of Symptom Onset) Among the total 181 patients, 155 patients (86.2%) presented to a provider (any provider, outpatient, or ED) within 1 week (168 hours) of visual loss. Among these 155 patients, the frequency of comprehensive inpatient evaluation by a stroke specialist increased significantly over the past decade, whereas the frequencies of outpatient or incomplete evaluations correspondingly decreased over the same time period (Table 5). All but 2 of the inpatient evaluations occurred at our institution; one occurred at the local Veterans Affairs hospital affiliated with our institution, and the second occurred at a local ED where appropriate stroke workup was performed. CONCLUSIONS Our study showed that 50% of the patients with a CRAO evaluated at our institution over the past decade presented first to an outpatient eye care provider, and 40% presented directly to an ED, with 86% of patients presenting within 1 week of visual loss. The rate of comprehensive evaluations by a stroke specialist for these patients increased from 44% to 82% over the past decade, with only 10% having an incomplete evaluation from 2017 to 2020. Patients continued to see multiple providers before presenting to our institution, but the number of patients who presented directly to us increased from 17% to 56% over the 10 years of the study, and the time to presentation to our institution decreased. Surprisingly, those with a history of ocular conditions, stroke, or on antithrombotic treatments did not present faster than those without. As expected, patients who presented to a local ED or local eye care provider lived further away. Before presenting to our institution, less than half of all patients received some form of vascular imaging and only about one third had a brain MRI, suggesting that any delay in presentation to our institution was not because the patient was undergoing a complete stroke workup elsewhere. Six percent of patients with CRAO received an intervention to lower the intraocular pressure (e.g., anterior chamber paracentesis, ocular massage, or topical medications), and none received thrombolytics before being evaluated at our institution. Few studies have investigated the referral patterns of patients with acute CRAO. In one report of 91 patients with acute CRAO presenting from 2017 to 2020 in North Carolina, only about half of patients (57%) presented to an ED (13). By contrast, a 2018 study from Germany found an increased rate (from 52.2% to 97.4%) of inpatient evaluation for patients with CRAO who presented to their ED 484 (14), after implementation of an intrahospital protocol, supporting one effective means for improving patient care. One of the main tenets in the management of CRAO is evaluation for the underlying cause (e.g., source of emboli or giant cell arteritis) with the goal of preventing a second ischemic event (15). To that end, there has been a rapidly growing consensus that patients with acute CRAO should be immediately referred to a local ED affiliated with a stroke center (3,16–19). The risk of another ischemic event in the subsequent weeks after a CRAO is high; one study found that 32% of patients had a stroke, myocardial infarction, or died within 2 years after a CRAO (20). Furthermore, 34%– 65% of strokes occur within 4 weeks of the CRAO, most often in the first 1–2 weeks (21,22), highlighting the need for timely intervention to prevent this second ischemic event. In addition, about one third of patients with CRAO have ipsilateral critical carotid artery disease which is best treated shortly after the acute ischemic event (20). Indeed, the number needed to treat for carotid endarterectomy quickly increases from 5 (for patients treated within 2 weeks of the ischemic event) to 125 (for those treated after 12 weeks) (23), and the AHA/ASA recommend that carotid endarterectomies occur between 48 hours and 7 days after an acute ischemic event (1). In our study, the median time to presentation for all patients to the first provider was 24 hours, suggesting that many patients are presenting quickly enough to benefit from emergent secondary