June 3, 2022 Conference


AM Session - Cataract Surgery in the Setting of Corneal Disease (with Inaugural Hal M. Freeman Lecture)

Program objectives:

1.  Increasing the understanding of the importance of managing ocular surface disease in maximizing post-surgical outcomes.

2. Improving understanding of the perioperative management of congenital and acquired corneal and anterior segment conditions which may affect the outcomes of cataract surgery.

3. Discussing the challenges of using new lens technology in both academic and private practice settings in patients with a variety of corneal conditions.


Cataract Surgery in Anterior Segment Dysgenesis: Lessons from Pediatric Ophthalmology
Deborah K. VanderVeen

Anomalous anterior segment structures pose several unique challenges that should be considered to maximize visual outcomes, and congenital forms of dysgenesis vary from abnormalities that are acquired later in life. While pediatric ophthalmologists typically make the diagnosis and manage these patients through the amblyogenic years, adult eye care providers can benefit from understanding both short and long-term challenges. Case presentation format will be used to review some considerations for non-surgical and surgical management in more commonly encountered conditions such as Axenfeld Reiger anomaly, Aniridia/iris hypoplasia, Peters anomaly, congenital cataract, and more severe types of combined anterior segment anomalies. Understanding the evolution of eye findings over time can help the adult practitioner manage such patients in adulthood, the most common being development of glaucoma in 50%. While the genetics of ASD is complicated, correlation of clinical findings with genetic mutations can provide guidance for follow up, management, and heritability.


Pre and Postoperative Management of the Neurotrophic Cornea
Erin Fogel

Neurotrophic Keratitis (NK) is a degenerative corneal disease characterized by damage to the trigeminal nerve, leading to loss of corneal sensation and breakdown of the corneal epithelium. The ophthalmic division of the trigeminal nerve innervates the cornea, lacrimal glands, conjunctiva, and eyelids and is responsible for blinking and tear secretion as protective reflexes. Advanced stages of NK include recurrent epithelial defects, stromal scarring, ulceration, and perforation. Etiologies of NK are numerous, but the most common causes include herpes simplex and herpes zoster keratitis, and neurosurgery for trigeminal neuralgia or acoustic neuroma, diabetes and contact lens wear. While NK falls under the category of Ocular Surface Disease (OSD) and appears in the ASCRS preoperative OSD algorithm, it presents unique diagnostic challenges for the cataract surgeon in its early stages as patients don’t complain of discomfort. Furthermore, while the treatment of OSD and NK overlap initially, more intensive interventions are needed as NK progresses. Punctal occlusion, preservative free drops, serum tears, amniotic membrane, and tarsorrhaphy have long been employed but the availability of a recombinant human nerve growth factor drop, cenegermin, has added to our armamentarium. We will discuss strategies to optimize the ocular surface preoperatively as well as intraoperative measures to deal with accompanying corneal scars and thinning. Finally, we will consider post operative considerations to prevent further breakdown of the epithelium and maximize the outcome of our cataract surgery patients with NK.


Peri-operative Management of Viral Keratouveitis: HSV and VZV
Priya Janardhana

Cataract surgery can cause reactivation of herpes keratouveitis in patients with a history of this or cause primary occurrence of herpes keratitis and uveitis post cataract surgery. It is important to know how to identify these patients clinically and know how to treat these patients and prophylactically manage patients with a history of keratouveitis. We will discuss these various clinical scenarios and potential complications, and how to treat these patients. 


Cataract Surgery in the Setting of Corneal Disease
Esen Akpek

Corneal clarity or shape abnormalities pose significant challenges during pre-operative planning for cataract surgery. Clinicians need to account for individual patient's unique circumstances and expectations to decide whether or not to combine cataract removal with corneal transplantation and the type of corneal transplantation. Additional challenges include obtaining precise biometry for the lens implant and choosing which formula to use for the intraocular lens calculations. Furthermore, removal of the cataract can be demanding, whether combined with a simultaneous keratoplasty or performed independently. We will discuss each scenario by reviewing patient cases and offering best practices to avoid potential pitfalls.


Toric/Multifocals in the Setting of Refractive Surgery
Santiago Villazon

As millions of refractive surgeries have been performed over the past 25 years, these patients are now entering the age of needing cataract surgery. These patients are expecting LASIK results with their cataract surgery and would like to explore the newer intraocular lens technologies to help decrease their dependence on glasses for distance, intermediate and near. Unfortunately, not everyone is an ideal candidate. Absolute and relative contraindications will be reviewed to select the ideal patients. In order to succeed in these patients, precise refractive outcomes for their cataract surgery are essential. Prior to implanting newer technology IOLs, surgeons will first need experience with performing routine cataract surgery on laser vision correction patients, especially understanding how to calculate IOL power after laser vision correction. Options that are available to all cataract surgeries such as Monofocal, Extended Depth of Focus, Multifocal, Monovision, and Toric IOLs will be reviewed. Newer IOL categories, such as the Light Adjustable IOLs will be explored.


To triple or not to triple?
Emma Davies

The decision for best surgical intervention for patients with Fuchs corneal dystrophy and cataracts can be challenging given a variety of options and complex interactions between surgeries. Careful consideration of severity of Fuchs corneal dystrophy and density of cataract should be weighed against the patient’s visual goals. Several important pre-operative measurements should be obtained to determine the severity of Fuchs corneal dystrophy to guide surgical planning including central corneal thickness, central endothelial cell count, and peripheral endothelial cell count. Seitzman et al. demonstrated that patients with a pre-operative central corneal thickness <640µm had good visual outcomes after cataract surgery while 22.2% of patients with central corneal thickness >640µm developed corneal decompensation requiring keratoplasty1. On specular microscopy, central endothelial cell count <500 cells/mm2 leads to corneal decompensation and should be taken to indicate the need for corneal surgical intervention2. Central endothelial cell count >1000 cells/mm2 indicates that careful cataract surgery alone should be well tolerated2. The decision between Descemet’s membrane stripping only (DSO) versus Descemet’s membrane endothelial keratoplasty (DMEK) should be made according to peripheral endothelial cell count (DSO requires 500-1000 cells/mm2 in the mid-periphery) and patient’s ability to wait 4-6 weeks after surgery for corneal clearance.   

The decision regarding combined or staged surgeries depends on patient preference, surgeon preference, and overall risks. Additional factors that need to be considered are an extremely dense lens that would be too challenging or too inflammatory to sustain a combined surgery or an extremely edematous cornea that prevents adequate surgical view for cataract surgery. The benefits of combined surgery are reduction in the number of operations, cost, and recovery time3. Whereas staged DMEK allows for better evaluation of optical biometry after the corneal graft heals3. There is conflicting data regarding graft health with staged DMEK surgery but overall endothelial cell loss in graft tissue ranges from 5% to 25%3.  

Overall, pre-operative determination of the severity of Fuchs corneal dystrophy and cataract should precisely guide the decision for type of surgery and timing of surgery.  


Optimizing the Ocular Surface Prior to Cataract Surgery can Prevent Complications and Patient Dissatisfaction
Esen Akpek

Dry eye and blepharitis are common ocular surface diseases that can reduce patient quality of life due to discomfort symptoms and fluctuating vision. Cataract surgeons need to assess for pre-existing ocular surface disease and instigate treatment prior to surgery to minimize poor outcomes. Among patients with untreated ocular surface disease, endophthalmitis, a rare but devastating complication, is a well-known risk, despite unremarkable cataract surgery. Reduced accuracy of biometry leading to unexpected post-operative refractive error or significant visual symptoms due to irregular corneal surface are less well known issues that result in patient dissatisfaction. Routine cataract surgical evaluation should include an assessment of ocular surface disease presence.


Introducing Premium Lenses to Resident Teaching
Susannah G. Rowe

Increasingly, premium lenses represent an important option for patients undergoing cataract surgery and for their surgeons. Thus, as part of training, residency programs must prepare new cataract surgeons to effectively navigate the rapidly changing world of refractive cataract surgery. In order to to optimize patient satisfaction and functional outcomes, new graduates need to know how to: evaluate emerging technologies; measure and assess patients as candidates for premium lenses; calculate the correct lens power; educate and guide patients in IOL choices; implant premium lenses, and parse relevant ethical and financial decisions.

To support this didactic goal and encourage use of premium lenses, many premium lens manufacturers offer a supply of at-cost premium lenses to teaching institutions for purposes of resident education. Such programs typically include a limited inventory and requirements with respect to resident involvement, posing opportunities as well as ethical conundra for academic cataract surgeons. Participants must consider principles of equity, obligations to respect patients' wishes and provide excellent care, duty to prepare future surgeons to provide excellent care to future patients, and impact on the practice. Approaches include collaborating with manufacturer's teaching programs, negotiating variations on requirements, and opting out altogether.

We will review some of the didactic, ethical, equity and practical considerations inherent to teaching residents to use premium lenses.


Undoing New Technology: Cataract surgery in the setting of ICLs, Intacts, Raindrop and Smile
Sara Bozorg

Patients who have undergone prior refractive surgery with either ICLs, Intact, Raindrop or Smile pose a challenge for surgeons for a number of reasons. These include difficulty in accurately choosing the correct IOL power, unique intraoperative challenges, and the extra chair time needed to explain to the patient that despite best efforts they still may need glasses for all distances. A thorough examination of the patient and eye will help to create an individual plan that will answer each of the following questions: - Should a device be removed prior to cataract surgery? - Which formulas will give the most accurate IOL power calculation? - How will cataract surgery itself be different? - Does the patient understand the limitations and extra risk involved in their surgery?


