On this episode of Retina Mentor Moments, John W. Kitchens, MD, sits down with Durga Borkar, MD, MMCi, a fellowship-trained vitreoretinal surgeon and attending physician at Duke University, to discuss career path advice for medical trainees and the potential role of artificial intelligence (AI) in managing retinal diseases.
Applications of clinical informatics in ophthalmology
Dr. Borkar first learned data analytics and basic computer programming at university and took time off during medical school to pursue clinical outcomes research leveraging large datasets. This laid the foundation for her to then obtain a Master’s degree in clinical informatics.
She remarked that having a background in quantitative research early on in her medical career helped propel her forward to pursue big data research projects. Ophthalmology has many helpful algorithms, noted Dr. Borkar; however, in her opinion, with the exception of diabetic retinopathy, many of these algorithms have been difficult to implement in clinical settings.
She explained that this is where clinical informatics could be useful, as it is largely concerned with taking a data set, applying an algorithm to it, and then integrating the algorithm into a user interface that a surgeon could realistically use in clinical practice.
Advice for residents considering different subspecialties
Dr. Borkar emphasized that during residency, she was heavily involved in cornea research and only realized that she wanted to pursue a retina fellowship 3 months before she had to apply for fellowships. Consequently, she recommended that residents make an equal effort in all their rotations.
Networking and making connections during medical training
Additionally, in her experience, many students tend to go into residency confident that they know exactly what subspecialty they want to focus on (she thought this as well). However, she advised trainees to be open-minded and to make connections in all of their rotations because their feelings might change towards the end of residency. Dr. Kitchens added that it’s never too late to make that change; as long as residents don’t burn bridges, the attending physicians will respect the decision to change subspecialties.
She mentioned that retina distinguishes itself from other subspecialties in that many private practices are heavily involved in research, so they tend to operate similarly to university-led academic spaces. Dr. Borkar also mentioned that since retina fellowships are primarily in academic institutions, outgoing fellows can have a unique opportunity to meet and network with a relatively small group of people spanning the country. In addition, asking friends who are current fellows at different locations about their experiences is a helpful approach to understanding the workplace culture of a job.
When it comes to academic jobs, it’s never too early to start reaching out for a position, she noted. Academic institutions tend to take longer to post a job and hire, even if they are actively considering adding a new position. As such, she recommended that if a trainee is interested in a specific academic job or interested in working in a particular location, it is important to reach out as early as possible—even in the first year.
Having reliable mentors is critical after fellowship, remarked Dr. Borkar, because training is a time when physicians are on a straight path and their job is to become a retina surgeon, for example. However, once the fellowship is complete, surgeons must mold their own path—which can be overwhelming. Consequently, having senior colleagues who can act as guides in finding and landing the first ophthalmology job is helpful.
Flexibility in career paths during medical training
Of note, Dr. Borkar had been an attending physician for 6 to 7 months prior to the coronavirus disease 2019 (COVID-19) pandemic. While the hospital she worked at was busy with urgent cases, the overall case volume was lower due to the pandemic, as elective surgeries were initially delayed.
As a result, she took this opportunity to get a Master’s degree in clinical informatics, which allowed her to develop a more formal skill set to pursue her research interests. Now, she tells fellows and new attendings to view their career paths with a flexible mindset and take advantage of opportunities as they come.
Research on private equity acquisitions and patient care/access
Dr. Borkar recently received a grant from the National Institute for Health Care Management (NIHCM) Foundation to research the spillover effects of private equity (PE) acquisitions of physician practices on local market competitors (i.e., academic institutions).1
While PE acquisitions have grown rapidly in healthcare, there is still a dearth of knowledge on the effects on providers in the same market that are not acquired.2 This study will use Medicare claims data to examine the impact of PE acquisitions on patient access, with a focus on complex surgical cases in ophthalmology.
Dr. Borkar is working with a healthcare economist at Brown University, Yashaswini Singh, PhD, to assess how the patient journey varies, for example, in retinal detachment care in areas that have a high saturation of PE firms. She added that Dr. Singh is an expert in this area, as she studies how PE acquisitions impact patient access and care across medical subspecialties.