prevention. However, patients with acute CRAO need to be diagnosed and triaged to a stroke center as quickly as possible after visual loss for optimal care, and to that end, there is still substantial room for improvement. The patients in our study presented to numerous types of providers, highlighting the fact that all healthcare providers, especially all ophthalmologists, regardless of subspecialty, and optometrists, need to recognize and understand the management of acute CRAO. Interestingly, patients with a history of stroke did not present to an ED faster, despite the expectation that this cohort would have received additional education about the signs and symptoms of stroke and the need for urgent evaluation. Equally surprising, patients with an ocular history presented slower to a provider (any provider) and to the ED than those without an ocular history. It is possible that those patients already established with an eye care provider may have been more likely to attempt a less urgent outpatient evaluation rather than immediately presenting to a local ED or they may have erroneously ascribed their new visual symptoms to prior established ocular diagnoses. A second tenet in the management of CRAO is the direct treatment of the CRAO with the goal to reverse vision loss. Although there remains a lack of Level 1 evidence for acute treatment of nonarteritic CRAO (3), previous reports suggest that recanalization of the occluded artery with thrombolysis may improve visual outcomes. Schrag et al’s 2015 meta-analysis showed a visual recovery Flowers et al: J Neuro-Ophthalmol 2021; 41: 480-487 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 1. Breakdown of the first provider seen for all 181 patients after visual loss onset. Forty percent (73/181) of patients presented directly to an emergency department (ED), and 50% (91/181) of patients presented to an outpatient eye care provider. rate of 17.7% in the natural history group, 7.4% in the conservative management group (including ocular massage, anterior chamber paracentesis, and hemodilution), and 31.8% in the thrombolysis group (24). The conservative management group was significantly worse than the natural history group, with a number needed to harm of 10, consistent with other reports of lack of efficacy of so-called conservative treatments for acute CRAO (3,6,18,25). In the recent North Carolina study of 91 patients with acute CRAO (13), 47.3% of patients received at least one conservative therapy compared with only 6.4% of our patients. It is generally agreed that emerging treatments will be more effective if administered quickly, within a short time window after visual loss onset (17,18,26,27). In their study of intravenous thrombolysis treatment for acute CRAO, Schrag et al (24) found a significant time effect, with a 50% rate of visual recovery in those patients treated within 4.5 hours. The most recent recommendations based on existing literature (although not on the results of clinical trials) suggest that intravenous thrombolysis be administered within 4.5 hours of visual loss to not only maximize the potential for visual recovery by facilitating arterial recan- alization before retinal cell death but also to limit potential hemorrhagic cerebral complications (1,18,24). Although the optimal treatment window remains debated (17,18), all current clinical trials are using the time limit of 4.5 hours (clinicaltrials.gov: NCT04526951 and NCT03197194). Although our study showed some improvement in the referral patterns of patients over time, many patients presented to a stroke center too late to benefit from any potential acute treatment and many were evaluated more than a few days after visual loss. Such delays in presentation are likely explained by both general public and health care provider unawareness that a CRAO is the equivalent of a stroke and should be evaluated emergently. Although several successful outreach programs launched in the United States over the last few decades have educated the public on signs of stroke and the importance of calling for emergency services, most did not explicitly include