PM Session - Ethics: Increasing Diversity and Health Equity in Ophthalmology (with Hutchinson Lecture)

Program Objectives:

1.  Increasing awareness about the current state of diversity in the ophthalmology workforce

2.  Discuss the impact on delivery of eyecare to underserved communities

3. Present the challenges posed to assessing ophthalmic interventions /research by a lack of diversity


The Ethics of the ACO and Ophthalmology
Christopher Andreoli

•The goals of the ACO model are laudable:  improve quality, outcomes, and create a sustainable financial model
•Adoption has not been universally successful
•Potential for abuse exist
•The medical profession should lead the future generations of sustainable high quality care or risk future loss of control to regulators and insurers


Telehealth Encourages More Minority Patients With Diabetes to Return for In-Person Care During the COVID-19 Pandemic
Shiyoung Roh

Eye examinations are important for the early detection of diabetic retinopathy (DR) and allows treatment capable of preventing vision loss and associated disability.  Nearly half of patients with diabetes mellitus (DM) do not obtain eye examinations even when they are covered by health insurance.  Importantly, historically marginalized populations have faced barriers to accessing eye care with rates of annual diabetic eye screening as low as 25% in some underserved populations.

In our multispecialty hospital clinic, we evaluated the impact of telehealth (TH) encounters on the likelihood of patients with diabetes mellitus (DM) return for subsequent in-person eye examination during the period of the coronavirus disease 2019 (COVID-19) pandemic.  COVID-19 caused a precipitous decline in return visits: 42% in 2020 versus 68% in the same period in 2019 (?2 = 874.246, P < 0.001).  The impact of COVID-19 was most pronounced for patients from racial and/or ethnic minorities: 45% White vs 36% non-White and/or Hispanic had return visits (?2 = 15.035, P < 0.001).   We found that TH correlated with completion of an in-person eye examination: (OR 1.5; 95% CI 1.3 to 1.7; P < 0.001).  Of those who completed a TH visit, 50% returned for subsequent in-person eye examinations compared with 40% of those who did not complete a TH encounter (?2 = 36.740, P < 0.001).  Importantly, TH delivered by an ophthalmologist was twice as great for non-White and/or Hispanic patients compared with White patients (OR 4.1[95% CI 2.8-6.3] vs. 2.1 [95% CI 1.8-2.5], z-value 3.119, P<0.001).

Conclusion: In our study, TH delivered during the COVID-19 pandemic increased the rate at which patients with DM returned for in-person eye examinations and TH delivered by an ophthalmologist more positively impacted minority patients. Future studies are needed to determine if TH delivered as part of regular eye care can help close gaps important for achieving the best health outcomes for our patients.


Ethical Issues in Healthcare: Medicolegal Aspects of Care in Ophthalmology
Christopher Teng

Malpractice is professional negligence and medical malpractice is the negligence of a doctor.

Negligence is the failure to use reasonable care under the circumstances, doing something that a reasonably prudent doctor would not do under the circumstances, or failing to do something that a reasonably prudent doctor would do under the circumstances.

Medical Malpractice lawsuits are most common in

  1. Diagnosis 34%
  2. Surgery 22%
  3. Treatment 19%

Average indemnity in ophthalmology

  • Cataract/Cornea $304,476
  • Retina $270,141
  • Oculoplastics $222,471

The power of the n of 1

  • Averages don’t matter, statistics don’t matter
  • The n of 1 is the patient in front of you
  • Do what’s best for them, when they need it

5 Effective Risk Prevention Strategies

  1. Achieving a good result
  2. Establishing and maintaining the physician patient relationship
  3. Effective informed consent (emphasis on risks)
  4. Thorough documentation and good medical records
  5. Identifying and managing litigious patients


Ethical Issues in Genetic Testing: Principles and case studies from pediatric ophthalmology
Deborah K. VanderVeen

Availability of genetic testing for ocular diseases is rapidly increasing, and with expansion of available testing comes a need to understand the role of the provider as well as implications for patients and families. Reasons for genetic testing may include the need to establish a diagnosis, to establish carrier status, or to predict future disease or heritability. Testing may reveal a mismatch between the classic phenotype and genotype findings, unexpected results, or a mutation may be identified for which there is lack of information or understanding about implications for an individual or family. Non-directive genetic counselling is an important responsibility on the part of the clinician, as well as recognizing patient concerns regarding maintenance of confidentiality or discrimination in access to health care, insurance coverage, or employment opportunities. A case presentation format will be used to illustrate use genetic testing in the pediatric ophthalmology setting, and how this impacted management for these cases.


Medicine and Ophthalmology: Working Toward Racial Equity
M. Roy Wilson

That some racial/ethnic groups experience better health than others is an issue of equity and justice. It is also an issue with economic and quality implications. The lack of a racially diverse physician workforce also impacts quality healthcare. Ophthalmology is among the least diverse of the medical/surgical specialties.


Innovations in Ophthalmology Education to Increase Diversity
Ninani Kombo

There is an abundance of research that demonstrates that diverse teams that include different kinds of thinkers outperform homogenous groups on complex tasks. This often translates into important and valuable results, for example, better problem-solving, greater innovation, and in the case of health care delivery, more equitable care.

Underrepresented groups in Medicine (UIM) include women, Blacks, Latino/Latinx, Native Americans, Hawaiian Natives, and Pacific Islanders. Members of the LGBTQ+ community are also UIM. The numbers of each of these groups in our population are not represented in the medical profession. The specialty of Ophthalmology is no exception. 

What are we doing right now? What can we do in education to address this? There are some organizations including the Rabb-Venable Excellence in Research Program, The Minority Ophthalmology Mentoring Program through the AAO that are dedicated to addressing this issue. 

Everyone can make a positive difference.  


Evolving Concepts in Race-based Medicine: My Personal Journey
M. Roy Wilson

Two of the major advances in understanding glaucoma in Blacks have been in defining race as a social construct and in the use of race as a variable in medicine. It is now well recognized that geography/ancestry of origin is more meaningful and useful than "race" and that social determinants of health are more important in explaining health disparities than "race." Yet race continues to be over-used in medicine. Race-based medicine perpectuates a false narrative that race is an essential, biological variable and translates this falsehood into clinical practice. Race-based medicine is bad medicine and should be avoided. 


Diversity in Ophthalmology: More Than Just Another [Pretty] Face
James Chodosh

This brief presentation will focus on the many meanings of diversity, and what a truly diverse profession of ophthalmology might look like.


Introduction of Session
David Ramsey

Posters


Quantitative wide-field swept-source optical coherence tomography angiography and visual outcomes in RAO
Yifan Lu

Purpose: Retinal artery occlusion (RAO) is an ophthalmic emergency that can lead to poor visual outcome and is associated with an increased risk of cerebral stroke and cardiovascular events. Fluorescein angiography (FA) is the traditional diagnostic tool for RAO; however, wide-field swept-source optical coherence tomography angiography (WF SS-OCTA), as a nascent imaging technology, is able to provide quick and non-invasive angiographic information with a wide field of view. In this study, we looked for associations between OCT-A vascular metrics and visual acuity in patients with prior diagnosis of RAO. Methods: Patients with diagnoses of central retinal artery occlusion (CRAO) or branched retinal artery occlusion (BRAO) were included. A 6mm x 6mm Angio and a 15mm x 15mm AngioPlex Montage OCT-A image were obtained for both eyes in each patient using the Zeiss Plex Elite 9000 WF SS-OCTA device. Each 6mm x 6mm image was divided into nine Early Treatment Diabetic Retinopathy Study (ETDRS) subfields. The average measurement of the central foveal subfield, inner ring, and outer ring was calculated for each parameter. Non-perfusion area (NPA) was manually measured using 15mm x 15mm Montage images. A linear regression model was utilized to identify a correlation between the imaging metrics and visual acuity. A P-value less than 0.05 was considered to be statistically significant. Results: Twenty-five subjects were included in the study. For RAO eyes, there was a statistically significant negative correlation between vision and retinal thickness as well as superficial capillary plexus vessel density (SCP VD). A negative correlation was found between vision and deep capillary plexus vessel density (DCP VD) without statistical significance. There was a positive correlation between vision and choroidal thickness as well as choroidal volume without statistical significance. No statistically significant correlation was found between vision and the above metrics in contralateral eyes. For NPA measurements, no significant correlation was found between vision and NPA. Conclusions: This is the first study to investigate the utility of WF SS-OCTA in RAO and to demonstrate correlations between various retinal vascular imaging metrics and visual outcomes. Further investigations should explore the associations between these imaging findings and cardiovascular risk as RAO patients are at elevated risk for symptomatic stroke. The results of this study provide a basis to understand the structural changes involved in visual outcomes in RAO. Furthermore, they may help guide management of RAO and prevention of cerebral stroke and cardiovascular accidents in patients with RAO.