The investigators will research, for instance, how PE acquisition may impact the way a retinal detachment patient would access care in terms of the number of visits they have with a retina specialist, the number of emergency room (ER) visits, and the time to retinal detachment surgery.
This project came about, in part, from internal data at Duke University, where they have seen a change in the number of urgent cases that are referred, reoperations, and diabetic tractional retinal detachments since a local retina group was acquired by PE. She added that this local trend doesn’t necessarily translate to the national level, so this study seeks to elucidate national trends in PE acquisitions and patient care.
Anecdotes on private equity practices
Dr. Kitchens noted that his practice was acquired by a PE firm earlier this year. While discussing PE on a vit buckle panel, one of the panelists noted that being acquired by a PE firm felt similar to going from owning a home to renting an apartment.
He added that, at times, this work culture change can have an emotional impact on surgeons. Dr. Kitchens emphasized that at his PE practice, the PE firm does not dictate to physicians how to take care of patients, such as drugs that patients are prescribed or medical procedures.
Dr. Borkar remarked that she has heard this perspective from her surgical colleagues and peers. Currently, there is an abundance of anecdotal information from surgeons in and outside of PE groups, and so she hopes this study will provide information and clarity on patient access and care.
A recent issue that Dr. Kitchens’ practice has encountered is hospitals rescinding on-call operating room (OR) hours for vitreoretinal surgery because they can’t afford the after-hours staff and are focused on other more profitable surgeries.
As a result, his practice doesn’t have a location to operate on call, which he is concerned will translate to local academic hospitals seeing more retinal detachment referrals and emergency cases that occur outside of regular work hours.
Ultimately, Dr. Borkar’s research isn’t meant to make a statement about the physicians in PE groups, but instead describe how PE practices impact the overall healthcare system.
Creating algorithms with large EHR datasets
Outside of healthcare delivery, Dr. Borkar’s research is focused on clinical outcomes using large electronic health record (EHR) datasets.
For example, she did research in collaboration with Verana Health (i.e., the American Academy of Ophthalmology’s technology partner for the Intelligent Research in Sight [IRIS] Registry) to analyze the clinical features and anti-vascular endothelial growth factor (VEGF) treatment outcomes in patients diagnosed with coexisting geographic atrophy (GA) and neovascular age-related macular degeneration (nAMD).3
Additionally, she recently contributed to the FARETINA-DME Study, which evaluated real-world data from over 12,000 eyes on injection frequency and early clinical response of diabetic macular edema (DME) patients to faricimab.4
Implementation of AI in clinical practice
Dr. Borkar has also researched how AI algorithms are developed using real-world images—specifically for GA progression algorithms. She noted that algorithms using real-world data can differ from algorithms developed from more “clean-cut” clinical trial datasets because images captured in the clinic can have imperfections that impact their readability.
When it comes to implementing AI algorithms into clinical practice, Dr. Kitchens noted that he sees close to 80 patients a day, so he doesn’t have the time to upload images to the cloud, wait for them to be analyzed, and then have them sent back the next day to share the results with patients.
Dr. Borkar highlighted that part of integrating this technology into clinical practice is having device manufacturers add these algorithms and their associated cloud data to existing devices so surgeons can access them quickly. She added that once AI algorithms become more widely available, they will likely be an added subscription cost to existing imaging systems.
Conclusion: Favorite retina meeting
Dr. Borkar noted that The Retina Society is her favorite meeting to attend because it is smaller and fosters a space for meeting and networking with colleagues—which is important to her because she is still fairly early in her career.
This year, she will go to the American Society of Retina Specialists (ASRS) meeting in Stockholm, Sweden, and will present the Phase 1/2a findings from a clinical trial (NCG04744662) sponsored by ONL Therapeutics. The clinical trial evaluated the safety and efficacy of two doses (50μg and 200μg) of two intravitreal injections for ONL1204 ophthalmic solution dosed 3 months apart, targeting neuroprotection from GA.5
Additionally, she will attend The Retina Society meeting in Lisbon, Portugal, this year to present FARETINA-DME data on real-world anatomical outcomes with imaging data taken from the IRIS Registry.