visual loss as a symptom of stroke. One of the first educational campaigns, the FAST campaign, was launched in 1998 in the United Kingdom, and several other outreach programs have since followed (such as the Brain Attack Campaign by the American Academy of Neurology, the Stroke Awareness Foundation, and the AHA’s Get With The Guidelines– Stroke Program). These campaigns aim to increase the public’s general awareness of stroke symptoms and the need to call emergency services instead of consulting their local provider or drive themselves to a local ED. More recently, another mnemonic has been proposed that includes visual symptoms: BE-FAST, resulting in a reduction of the number of missed strokes from 14% to 4.4% in one retrospective study (10). Despite the importance of visual loss being a symptom of stroke, BE-FAST (or other mnemonics including visual symptoms) have not yet been widely adopted and patient care remains delayed (28). A recent study of patients with acute CRAO presenting within 2 weeks to an outpatient retina practice in the United States showed critical delays in “time to workup” in outpatient referrals (13.6 days) vs the ED (2.2 days) (19). Whereas we strongly agree with their proposition that all eye care providers establish a relationship with a stroke center for urgent referrals (and that screening for giant cell arteritis occur in all patients over 50) (15), it is essential that referral pathways be simplified and that the entire workup be deferred to the expert stroke neurologist in the ED. Eye care providers should be able to TABLE 5. Disposition of the 155 patients seen within 7 days of visual loss from 2010 to 2020 Inpatient comprehensive stroke evaluation Outpatient comprehensive stroke evaluation (by a neurologist) No evaluation, incomplete evaluation, or evaluation by a non-neurologist 2010–2013 2014–2016 2017–2020 17 (44%) 4 (10%) 18 (46%) 21 (55%) 4 (10%) 13 (34%) 64 (82%) 6 (8%) 8 (10%) The frequency of inpatient evaluations increased compared with outpatient and inappropriate evaluation, which decreased over time. Fisher exact test, P , 0.001. Flowers et al: J Neuro-Ophthalmol 2021; 41: 480-487 485 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution evaluate patients with acute visual loss emergently, to make the correct diagnosis and immediately send patients with acute CRAO to the closest ED affiliated with a stroke center for rapid workup and potential treatment or inclusion in a clinical trial. For those patients presenting directly to an ED with acute visual loss, urgent ophthalmology consultation, nonmydriatic fundus photography, and/or optical coherence tomography with teleophthalmology consultation should allow for the rapid diagnosis of CRAO and immediate activation of the stroke call (either on site or using existing telestroke networks). One could even envision referral patterns similar to what is ideally currently performed for patients with suspected cerebral stroke. After hospitals successfully improved ‟door to needle” time for stroke patients, focus shifted to improve prehospital care efficiency and mobile stroke transport units (MSTUs) were developed. MSTUs are equipped with CT scanners, trained medics, and access to a neurologist (either in person or by telemedicine) who coordinate care to deliver thrombolytic therapy before the patient even reaches the hospital. This has been shown to decrease treatment time to 25 minutes, and other (more affordable) telestroke units are being investigated (29). With this infrastructure already in place, the addition of portable nonmydriatic fundus cameras and/or optical coherence tomography machines becomes a relatively small obstacle. Automated interpretation of ocular fundus imaging with recently developed artificial intelligence systems will likely enhance this technology in the near future (30,31). There are several limitations to our study, mostly related to its retrospective nature. The data collected heavily relied on proper documentation of symptom onset and studies performed at outside institutions, and such data were not available on all patients. In addition, the