Relationship Between Time to OR, Final Visual Acuity and Endophthalmitis in a Retrospective Study of Open Globe Injuries
Kevin Makhoul, Racquel Bitar, Grayson Armstrong, Marguerite Weinert, Francesca Kahale , Ta Thong, Alice Lorch, Dean Eliott, Marisa Tieger

Introduction; 

Repair of open globe injuries within 24 hours from time of injury is the current standard of care. This standard was developed as a result of previous reports suggesting poor visual outcomes and increased rates of endophthalmitis following delayed repair. This retrospective study evaluated 778 open globe injuries repaired by the trauma service at the Mass Eye and Ear to assess the impact of time to OR on visual outcomes and rates of endophthalmitis.

Methods;

A retrospective chart review was performed for open globe injuries repaired between January 2012 to January 2022. Multivariate regression was completed with time to OR, age, sex, race, ethnicity, laterality, maximal zone of injury (MZI), uveal prolapse, lensectomy, IOFB, APD, mechanism of injury (MOI: rupture vs penetrating laceration), and post-op surgeries included as covariates with final VA as the outcome and separately with endophthalmitis as the outcome. VA was converted from Snellen to LogMAR. Time to OR was expressed on a binary scale of within 24 hours versus greater than 24 hours. 

Results;

778 cases of OGI were analyzed. Multivariate analysis revealed a coefficient of 0.105 (p = 0.224) between time to OR and VA. Significant correlation coefficients were found for: MZI (0.307, p = 0.000), age (0.007, p = 0.004), MOI laceration vs rupture (-0.500, p = 0.000), uveal prolapse (0.309, p = 0.000), APD (0.808, p = 0.000), and lensectomy (0.221, p = 0.012). Time to OR greater than 24 hours had a 3.68% increased risk for endophthalmitis (p=0.004).

Conclusions;

The study suggests time to repair of open globe injury is not a clinically significant prognosticator for final VA but was associated with higher rates of endophthalmitis. Other variables that did have a statistically significant impact on final visual acuity include: zone III injuries, older age, rupture, uveal prolapse, presence of an APD and lens violation.


Descemet’s Stripping Endothelial Keratoplasty Outcomes in Complex Eyes
Sila Bal, Yvonne Wang, Kevin Ma, Rohan Singh, Jimena Tatiana Carreno Galeano, Nayan Sanjiv, Ula Jurkunas, Jia Yin

PURPOSE: Outcomes of Descemet’s stripping endothelial keratoplasty (DSEK) in very complex eyes METHODS: Retrospective review of DSEK outcomes in eyes with ocular comorbidities at Mass Eye and Ear between 2014 and 2020 RESULTS: The cohort included 254 patients, mean age 69.8 years, females 133 (52.4%). Median number of previous intraocular surgeries was 2 (range 0-6); 59.4% patients had prior glaucoma surgery and 78.7% prior vitrectomy. Overall graft failure rate was 27%; graft survival rate was 88.2% and 82.7% at one and two years, respectively. Graft rejection rate was 2.8% and rebubble rate was 20%. Logistic regression revealed that previous glaucoma surgery (OR 2.2, p=0.02) and previous penetrating keratoplasty (OR 2.3, p=0.045) were associated with graft failure. CONCLUSION: DSEK is a viable option for treating corneal edema in eyes with multiple comorbidities. While graft survival is moderate, it is limited in eyes with a history of glaucoma surgery or penetrating keratoplasty.


Program for Underrepresented Minority Peer Mentorship (PUMP): A Structured Mentorship Program to Improve Diversity in the Ophthalmology Workforce
Sila Bal, Nakul Singh, James Chodosh, Ankoor Shah

Purpose: To develop a structured, near-peer mentorship program between Harvard Medical School students who identify as underrepresented in medicine (URiM) and Massachusetts Eye and Ear/Harvard Ophthalmology residents. Methods: A novel, structured, near-peer mentorship program was developed matching student mentees with resident mentors and implemented between January 2021 and June 2022. The program was divided into four components including education (e.g., lectures and grand rounds), clinical (e.g., shadowing, wet labs, and formal skills labs), research, and career development. Two surveys were developed using previously validated and accepted mentorship evaluation surveys and administered to the mentors and mentees. Statistical analysis of survey questions was performed using basic descriptive statistics of aggregate medians of each survey question response. Harvard Medical School Institutional Board Review assessment determined the surveys to fall under quality improvement. Results: Sixteen mentees participated in the program over a 1.5-year period. Of those, 3 students chose a path outside of ophthalmology. Five students completed the mentee survey, 2 students reported their race/ethnicity as Black, 2 as Latinx, and 1 as East Asian. Notable survey results from the mentee survey included a median score of 5 (IQR=1) when asked “I have increased confidence in my ophthalmological clinical skills as part of this program,” 5 (IQR=0) when asked “this program increased my interest in ophthalmology,” 5 (IQR=0) when asked “I am more likely to pursue a residency in ophthalmology because of participation in this program,” and 5 (IQR=0) when asked “how likely are you to recommend this program to other medical students” where 0 corresponds to strongly disagree, 1 to disagree, 2 to slightly disagree, 3 to slightly agree, 4 to agree, and 5 to strongly agree. Five residents completed the mentor survey with a median score of 5 (IQR=1) in response to “I grew as an educator as part of this experience” and 4 (IQR=1) in response to “I felt that I was able to effectively engage with the students during this mentorship program.” Conclusion: There are significant racial, ethnic, and gender/sexual orientation disparities in the ophthalmology physician workforce with respect to the demographics of the US population. Addressing these disparities requires a multifaceted approach, including improving the pipeline of students interested in ophthalmology. We found a positive impact of a novel, structured, near-peer mentorship program on URiM medical students’ perspective of the field. Similar structured mentorship programs and the development of longitudinal relationships with mentees may encourage students to pursue ophthalmology.


High positive predictive value of fluorescein angiography contiguous, perinerve retinal vascular leakage pattern for birdshot chorioretinopathy
Ashley Li, Atitaya Apivatthakakul, Lucia Sobrin

Purpose:

Birdshot chorioretinopathy (BSCR) is a posterior uveitis with ovoid yellow-white choroidal lesions. If untreated, patients’ vision declines, so early diagnosis is critical. While indocyanine green angiography (ICGA) is used to identify subtle lesions, ICGA is not always available. Fluorescein angiography (FA) is more common and a contiguous, perinerve retinal vascular leakage pattern has been described in BSCR patients. To facilitate BSCR diagnosis, we determined the sensitivity and positive predictive value of a contiguous, perinerve retinal vascular leakage FA pattern.

Methods:

Mass General Brigham patients with a FA Common Procedural Terminology code and mention of BSCR in the free text of notes were identified. BSCR diagnosis required HLA-A29 positivity and typical lesions on fundus photos and/or ICGA hypofluorescent spots. Chart review was performed to confirm BSCR vs. other uveitis/retinal vasculitis diagnosis and symptom duration. The first FA was analyzed for a contiguous, perinerve pattern of leakage by two readers. A perinerve leakage pattern was defined as leakage primarily around the optic nerve and along the larger arcade vessels. We compared the rates of this FA pattern in BSCR vs. other posterior uveitis/retinal vasculitis patients using the chi-square test and determined the sensitivity and PPV of this FA pattern for BSCR diagnosis. All statistical analyses were performed in STATA. 

Results:

60 BSCR patients and 120 patients with other posterior uveitis and/or retinal vasculitis were identified. A perinerve FA pattern was more common in BSCR patients vs. patients without BSCR (56.7% vs. 6.7%, p=7.432 X 10-14). The sensitivity and specificity of the FA pattern were 56.7% and 93.3%, respectively. The PPV was 80.95%. BSCR patients with an FA perinerve pattern had a shorter time from symptom onset to FA vs. BSCR patients without the pattern (212 vs. 1331 days, p=.0014, t-test).

Conclusions:

An FA contiguous, perinerve retinal vascular leakage pattern is a useful tool to identify potential BSCR for further imaging and serological testing. This pattern is more common when the interval between symptom onset and FA is shorter.


The Impact of Race on Surgical Outcomes for Rhegmatogenous Retinal Detachments
Jeannie Xu, Samaneh Davoudi, Jamie Yoon, Manju L. Subramanian, Steven Ness

Sociodemographic factors including non-White race and lower regional mean household income (MHI) are potential risk factors for delayed presentation, lower single surgery success rates (SSSR), and worse visual outcomes in patients with rhegmatogenous retinal detachment (RRD). The purpose of this study is to reevaluate these findings in a safety-net hospital patient population and to determine if differences in preoperative characteristics other than sociodemographic factors may account for variations in outcomes.

A retrospective cohort study on patients who underwent surgical repair of RRDs at a US safety-net hospital was conducted. Patients were excluded if they had previous RRD repair, traction or serous RD, or did not undergo surgical repair. Demographic, clinical, and operative findings were collected from chart review. The primary outcome was the SSSR between White and Minority patients with secondary outcomes comparing preoperative characteristics and visual outcomes between groups.

A total of 195 patients were categorized as White (n=71) and Minority (n=124) based on self-reported race. SSSR was similar between groups (73.2% White vs. 73.4% Minority, p=0.98). Minority patients were significantly younger, had lower MHI, were more likely to have multiple retinal breaks, and had a higher incidence of preoperative PVR. While univariate testing found race as a predictor of visual outcomes, multivariate analysis showed that only macula status, presence of PVR, and SSSR as significant predictors of visual outcomes. Our results suggest that factors other than socioeconomic status may account for differences in outcomes in RRDs in Minority patients. The higher rate of PVR in Minority patients, if verified in future studies, could impact clinical management of RRD in this patient population.