home zip code was used as a proxy for distance, which may not reflect the exact distance traveled at the time of visual loss. The findings of this study may be unique to our institution which is located in a very large metropolitan area with relatively easy access to numerous healthcare providers and facilities affiliated with stroke centers. This likely explains why patients who lived within 12 miles of our institution presented faster to any provider than those living further away in rural Georgia where access to care may require longer travel. Despite the abundance of literature on the need for timely presentation to a stroke center for the acute management of CRAO, patients with acute visual loss are still being evaluated by multiple outpatient providers and undergoing incomplete evaluations instead of presenting immediately and directly to an ED affiliated with a stroke center. Although definite improvement has occurred over the past decade at our institution, delays in presentation remain too long to allow for very acute treatments,32 should some become available. Such delays underscore the barriers to performing clinical trials evaluating the potential benefits of very acute treatments for CRAO. Further educational interventions for both health care providers and the general population are necessary. 486 REFERENCES 1. Powers WJ, Rabinstein AA, Ackerson T, Adeoye OM, Bambakidis NC, Becker K, Biller J, Brown M, Demaerschalk BM, Hoh B, Jauch EC, Kidwell CS, Leslie-Mazwi TM, Ovbiagele B, Scott PA, Sheth KN, Southerland AM, Summers DV, Tirschwell DL. Guidelines for the early management of patients with acute ischemic stroke: 2019 update to the 2018 guidelines for the early management of acute ischemic stroke: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke. 2019;50:E344–E418. 2. Olsen TW, Pulido JS, Folk JC, Hyman L, Flaxel CJ, Adelman RA. Retinal and ophthalmic artery occlusions preferred practice pattern. Ophthalmology. 2017;124:P120–P143. 3. Flaxel CJ, Adelman RA, Bailey ST, Fawzi A, Lim JI, Vemulakonda GA, shuang YingG. Retinal and ophthalmic artery occlusions preferred practice pattern. Ophthalmology. 2020;127:P259–P287. 4. Atkins EJ, Bruce BB, Newman NJ, Biousse V. Translation of clinical studies to clinical practice: survey on the treatment of central retinal artery Occlusion. Am J Ophthalmol. 2009;148:172–173. 5. Abel AS, Suresh S, Hussein HM, Carpenter AF, Montezuma SR, Lee MS. Practice patterns after acute embolic retinal artery occlusion. Asia Pac J Ophthalmol (Phila). 2017;6:37–39. 6. Youn TS, Lavin P, Patrylo M, Schindler J, Kirshner H, Greer DM, Schrag M. Current treatment of central retinal artery occlusion: a national survey. J Neurol. 2018;265:330–335. 7. Flynn D, Ford GA, Rodgers H, Price C, Steen N, Thomson RG. A time series evaluation of the FAST national stroke awareness campaign in England. PLoS One. 2014;9:e104289. 8. Dombrowski SU, White M, Mackintosh JE, Gellert P, AraujoSoares V, Thomson RG, Rodgers H, Ford GA, Sniehotta FF. The stroke “Act FAST” campaign: remembered but not understood? Int J Stroke. 2015;10:324–330. 9. Alberts MJ, Wechsler LR, Jensen ME, Latchaw RE, Crocco TJ, George MG, Baranski J, Bass RR, Ruff RL, Huang J, Mancini B, Gregory T, Gress D, Emr M, Warren M, Walker MD. Formation and function of acute stroke-ready hospitals within a stroke system of care recommendations from the brain attack coalition. Stroke. 2013;44:3382–3393. 10. Aroor S, Singh R, Goldstein LB. BE-FAST (balance, eyes, face, arm, speech, time): reducing the proportion of strokes missed using the fast mnemonic. Stroke. 2017;48:479–481. 11. Man S, Schold JD, Uchino K. Impact of stroke center certification on mortality after ischemic stroke: the medicare cohort from 2009 to 2013. Stroke. 2017;48:2527–2533. 12. Hayreh SS. Prevalent misconceptions about acute retinal vascular occlusive disorders. Prog Retin Eye Res. 2005;24:493–519. 13. Lee KE, Tschoe C, Coffman SA, Kittel C, Brown PA, Vu Q, Fargen KM, Hayes BH, Wolfe SQ. Management of acute central retinal artery occlusion, a “retinal stroke”: an institutional series and literature review. J Stroke Cerebrovasc Dis. 2021;30:105531. 