A Novel 3D Printed Silicone Simulation Model for Effective Instruction of Marginal Eyelid Laceration Repair
Nicole Lifson

Purpose: Ophthalmology residents are often called upon to repair complex eyelid lacerations, particularly marginal eyelid lacerations, with limited prior experience. Using 3D printing technology we created a novel silicone model of the eyelid to allow residents to easily and affordably practice their surgical skills. We performed a prospective study to determine the efficacy of these models as a training tool for marginal eyelid laceration repair.

Methods: 3D design software was used to create a model of the face and eyelids which served as the scaffold for a silicone eyelid mold. A total of 18 ophthalmology residents participated in a training session on marginal eyelid laceration repair, practicing on these models. Prior to and following the session, residents anonymously responded to a survey assessing their degree of confidence in repairing marginal and non-marginal eyelid lacerations, as well as their overall opinion of the workshop. Questions related to their confidence and ability utilized a 5-point Likert Scale: 1. Poor 2. Fair 3. Good 4. Very Good. 5. Excellent. The responses were compared using the Wilcoxon Signed-Rank test.

Results: Following the training session, residents showed a significant improvement in confidence and knowledge of marginal eyelid laceration repair. Amongst all residents, the average confidence level for repairing marginal eyelid lacerations increased from 2.33 to 3.22 (p=0.005) following training with the silicone models. The residents with no prior marginal eyelid repair experience reported a larger increase in the degree of confidence with scores rising from 1.14 pre-session to 2.29 post-session (p=0.01). Overall, attitudes regarding the training session were positive. Residents found the simulation to be beneficial with a mean score of 4.47 (SD 0.62). When asked to compare this model to other eyelid simulations for eyelid laceration repair, the silicone model was found to be superior with a mean score of 4.50 (SD 0.53).

Conclusions: 3D printed silicone eyelid models provide an effective and realistic simulation of marginal eyelid laceration repair.


Automated machine learning models for diabetic retinopathy screening using handheld fundus cameras in a low-resource community screening program
Cris Martin Jacoba, Duy Doan, Joseph Paolo Silva, Lizzie Aquino, Recivall Salongcay, Claude Salva, Kexin Zhang, Tunde Peto, Jennifer K. Sun, L. Paul Aiello, Paolo Silva

Introduction:

Evaluating diabetic retinopathy severity from retinal images is one of the most common uses of AI in medicine, with two current FDA-approved algorithms in the US.1 However, commercially available algorithms are expensive, not universally generalizable, and limited to specific retinal cameras and fields, limiting broader clinical use. The generalizability and validity of AI algorithms are particularly challenged in translating findings across different clinical settings, adapting to patient population-level differences and local clinical idiosyncrasies.

Automated machine learning, or AutoML, allows the development of code-free deep learning models that decrease the barrier to using artificial intelligence to address relevant clinical needs.2 The ease and customization potential of AutoML allows the development of models based on specific use cases driven by the clinical need. This is particularly useful in low resource settings where the financial incentive of commercial ventures is low but the clinical need is high. In this study, we used Google Cloud Platform’s AutoML Vision to train models based on locally sourced data from community-based programs in the Philippines to detect referable DR.

Methods:

AutoML Vision (Google Cloud) models were generated based on previously acquired 5-field (macula centered, disc centered, superior, inferior, temporal macula) retinal images (17,237 images, 3447 eyes) from the Philippine DRSP using the recommended 240 node hours. Each individual image was labeled based on the International DR and DME classification and performed by four certified graders from a centralized reading center (RC), with secondary adjudication done by a senior retina specialist. Images for the initial model were split 8-1-1 for training, optimization and testing to detect referable DR [(refDR), defined as moderate nonproliferative DR or worse or any level of diabetic macular edema (DME). The AutoML platform provided model evaluation metrics by showing the area under the precision-recall curve (AUPRC), and the confusion matrix for initial model performance.

Internal validation of the autoML model was performed using a holdout image set that used a 2-field imaging protocol (disc and macula-centered, 225 eyes) using the same device in the same population, evaluated by the same RC. External validation was performed using another published image set (macula-centered, 207 eyes) which imaged a different patient population using tabletop retinal cameras. 3 This dataset is publicly available from the Eye Picture Archive Communication System (EyePACS) Kaggle, USA, and originates from the Joint Shantou International Eye Centre, China. Sensitivity, specificity, positive predictive value, negative predictive (SN, SP, PPV, NPV), accuracy and F1 score for refDR were calculated.

Results:

Training set distribution of DR severity by RC: no DR 75.6%, mild NPDR 11.5%, moderate NPDR 5.9%, severe NPDR 3.3%, PDR 3.7%. DME severity was: no DME 86.6%, DME 13.4%. In the training set refDR was present in 17.3%, non-refDR 82.7%. The model’s AUPRC was 0.995 with a precision and recall of 97% using a score threshold of 0.5. The confusion matrix showed true positives were 89%, true negatives were 99%, false positives were 1%, and false negatives were 11%.

Internal validation set distribution by RC: no DR 54.2%, mild NPDR 17.8%, moderate NPDR 9.8%, severe NPDR 3.3%, PDR 5.8%, ungradable 1.8%. DME severity was: no DME 62.7%, DME 6.2%, center involved DME 19.1%, ungradable 12.0%. RefDR was present in 39.1% of images. The SN, SP, PPV, NPV, accuracy and F1 scores were 0.96 (95% CI: 0.88-0.99), 0.98 (95% CI: 0.94-0.99), 0.96 (95% CI: 0.88-0.99), 0.98 (95% CI: 0.94-0.99), 0.97 and 0.96, respectively. In some false positive eyes, despite the image not approaching refDR thresholds, the eyes had a different pathology that warranted ophthalmology referral, such as macular holes and macular degeneration.

External validation set distribution (published Kaggle set): Non-refDR 57.5%, refDR 42.5%. The autoML model deployed on the imaging set using high-quality photos from tabletop cameras performed well, with SN, SP, PPV, NPV, accuracy and F1 scores at 1.0.

Conclusion:

This study demonstrates the accuracy and feasibility of low-cost autoML models for identifying refDR developed for a DRSP using handheld retinal imaging in a low-resource setting community program. The performance approaches and potentially outperforms published diagnostic accuracy metrics of commercial models used for DRSP. These data emphasize the use of local data in the development and optimization of machine learning models to potentially improve performance in the populations that they will be used. Furthermore, the use of autoML may increase access to machine learning models adapted for specific programs that are guided by clinicians to rapidly address disparities in patient care.


Effect of optic nerve sheath fenestration on visual recovery in pseudotumor cerebri syndrome
Yi Ling Dai, David Ramsey, Geetha Athappilly-Rolfe, Susan M. Tucker

Methods: Retrospective case series of patients with PTCS who underwent ONSF between 1998 and 2017.  Visual acuity (VA), papilledema grade, visual field (VF) mean deviation (MD), and retinal nerve fiber layer (RNFL) thickness were assessed prior to and after ONSF.


Results: Seventeen patients, all female,  aged between 13 to 36 years, underwent unilateral ONSF. Follow up averaged 40.1 months. Visual acuity remained intact in both eyes before and after surgery. The grade of papilledema improved in the operated eye from 3.3 ± 1.3 to 0.3 ± 0.7 and the non-operated, fellow eye from 3.0 ± 1.6 to 0.18 ± 0.4. RNFL thickness improved in the operated eye from 374 µm to 90 µm and from 306 µm to 108 µm in the fellow eye. There was an exponential rate of improvement in papilledema and RNFL thickness, with the greatest improvement occurring within the first 30 days. VF MD improved steadily in both eyes up to 12 months. 

Conclusions: ONSF leads to rapid improvement in papilledema and a steady recovery in VF. 


Evaluation of Normalized Bruch’s Membrane Opening-Minimum Rim Width and Retinal Nerve Fiber Layer Measurements for Visual Field Progression Prediction
Chhavi Saini, Jessica Sun, Julia Devlin, Christine Xu, Mengyu Wang, Scott H. Greenstein, Stacey Brauner, Lucy Shen

Methods: Baseline swept-source OCT imaging was obtained for POAG patients and unaffected healthy controls. POAG patients with baseline Humphrey visual field (HVF) mean deviation (MD) ?-12dB and ?4 reliable HVFs available after the baseline OCT scan were included. If both eyes were eligible, one eye was chosen randomly. Retinal nerve fiber layer (RNFL) thickness was measured automatically and BMO-MRW was manually calculated from 12 radial B-scans and normalized for BMO size. Global VF progression and progression in the paracentral region were defined by trend-based point-wise linear regression criteria using total deviation plots.

Results: Forty POAG patients and 39 controls were included. Average RNFL, minimum RNFL, average BMO-MRW, and minimum BMO-MRW were all significantly lower in POAG patients compared to controls (p<0.001 for all). POAG group had a baseline HVF MD of -4.1±3.1dB and an average follow-up time of 5.4±1.8 years. Fourteen POAG patients (35%) showed global HVF progression, while 7 (17.5%) showed paracentral progression. Average BMO-MRW and minimum BMO-MRW were significantly lower in POAG progressors vs. non-progressors (191.7±38.6µm vs. 248.1±59.8µm, p=0.003 and 83.9±40.9µm vs. 144.0±63.0µm, p=0.003, respectively). POAG patients with and without paracentral VF progression did not differ in baseline structural measurements. In receiver operating characteristics, the calculated area under the curve (AUC) for models to predict global VF progression were: 84.6% for average RNFL, 83.0% for minimum RNFL, 81.0% for average BMO-MRW and 78.6% for minimum BMO-MRW (p=0.79, comparing ROC area estimates from all curves).

Discussion: POAG patients with VF progression had significantly lower baseline BMO-MRW measurements compared to non-progressors. BMO-MRW measurements show comparable AUC for predicting subsequent visual field loss, compared to established structural parameters, suggesting utility of BMO-MRW in glaucoma monitoring.


Engagement of Patients with Diabetic Retinopathy by Telehealth Physician Extenders Improves Follow Up
Joseph Anaya, Rachel Munzar, Claudia Lasalle, Shiyoung Roh, David Ramsey

Methods: Established patients with DR who had not been seen within the last year were identified as LTF. A TH Physician-Extender performed an outbound call to LTF patients with an offer of symptomatic screening for new eye problems and assistance in scheduling an in-person clinic visit. Voicemail reminders to follow-up were left for patients who could not be reached. The intervention was performed over a period of 60 days, limited by staff availability. LTF patients who did not receive a call retained the ability to schedule an appointment by means of standard operating procedure (SOP). Appointment schedule rates were assessed 30 days after the intervention period (Program Assessment Timepoint). Call times were digitally measured, and labor cost of the intervention estimated.

Results: Out of 2514 patients with DR, 424 were identified as LTF. One hundred fifty seven patients were called within the 60-day study period (Intervention Group); the remaining 268 patients formed the SOP group. There were no statistically significant differences in age, sex, or ethnicity/race between the intervention and SOP groups. Of the 157 outbound calls, 66 (42%) reached patients; 80 (51%) went to voicemail (VM); 9 (6.0%) were unanswered and without voicemail; and 2 (1%) identified newly deceased patients. Of the 66 reached patients, 30 (45%) scheduled an appointment, 8 (12%) requested a call back, 25 (38%) indicated transfer or decline of further care, 8 (12%) requested a callback, and 3 (5%) indicated self-planned callback. At the time of program assessment, 48 (31%) intervention group patients had scheduled an appointment and 22 (14%) had completed a return visit. This was more than twice the respective rates of SOP group patients (14%, X=17.39, p<0.05, and 7%, X=5.44, p<0.05, respectively). The measured call duration was a mean 2.3 (SD 1.9) minutes, yielding an estimated $1.40 per call and $4.70 per appointment scheduled.

Conclusion: Employing TH Physician-Extenders to engage LTF patients can help improve the rate at which patients with DR return for eye care. Such an intervention program may be performed at a reasonable cost and may be considered in other clinical settings where patients often fail to access healthcare services.


In-State versus Out-of-State Ophthalmic Telemedicine Utilization during the COVID-19 Pandemic
Kanza Aziz, Jade Moon, Anjali Devgan, Alice Lorch, David Friedman, John Miller, Grayson Armstrong

Purpose: During the COVID-19 pandemic, regulatory changes in the United States allowed physicians to practice telemedicine across state lines. Data on the use of interstate ophthalmic telemedicine during the pandemic are limited. We aimed to evaluate the geographic characteristics and interstate utilization of telemedical care as compared to in-person care at a tertiary eye care center during the pandemic.

Methods: In this single-center, retrospective, cross-sectional study at Massachusetts Eye and Ear (MEE) from January 1 to December 31, 2020, clinical encounters were reviewed to extract patient and visit characteristics. In-person versus telemedical visit types were identified based on institutional categories and billing codes. Residential zip codes were used to estimate geographic characteristics of patients including distance from MEE and in-state versus out-of-state status. Pearson chi-squared tests were used to compare telemedical and in-person care groups. Further, a multivariate regression model was used to compare the groups by in-state versus out-of-state status after adjusting for age, sex, race/ethnicity, primary language, and level of education.

Results: A total of 1911 telemedical patients (2262 encounters) and 65763 in-person patients (147211 encounters) were included. The median (interquartile range; IQR) age of telemedicine patients was 61 (43-72) years, 62% of which were female. The median (IQR) age of in-person patients was 63 (49-72) years, 58% of which were female. Telemedicine patients included 14.7% (n=281) out-of-state patients, as compared to 12.0% (n=7876) out-of-state in-person patients (p<0.001). Regarding distance, 42.5% of telemedicine patients and 47.5% of in-person patients lived <10 miles (p<0.001), 41.9% and 41.3% lived 10-50 miles (p=0.611), 8.8% and 7.0% lived 51-100 miles (p=0.002), 3.9% and 2.5% lived 101-250 miles (p<0.001), and 3.0% and 1.8% lived >250 miles (p<0.001) away from MEE, respectively. After adjusting for covariates, there was no difference in the odds of using telemedical care as compared to in-person care between in-state and out-of-state patients (OR: 1.16, 95% CI: 0.99-1.35).

Conclusions: A significantly greater proportion of telemedical care, as compared to in-person care, was provided to out-of-state patients at a large eye care center during the pandemic. Moreover, a significantly greater proportion of telemedical care was utilized by patients living further away from the eye center. Out-of-state patients were as likely as in-state patients to use telemedical care as compared to in-person care, after adjusting for covariates. Proposals to revert to pre-pandemic policies requiring in-state telemedicine could set back forward progress made during the pandemic, including negative impacts on access to care and continuity of care for established patients. Expanded interstate telemedicine licensure and scope could help advance the efficiency and deployment gains seen during the pandemic.


Association of Social Media Metrics with Academic Citations of Ophthalmology Research
Gagan Kalra, Ravi Parikh, Grayson Armstrong

Aim: To study the association between social media metrics and academic citations of ophthalmology research.
Methods: Using Web of Science (WoS), digital object identifiers (DOI), publication year, and the number of citations were collected for all research items published in 10 major ophthalmology journals between the years 2011 and 2021. The exclusion criterion was the lack of DOI on WoS. The number of unique social media mentions (Twitter users sharing the research item) and Mendeley readers were collected from Altmetric. Pearson correlation coefficient and student’s t-test were used in evaluating correlations and comparing means between groups.
Results: A total of N=35,488 research items were included in this analysis. The number of Mendeley readers per item positively correlated with the number of citations in WoS (r=0.44, p=0.021). For research items published between 2016 and 2021, items mentioned by ?1 Twitter user had significantly higher academic citations compared to those with no Twitter mentions (8.5 vs 5.4, p<0.001). This difference was significantly larger for the 2011-2015 research items (25.5 vs 20.7, p<0.001).
Conclusion: We found a positive association between social media metrics and academic citations. Future research is needed to further elucidate this association.


Physician-to-Physician eConsultations to Ophthalmologists at an Academic Medical Center
Noha Sherif, Alice Lorch, Grayson Armstrong

Objective: Despite the known advantages of electronic consultation (eConsult) to patients, providers, and healthcare systems, including increased timely access to specialist consultation, decreased resource waste, and improved care coordination, the role of eConsults in ophthalmology has yet to be explored.1,2 We conducted a retrospective chart review to evaluate the feasibility and diagnostic accuracy of a physician-to-physician eConsults ophthalmology program and characterize the types of clinical questions and ophthalmic learning needs of healthcare providers.

Methods: Ophthalmology eConsults were reviewed to identify submission-to-response time, primary diagnoses made by eConsultants, and referral outcomes including symptom resolution, completion of follow-up appointments, and presentation to an emergency department (ED). eConsults underwent thematic review and categorized based on the type of clinical question asked, urgency, and ophthalmic condition addressed. Demographic data on patients and referring providers were also collected.

Results: One hundred ophthalmology eConsults were placed, and 100% were responded to by an ophthalmologist. An average of 1.6 days and a standard deviation (SD) of ±1.9 elapsed from the time an eConsult order was placed to the time the consultation was completed. Of the eConsults, 62% (n=62) were recommended for an in-person evaluation; of these, 48.4% (n=30) presented to an ophthalmologist. For patients who had an in-person follow-up, concordance of diagnoses between eConsultant and in-person evaluation was observed in 93% (n=28) of cases. On average, in person evaluations occurred within 28.9 (SD ±27.4) days of the eConsult response. The most common clinical inquiries were related to appropriate triage and referral (24.4%), management (22%), and diagnosis (19.7%). All eConsults were nonurgent. The most common ophthalmic condition addressed was chalazia/hordeola (14%). Only 5% of patients presented to an ED for the same ophthalmic concern addressed by eConsult.

Conclusions: Ophthalmology eConsultants can diagnose, triage, and manage nonurgent ocular conditions with high accuracy through electronic medical record review. eConsult programs can also facilitate timely access to specialty care.


Systemic Complement Activation in Non-Exudative Age-Related Macular Degeneration
Jonathan Lin, Stylianos Serghiou, Joan Miller, Demetrios Vavvas

Purpose. Although complement inhibition has emerged as a possible therapeutic strategy for certain forms of advanced age-related macular degeneration (AMD), it is unknown what portions of the complement pathway are dysregulated in AMD or when this dysregulation occurs relative to AMD stage; in this meta-analysis, we tested the hypothesis that complement dysregulation may occur during early/intermediate but not advanced non-exudative AMD.

Methods. Using PubMed, Google Scholar, and Embase, we identified articles from database inception to October 11, 2020 that reported systemic complement activation profiles in patients with early/intermediate non-exudative AMD or geographic atrophy (GA) and non-AMD controls. We meta-analyzed the data by generating multi-level random-effects models using restricted maximum likelihood estimation. Standard errors were adjusted using the Knapp-Hartung correction. Risk of bias was assessed using a modified Newcastle-Ottawa score.

Results. The seven meta-analyzed studies included 710 independent participants (mean per study: 101; standard deviation: 35) with 23 effect sizes. Compared with non-AMD controls, patients with early/intermediate non-exudative AMD (N=246) had significantly higher systemic complement activation as quantified by the levels of complement proteins generated by common final pathway activation (standardized mean difference [SMD]=0.52 [95% confidence interval (CI): 0.19 to 0.86]) and significantly lower systemic complement inhibition (SMD=-0.57 [95% CI: -1.07 to -0.07]). In contrast, there were no statistically significant differences in systemic levels of complement common final pathway activation products (SMD=0.34 [95% CI: -0.05 to 0.74]) or complement inhibition (SMD=-0.02 [95% CI: -0.29 to 0.24]) in patients with GA (N=178) versus non-AMD controls.

Conclusion. We provide evidence that systemic complement over-activation is a feature of early/intermediate non-exudative AMD; no such evidence was identified in patients with advanced non-exudative AMD, suggesting that complement inhibition may have improved efficacy for patients with early disease rather than for patients who have already developed GA as has been studied in multiple clinical trials to date.


Optical Coherence Tomography Angiography in the Evaluation of Vascular Patterns of Ocular Surface Squamous Neoplasia During Topical Medical Treatment
Despoina Theotoka, Zhiping Liu, Sarah Wall, Anat Galor, Ghada Al Bayyat, William Feuer, Jianhua Wang, Carol Karp

Optical coherence tomography angiography (OCTA) was utilized to examine changes in ocular surface squamous neoplasia (OSSN) vascular patterns over time in individuals treated with topical medical therapy. Ten individuals with OSSN diagnosed by clinical examination and high resolution (HR)-optical coherence tomography (OCT) were recruited. All individuals received topical immuno- or chemotherapy. OCTA images were obtained and analyzed at three points: presentation, mid-treatment and tumor resolution. Tumor metrics including tumor area (TA), tumor volume (TV), tumor depth (TD), and total tumor density (TTD) were calculated. Vessel area density (VAD) was also quantified within the OSSN, the subepithelium under and adjacent to the OSSN and the subepithelium of the uninvolved, contralateral eye. Vascular network changes were also subjectively evaluated. TA, TV, TD and TTD all significantly decreased with time (p<0.001). The mean VAD within the OSSN significantly decreased (p<0.001) between visits (presentation: 26.52±6.8%, mid-treatment: 7.19±5.88%, tumor resolution: 0.11±0.34%). The mean subepithelial VAD under the OSSN also decreased with time (23.22±11.03%, 20.99±5.99% and 19.58±7.08%), and after resolution the sub-tumor VAD (19.58±7.08%) was comparable to the subepithelial VAD in the contralateral eye (15.47±4.37%, p>0.05). The mean VAD in the subepithelium adjacent to the OSSN increased with treatment, then decreased significantly between mid-treatment and resolution (23.26±4.54, 28.30±7.43% and. 21.68±6.10%, p=0.009). Qualitatively, the tumor subepithelial vascular network was complex and dense but with tumor resolution appeared less tortuous and similar to the uninvolved eye. OCTA provided insight into the pathophysiology of tumor angiogenesis, showing decreased vascular density and normalization of vascular networks associated with tumor resolution.


In-person revisits among ophthalmic telemedicine patients during COVID-19
Jade Moon, Shefali Sood, Ravi Parikh, Grayson Armstrong

Purpose: Teleophthalmology may present a cost-saving alternative to in-person care and can improve access to care among disadvantaged groups. However, it remains unclear if telemedicine for ophthalmic concerns can completely replace in-person office visits. Duplicative care, in which a patient utilizing telemedicine requires an in-person appointment in a short interval, is additive and may be a disservice and an economic burden to patients, providers and the healthcare system alike. Thus, in this study, we explore the downstream impact of a visit initiated via telemedicine. We sought to compile the in-person revisit rate over 7 days after the initial visit. We investigated clinical and sociodemographic trends in utilization to identify risk factors for an early revisit.  

Methods: Retrospective review at Massachusetts Eye and Ear in 2020. Revisits were defined as in-person visits occurring within 1 week of a previous, related telemedical visit. Clinical characteristics of revisits were evaluated. Multivariate regressions of socio-demographic factors (age, gender, race/ethnicity, education, insurance status, employment, distance from ophthalmology clinic) were performed.

Results: Out of 2262 telemedicine (159 [7.0%] telephone, 2103 [93.0%] video) by 1911 patients, 139 (6.1%) resulted in a revisit. 32.7% of telephone visits resulted in a revisit compared to 4.1% of video visits (P=0.001). Vitreous opacities was the diagnosis with highest revisit rate (23.8%), accounting for 3.6% of revisit diagnoses and 0.9% of overall diagnoses (P=0.01). There was no difference in revisit rates across subspecialties. In multivariate analyses, Medicaid insurance status was significantly associated with a revisit (P<0.05).

Conclusion: While the in-person revisit rate after telemedical visits was low (6.1%), there was variation depending on telemedicine type, visit diagnosis and subspecialty service. Individuals with Medicaid insurance coverage were significantly more likely to revisit within 1 week. Telephone/ audio-only visits were significantly more likely to lead to an in-person revisit within 1 week compared to video visits. We find that in-person care utilization after a virtual visit is dependent on clinical and sociodemographic factors. More research is warranted on the downstream impact of telemedical care and may have significant implications for healthcare service delivery, cost of care, and visit burden.


Outcomes of Far Posterior Open Globe Injuries: The Case for Zone 4
Isaac Bleicher, Laurel Tainsh, Eric Gaier, Grayson Armstrong

Methods:

We performed a retrospective review of Z3 OGIs with at least 30 days of follow-up treated at a tertiary care center over a 10-year period. Far posterior OGIs (pZ3) were defined as those with wounds extending ?10mm posterior from the corneal limbus. Secondary surgeries, visual outcomes, and anatomic outcomes were compared between pZ3 and anterior Z3 (aZ3) eyes at presentation, the 3 months and the latest available timepoints after injury. Chi-square and Mann-Whitney U tests were used to compare pZ3 and aZ3 outcomes.

Results:

Of 300 eyes, 258 met inclusion criteria. Of these, 62% were classified as pZ3. At final follow-up, pZ3 eyes were more likely to be no light perception (NLP) (pZ3: 49%, aZ3: 33%), hand motion (HM) or worse (pZ3: 75%, aZ3: 57%) or <20/400 Snellen acuity (pZ3: 84%, aZ3: 65%) (p ?0.02). The effect of zonal group on HM and <20/400 acuity persisted even when excluding patients presenting NLP (p ?0.05). These relationships were also seen at 3 month follow-up for NLP, £HM, and £20/400 visual acuities (all p < 0.01) Compared to presentation, pZ3 eyes were less likely to have acuity improve (pZ3: 33%, aZ3: 44%) and more likely to have worsened acuity (pZ3: 19%, aZ3: 11%) (p<0.02). pZ3 eyes were more likely to become phthisical or be eviscerated/enucleated (pZ3: 56%, aZ3 40%, p<0.02). pZ3 and aZ3 eyes underwent means (±SD) of 1.5±1.6 and 1.4±1.5 secondary procedures, respectively (p>0.76).

Conclusion:

Eyes with OGIs extending ?10 mm posterior to the corneal limbus have poorer visual and anatomic outcomes compared to those limited to the more anterior Z3. While the potential for recovery in posterior OGIs necessitates careful assessment and emergent repair in all cases, further zonal categorization within zone 3 injuries may help improve prognostic precision and refine surgical approaches.


Longitudinal Satisfaction of Oral Sedation versus Standard of Care Intravenous Sedation for Ocular Surgery
Minali Prasad, Deniz Gaberz-Mah, Jeannie Xu, Daniella Zubieta, Marissa Fiorello, Manju L. Subramanian

Introduction: This is a follow up study to the oral versus intravenous (IV) sedation study (OIV) which measured patient satisfaction of oral to IV sedation for various types of eye surgery. Our aim was to longitudinally follow up to determine if patient satisfaction was maintained long term for both types of sedation.

Methods: Patients were interviewed with the same satisfaction survey given in the original study. Statistical analysis involved t-tests for noninferiority of the mean satisfaction scores (MSS). The total sample size estimated to obtain statistical significance with a power of 90%, 1:1 randomization of oral triazolam to intravenous midazolam and a noninferiority margin of 0.5 was 70, or about 35 in each anesthetic group. We compared the original MSS (OMSS) and the follow up MSS (FMSS) for each sedation type, and then we combined the OMSS and FMSS for both anesthetic groups to measure long term satisfaction.

Results: Participants were interviewed at a median of 1226 days (range 754-1449 days) from their original surgery. The OMSS was 5.20±0.84 for the oral treatment group (OTG) (n=42) and 5.26±0.67 for the intravenous treatment group (ITG) (n=39), and noninferiority between groups was demonstrated with a difference in MSS of 0.0582 (p=0.0054). The FMSS was found to be inferior with a score of 5.13±0.94 for the OTG and 5.54±0.65 for the ITG, with a difference in MSS of 0.4024 (p=0.2952). Satisfaction scores for both the OTG and ITG combined (p=0.3190) or the OTG alone (p=0.6596) did not differ between the OMSS and FMSS. Scores in ITG increased longitudinally (p=0.0052).

Conclusion: When separated into oral and intravenous groups, long term satisfaction scores were significantly better in the IV sedation group, indicating that while patients were similarly satisfied in the immediate post-operative period, those receiving IV sedation were more satisfied long-term than those receiving oral sedation. Additionally, non-inferiority was not established between oral and IV sedation. However, patient satisfaction did not differ longitudinally when both treatment groups were combined.


Recovery of Vision in Open Globe Injury Patients with Initial No Light Perception Vision
Karen Wai, Noha Sherif, Marisa Tieger, Racquel Bitar, Kevin Makhoul, Grayson Armstrong

Purpose: The recovery of vision in open globe injuries (OGI) presenting with no light perception (NLP) vision is poorly understood, and patients presenting with NLP vision often undergo primary enucleation or evisceration to mitigate the risk of sympathetic ophthalmia. We conducted a retrospective chart review to identify patient characteristics, OGI features, and surgical events that may predict the recovery of any vision in patients initially presenting with NLP and inform clinical considerations for management.

Methods: The Massachusetts Eye and Ear (MEE) OGI Database was used to identify patients with NLP vision at presentation from January 1999 to March 2022 with at least one week of follow up. The medical records of these patients were analyzed to identify demographic characteristics; zone of injury; time from injury to surgical repair; presence of retinal detachment, vitreous hemorrhage (VH), intraocular foreign body (IOFB), vitrectomy, hyphema, amongst 17 other features; and visual acuity (VA) at last follow-up appointment. Logistic regression was used to analyze the relationship between regained vision at most recent follow-up visit and the above dependent variables (Python 3.10.1).

Results: A total of 152 OGI cases with NLP at presentation that met inclusion criteria were identified. Of these, the majority were zone 3 injuries (68.4%). The average time of OGI to repair was 0.85 days, and the average length of follow up in our study was 3.5 years. Of patients with initial NLP vision with OGI included in our study, 29 cases (19.1%) regained better than NLP vision at most recent follow-up. Regained vision ranged from VAs >20/500 (17.2%) to LP (58.6%). Vitrectomy was positively associated with regained vision (B=3.1, p<0.001) while zone 3 injury was negatively associated with regained vision (B=-1.94, p=0.05). Previous intraocular surgery approached statistical significance for a negative association with regained vision (B=-1.36, p=0.08)

Conclusion: While OGIs cause severe ocular morbidity, a significant proportion of OGIs with initial NLP vision regained some vision during follow-up care. Vitrectomy was positively associated with return of vision, while zone 3 injuries were negatively associated. These findings argue for primary closure of OGIs given potential for return of vision rather than primary enucleation, even in the presence of NLP vision at presentation.


Causes and Complications of Intraocular Lens Exchanges: Case Series, Retrospective Review, and Biomechanical Analysis
Arjun Watane, Veshesh Patel, Nimesh Patel, Jayanth Sridhar, Nicolas Yannuzzi

Background: IOL exchanges may be carried out for a variety of clinical reasons, and several techniques for the insertion and/or exchange of a secondary IOL have been developed. Complications may include recurrent dislocations, suture erosions, haptic breakages, or haptic-optic avulsions.

Methods: A) Retrospective, multi-center case series in patients with scleral-sutured enVista MX60 dislocations secondary to eyelet fractures. B) Retrospective, single-center chart review of patients who underwent IOL exchange between 2014-2020. C) Suture geometry and force vectors were approximated and tested on enVista MX60 eyelets, comparing simple-pass and cow-hitch suture techniques. D) The force to break or avulse the haptic and haptic-optic junctions of CT Lucia 602 and AcrySof MA60 were compared.

Results:
A) 25 cases were identified, with 20 post-operative and 5 intraoperative eyelet fractures. Mean days until postoperative fracture was 96±125. New MX60s were sutured in 10 patients, of which 2 had repeat fractures.
B) 511 eyes of 489 patients were identified. Median time from initial procedure to IOL exchange was 47.5 months. Most common reason for exchange was IOL dislocation (46.6%). Iris-sutured more frequently dislocated than 4-pt scleral-sutured (P=.002), ACIOL (P=.01), 2-pt scleral-sutured (P=.03). Yamane more frequently had IOL tilt than ACIOL (P=.002), 4-pt scleral-sutured (P=.01), iris-sutured (P=.04).
C) Force to break enVista MX60 eyelets with simple-pass (2.42N; 0.11N) was significantly higher than with cow-hitch (1.40N; 0.26N) (P<.001) suturing.
D) Force to break haptics was significantly higher in CT Lucia (2.92 N; 0.53 N) than in the MA60 (1.04 N; 0.26 N) (P<.001). Force for haptic-optic avulsion of CT Lucia (1.10 N; 0.26 N) was significantly higher than in the MA60 (1.35 N; 0.18 N) (P=.048).

Conclusion: Scleral-sutured MX60 IOLs may experience intraoperative or postoperative eyelet fractures resulting in lens dislocation. Iris-sutured IOLs were associated with subsequent dislocations. Yamane scleral-fixated IOLs were associated with IOL tilt. Simple-pass suture method allowed a greater force before the enVista MX60 eyelet fractured compared to cow-hitch method. CT Lucia haptics required 2.8x greater force to break its haptic than the MA60 haptic, a clinically significant finding. 


Association of capillary plexus vessel density with risk of proliferative diabetic retinopathy in eyes with advanced nonproliferative diabetic retinopathy
Konstantina Sampani, Mohamed Elmasry, Paolo Silva, Lloyd Paul Aiello, Jennifer K. Sun

The risk of progression in eyes with PDR in 1 year increases with increasing DR severity from 15% in eyes with moderate NPDR to 52% in eyes with severe NPDR [1]. There is, however, substantial individual variability that is not reliably addressed by current DR staging systems. Thus, there is an unmet need to identify reliable, validated biomarkers to enhance ETDRS classification system [2] and establish individualized protocols.

In this study, treatment-naïve eyes with moderate to severe NPDR within the ETDRS fields on ultrawide field photographs were imaged using 3x3 mm scans on OCTA (AngioVue). The study group was 104 eyes (70 subjects) with moderate (N=49), moderately severe (N=23) or severe (N=32) NPDR at baseline and a mean follow-up (f/u) of 2.5±1.5 yrs. Seventeen eyes, all with baseline severe NPDR, developed PDR within 1.3±1.2 yrs.  In the whole cohort, no significant difference was found in mean baseline OCTA SCP, ICP or DCP VD among eyes with any step DR severity worsening or between eyes that did and did not progress to PDR [any-step/PDR progression: SCP: p=0.4/0.2; ICP: p=0.1/0.3; DCP: p=0.2/0.5]. In the whole cohort, time to event analysis, adjusted for age and OCTA signal strength index, demonstrated no significant relationship between mean baseline OCTA SCP, ICP or DCP VD and the risk of progression to PDR over the f/u period of time [SCP: HR=0.87, p= 0.05; ICP: HR=1.04, p=0.63; DCP: HR=1.01, p=0.84]. In the subgroup with severe NPDR eyes, increased mean baseline OCTA DCP VD, but not SCP or ICP VD, was significantly associated with increased risk of progression to PDR [SCP: HR=0.89, p= 0.13; ICP: HR=1.14, p=0.20; DCP: HR=1.46, p=0.01].

The findings from this study suggest that retinal capillary plexus VD is potentially associated with risk of PDR onset in eyes with severe NPDR. The association of higher PDR risk with higher DCP VD could be due to greater DCP perfusion from vascular shunting and capillary remodeling that precedes neovascularization, but further prospective, longitudinal studies are needed to elucidate the understanding of the relationship between OCTA VD and the risk of PDR onset.


Single-cell transcriptional profiling of the adult murine lacrimal gland in health and disease
Jacob Heng, Briana Winer, Loyal Goff, Jeremy Nathans

Objective: To characterize the single-cell transcriptome of the adult murine lacrimal gland in wild-type mice and mouse models of Sjogren’s syndrome

Methods: Extraorbital lacrimal glands from adult BALB/cJ male (n=4) and female (n=4) mice, as well as adult NOD.B10.H2b male mice (n=2) and adult MRL-lpr female mice (n=2) were dissected and dissociated for droplet-based single-cell RNA sequencing (sc-RNAseq, 10X Genomics). All mice were 4 months of age at the time of experiment. Cell type annotation and differential gene expression analysis were performed using the Monocle R package. Gene set enrichment analysis (GSEA) was carried out using the molecular signatures database (MSigDB) Hallmark pathways. Cell type markers and other genes of interest were validated using immunofluorescence.

Results: A total of 102,431 cells were profiled by sc-RNAseq and included all known major cell types in the lacrimal gland. There was notable sexual dimorphism in the expression of the secretoglobins, with most secretoglobins being preferentially expressed in male lacrimal glands. Analysis of the immune cell clusters identified cells of monocyte lineage, plasma cells, B cells, T cells, NK cells, mast cells and a cluster of GM-CSF-producing innate lymphoid cells (ILCs). Cells of monocyte lineage exhibited dynamic transcriptional profiles that included a sub-population of Cd163-positive, Pf4-positive macrophages. GSEA revealed significant upregulation of interferon signaling in NOD.B10.H2b and MRL-lpr lacrimal glands.

Conclusions: sc-RNAseq revealed sexual dimorphism in gene expression and diverse immune cell types in lacrimal glands of wild-type mice and mouse models of Sjogren’s syndrome. Interferon signaling appeared to be significantly upregulated in two mouse models of Sjogren’s syndrome. The role of novel immune sub-populations, including innate lymphoid cells and Cd163-positive macrophages, in lacrimal gland disease remains to be elucidated and warrants further investigation.


Comparison of surgical outcomes of Gore-Tex scleral sutured posterior chamber intraocular lens (PCIOL) versus sutureless scleral fixated PCIOL using the modified Yamane technique versus anterior chamber intraocular lens (ACIOL)
Yingna Liu, Paul Zhou, Andre Witkin, Chirag Shah

Background: For the management of poor capsular support, ACIOL, scleral sutured PCIOL and sutureless scleral fixated PCIOL using the modified Yamane technique are all viable options.1–3 However, few studies have directly compared the visual outcomes of these techniques.

Methods: Retrospective review of patients who had vitrectomy with either the Envista or Akreos scleral sutured PCIOL, CT Lucia sutureless scleral fixated PCIOL, or ACIOL with a minimum follow-up of 6 months was conducted. Pre- and postoperative best-available visual acuity (VA) and complications were recorded.

Results: Between eleven retinal surgeons, there were 47 sutured IOLs, 36 scleral fixated IOLs, and 59 ACIOLs.  VA improved in all groups, with no difference between groups.  The logMAR improvement from baseline to one year was 1.02, 0.97, and 1.17, for a final Snellen equivalent of 20/45, 20/45, and 20/62, respectively, in the sutured PCIOL, sutureless PCIOL, and ACIOL groups.  Complications rates, in general, were similar, with some significant difference.  Immediate post-operative transient vitreous hemorrhage and/or hyphema was more common after sutureless scleral IOL fixation than scleral sutured IOL (19.4% vs 2.1%, p = 0.008), but similar to the rate with ACIOL (8.5%).  CME was more common after ACIOL than after scleral sutured IOL (28.8% vs. 8.5%, p = 0.009), but similar compared to sutureless scleral fixation (16.7%).  Rates of elevated IOP (17.0%, 13.9%, 18.6%) and corneal edema (8.5%, 2.8%, 5.1%) were similar between scleral sutured, sutureless scleral fixation, and ACIOL, respectively.

Conclusion: Scleral sutured PCIOLs, sutureless scleral fixated PCIOLs, and ACIOL all result in similar visual improvement one year after surgery with similar complication rates.


An Artificial Intelligence (AI) Model for Screening Computed Tomography (CT) Imaging for Thyroid Eye Disease
Lisa Lin, Paul Zhou, Soomin Jeon, Jonathan Lu, Synho Do, Grace Lee

Purpose: Thyroid eye disease (TED) is an autoimmune disease with an array of clinical manifestations, which can be complicated by compressive optic neuropathy. It is important to identify patients with TED early to ensure close monitoring and treatment to prevent potential permanent disability or vision loss. Currently, there is no standardized approach to screening for TED, which can delay such treatment. Deep learning artificial intelligence (AI) algorithms have been utilized in ophthalmology and in other fields of medicine to detect disease.  This study aims to introduce an AI model to evaluate orbital computed tomography (CT) images for the presence of TED and potential compressive optic neuropathy.

Methods: Retrospective review of patients with dedicated orbital CT scans and with oculoplastics examination over a 10- year period at a single academic institution was performed. CT scans for patients with TED and normal controls were included, while those with other diagnoses, such as tumors or other inflammatory processes were excluded. CTs were preprocessed and adopted for the two-dimensional (2D) ResNET training to diagnose TED and compressive optic neuropathy.

Results: A total of 1368 photos from 95 CT scans (94 patients) were used to develop the AI model. The model can discern the region of interest of the CT scans with an accuracy of 93.7%,  diagnose TED versus normal control with an accuracy of 92.4% (F1 score of 0.94), and can identify TED with moderate to severe optic neuropathy with an accuracy of 85.1%.

Conclusion: The AI model developed in the study can accurately detect TED and further detect TED with optic neuropathy based on orbital CT. With further optimization and validation, this TED AI model could help guide frontline healthcare providers in the detection of TED and help stratify the urgency of a referral to an oculoplastic surgeon and endocrinologist


Global Current Practice Patterns for the Management of Traumatic Optic Neuropathy and Orbital Floor Fractures
Michael Fliotsos

Introduction: Facial trauma can have a variety of sequelae, including orbital floor fractures and traumatic optic neuropathy. Although these injuries are common, their management is controversial. Our goal was to summarize management practices at international high-volume eye trauma centers and identify areas of alignment and discrepancy.

Methods: An online survey was conducted from August 2020 to January 2021 to

capture management paradigms for orbital floor fractures and traumatic optic neuropathy at high-volume eye trauma centers globally. Trauma experts responded on behalf of general practices at their institutions.

Results: Response rate was 85.7% (36/42). Of the responding institutions, 54.5% (n=18/33) routinely administered systemic antibiotics for orbital floor fractures, with oral amoxicillin and clavulanic acid (n=13/18, 72.2%) and oral cefalexin (n=4/18, 22.2%) most common. Only 27.3% (n=9/33) of respondents administered systemic steroids, most of whom utilized oral steroids (n=7/9, 77.8%). Respondents considered the optimal timing for surgical repair of orbital floor fractures to be on average 11.46 ± 8.46 days from presentation, with oculoplastic surgeons participating in performing the repair at 72.7% (n=24/33) of responding institutions and oral and maxillofacial surgeons at 57.6% (n=19/33) of centers. Most experts did not routinely admit patients with isolated orbital fractures on presentation (n=26/33, 78.8%), however, 48.5% (n=16/33) of respondents admitted patients after the repair for an average of 3.50 ± 1.96 days. For cases of traumatic optic neuropathy, 63.6% (n=21/33) of respondents routinely administered systemic steroids, with 57.1% (n=12/21) preferring intravenous administration. High dose steroids (methylprednisolone 1 gm daily) were used by 36.4% (n=12/33) of responding centers and moderate doses (prednisone 1 mg/kg daily) were preferred by 21.2% (n=7/33).

Conclusions: While experts largely agreed on administering antibiotics for orbital floor fractures and steroids for traumatic optic neuropathy, responding trauma experts differed on many aspects of optimal management paradigms. Further investigations are needed into best treatment regimens to institute evidence-based care in ophthalmic trauma.


Perspectives of Resident and Attending Physicians on Common Ethical Dilemmas in Research
Michael Fliotsos

Methods: An online survey was distributed to 8 ophthalmology residency programs. Controversial vignettes (14) were developed based on common research scenarios regarding mentorship, honorary/ghost authorship, general authorship, and collaboration. Respondents assessed the ethicality of the situations on a Likert scale. 

Results: The response rate was 35.3% (77/218). 37.7% (n=29) were residents and 62.3% (n=48) attendings. 50.6% of responses identified ghost/honorary authorship situations as unethical, while 44.2% identified mentorship scenarios as unethical. More responses identified general authorship scenarios (i.e., order of authors) as ethical (35.1%) compared to other categories. Residents were more likely to report behavior as unethical than attending ophthalmologists (p=0.032).

Conclusions: There is a need for more prescriptive guidelines for authorship and mentorship ethics at all training levels to ensure consistency, fairness, and integrity of research. 


Risk of Posterior Subcapsular Cataract Development among Uveitis Patients Treated with Topical Difluprednate
Soraiya Thura, Priya Janardhana

Purpose: We sought to investigate the risk of posterior subcapsular cataract development (PSC) among patients treated for uveitis with 0.05% topical difluprednate.

Methods: This is a retrospective cohort study assessing patients with uveitis who were treated with topical difluprednate from January 1st, 2016 through October 30th, 2020 at an academic medical center. Thirty-four patients (fifty-one eyes) were evaluated through a chart review to collect clinical data which was then analyzed. Patients that received cataract surgery before being treated with difluprednate for uveitis were excluded.

Results: Sixteen eyes (31.4%) were observed to have had development or progression of PSC. For the eyes that developed a PSC, both the duration (mean of 21.7 weeks versus 15.1 weeks) and total cumulative dose (mean of 352 drops versus 288.3 drops) of topical difluprednate was higher compared to patients who did not develop a PSC. The average time to development of a PSC noted on eye exam after the initiation of difluprednate was 46 weeks. Nine eyes (30%) were noted to have ocular hypertension at some point during the follow-up period, and the eye pressures were controlled with pressure-lowering drops.

Conclusions: In our cohort of patients, we have found that an increased duration and cumulative dosing of difluprednate treatment may contribute to an increased risk of posterior subcapsular cataract formation in uveitis patients.  Confounding factors between the patients that developed PSC versus those that did not develop a PSC were similar and not statistically significant. This is one of only a few studies that have assessed the impact of topical difluprednate on cataract formation in an adult uveitis population, and the only study that assesses the timeline of PSC development in phakic patients while on difluprednate drops and the dosing correlated to this. We believe these findings will provide further clinical insight into the safety profile and dosing of difluprednate, as well as contribute to our knowledge of topical steroid treatment regimens for patients with uveitis.