14. Hoyer C, Kahlert C, Güney R, Schlichtenbrede F, Platten M, Szabo K. Central retinal artery occlusion as a neuro‐ ophthalmological emergency: the need to raise public awareness. Eur J Neurol. 2021;28:2111–2114. 15. Mac Grory B, Schrag M, Biousse V, Furie KL, Gerhard-Herman M, Lavin PJ, Sobrin L, Tjoumakaris SI, Weyand CM, Yaghi S; American Heart Association Stroke Council; Council on Arteriosclerosis, Thrombosis and Vascular Biology; Council on Hypertension; and Council on Peripheral Vascular Disease. Management of central retinal artery occlusion: a scientific statement from the American Heart Association. Stroke. 2021;52:e282–e294. 16. Arnold AC. Urgent evaluation of the patient with acute central retinal artery occlusion. Am J Ophthalmol. 2018;196:16–17. 17. Scott IU, Campochiaro PA, Newman NJ, Biousse V. Retinal vascular occlusions. Lancet. 2020;396:1927–1940. Flowers et al: J Neuro-Ophthalmol 2021; 41: 480-487 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution 18. Dumitrascu OM, Newman NJ, Biousse V. Thrombolysis for central retinal artery occlusion in 2020: time is vision! J Neuroophthalmol. 2020;40:333–345. 19. Vangipuram G, Yang L, Weigle MP, Blackorby BL, Blinder KJ, Dang S, Shah GK. Workup following retinal artery occlusion— experience from an outpatient retina clinic and the delay in workup. Graefes Arch Clin Exp Ophthalmol. 2021:1–8. doi: 10.1007/s00417-021-05135-x. 20. Lavin P, Patrylo M, Hollar M, Espaillat KB, Kirshner H, Schrag M. Stroke risk and risk factors in patients with central retinal artery occlusion. Am J Ophthalmol. 2018;196:96–100. 21. Park SJ, Choi NK, Yang BR, Park KH, Lee J, Jung SY, Woo SJ. Risk and risk periods for stroke and acute myocardial infarction in patients with central retinal artery occlusion. Ophthalmology. 2015;122:2336–2343.e2. 22. French DD, Margo CE, Greenberg PB. Ischemic stroke risk in medicare beneficiaries with central retinal artery occlusion: a retrospective cohort study. ophthalmology and therapy. Ophthalmol Ther. 2018;7:125–131. 23. Gocan S, Bourgoin A, Blacquiere D, Shamloul R, Dowlatshahi D, Stotts G. Fast-Track systems improve timely carotid endarterectomy in stroke prevention outpatients. Can J Neurol Sci. 2016;43:648–654. 24. Schrag M, Youn T, Schindler J, Kirshner H, Greer D. Intravenous fibrinolytic therapy in central retinal artery occlusion: a patient-level meta-analysis. JAMA Neurol. 2015;72:1148–1154. 25. Sharma R, Newman N, Biousse V. Conservative treatments for acute nonarteritic central retinal artery occlusion: do they work? Taiwan J Ophthalmol. 2021;11:16–24. Flowers et al: J Neuro-Ophthalmol 2021; 41: 480-487 26. Mac Grory B, Lavin P, Kirshner H, Schrag M. Thrombolytic therapy for acute central retinal artery occlusion. Stroke. 2020;51:687–695. 27. Hayreh SS. Acute retinal arterial occlusive disorders. Prog Retin Eye Res. 2011;30:359–394. 28. Biousse V, Nahab F, Newman NJ. Management of acute retinal ischemia: follow the guidelines! Ophthalmology. 2018;125:1597–1607. 29. Belt GH, Felberg RA, Rubin J, Halperin JJ. In-Transit telemedicine speeds ischemic stroke treatment: preliminary results. Stroke. 2016;47:2413–2415. 30. Milea D, Najjar RP, Zhubo J, Ting D, Vasseneix C, Xu X, Aghsaei Fard M, Fonseca P, Vanikieti K, Lagrèze WA, la Morgia C, Cheung CY, Hamann S, Chiquet C, Sanda N, Yang H, Mejico LJ, Rougier M-B, Kho R, Thi Ha Chau T, Singhal S, Gohier P, Clermont-Vignal C, Cheng C-Y, Jonas JB, Yu-Wai-Man P, Fraser CL, Chen JJ, Ambika S, Miller NR, Liu Y, Newman NJ, Wong TY, Biousse V. Artificial intelligence to detect papilledema from ocular fundus photographs. N Engl J Med. 2020;382:1687– 1695. 31. Kohane I. AI for the eye—automated assistance for clinicians screening for papilledema. N Eng J Med. 2020;382:1760– 1761. 32. Chan W, Flowers AM, Meyer BI, Bruce BB, Newman NJ, Biousse V. Acute Central Retinal Artery Occlusion Seen Within 24 hours at a Tertiary Stroke Center. J of Stroke Cerebrovasc Dis. [published online ahead of Print]. doi: 10.1016/ j.jstrokecerebrovasdis.2021.105988. 487 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited.