Lessons for the Practicing Ophthalmologist

Wills Eye Hospital 66th Annual Alumni Conference

Mark F. Pyfer, MD; Michael P. Rabinowitz, MD; David H. Perlmutter, MD; Murtaza K. Adam, MD; Margaret A. Greven, MD; Alessandra K. Intili, MD; Michael L. Dollin, MD; Christopher J. Brady, MD; Julia A. Haller, MD


July 22, 2014

Editorial Collaboration

Medscape &

This feature requires the newest version of Flash. You can download it here.

Editor's Note: The Wills Eye Hospital Annual Alumni Conference is a 3-day conference designed for comprehensive ophthalmologists, specialists, and allied health personnel. The conference reviews clinical, diagnostic, and therapeutic approaches to eye problems related to all ophthalmic subspecialties. Mark F. Pyfer, MD, Chair of the Wills Eye Hospital 66th Annual Alumni Conference, convened afterward with his Wills Eye Hospital colleagues to discuss for Medscape readers some of the highlights of this year's meeting.

Prostaglandin Periorbitopathy

Dr. Pyfer: Greetings and welcome. I am Dr. Mark Pyfer, Chair of the Wills Eye Hospital 66th Annual Alumni Conference, which was recently held in Philadelphia. This is a collaborative effort between Medscape and Wills Eye Hospital. We are going to be speaking to some of our residents and fellows who presented original research in a very well-received free paper session that started off the conference. First, Dr. Michael Rabinowitz, from oculoplastics, will tell us about his work on prostaglandin-associated periorbitopathy.[1]

Michael P. Rabinowitz, MD: We are looking at prostaglandin periorbitopathy, which is a syndrome in which the eye sinks into the eye socket because of relative fat loss within the eye socket and around the eye. You get a sinking of the upper eyelid where the superior sulcus becomes hollow. All of the literature says that the lid can droop as a result of this syndrome. However, our study, in contrast to the vast majority of the literature, is looking at this in patients treated in only one eye. So, we are able to distinguish the changes associated with this syndrome from the changes in the normal aging eyelid. We are finding that the eyelids are almost universally retracted as opposed to drooping. This is an important factor because it can distinguish this syndrome from the aging eyelid.

Dr. Pyfer: Is this associated with prostaglandin use?

Dr. Rabinowitz: Correct.

Dr. Pyfer: What should general ophthalmologists or glaucoma specialists be aware of when prescribing prostaglandin analogs for their patients with glaucoma?

Dr. Rabinowitz: The most important thing is to be aware of this syndrome. Be aware of the changes in the eyelids and the orbit, and also the relative enophthalmos that can result from the drop. These are important considerations especially in aging patients who already have a loss of orbital fat. When they are lying flat, in the supine position, the globe can sink back toward the orbital apex, causing separation between the lid and the cornea, which can result in exposure.

Dr. Pyfer: Is this especially noticeable when only one eye is being treated?

Dr. Rabinowitz: It is.

Automated Fundus Photography Analysis

Dr. Pyfer: Next, we have Dr. David Perlmutter, who presented his work on automated analysis of the vasculature in fundus photographs.[2] David, can you tell us about your work?

David H. Perlmutter, MD: We collaborated with the Department of Family Medicine at Thomas Jefferson University Hospital, and our goal was to use a software program called AVRnet to look at vessel changes in fundus photos. These were communicated through telemedicine, and we are hoping that we can show that changes in the vessel measurements will be useful measures of control of hypertension and diabetes.

Dr. Pyfer: Did the automated analysis agree with experts' opinions on the fundus vasculature?

Dr. Perlmutter: The values that we are measuring did agree with the current published values in the literature. We think that this software tool will be useful going forward in a long-term study. Such a study in the family medicine department is going to look at vascular risk factors, as well as such unique tools as finger capillaroscopy. We are hoping to correlate the vessel measurements with these risk factors and these additional measurements.

The Smartphone as an Ophthalmoscope

Dr. Pyfer: I am now joined by Dr. Murtaza Adam, who presented innovative work on the smartphone ophthalmoscopy reliability trial.[3]

Murtaza K. Adam, MD: This study came about because of work done at Emory, looking at the use of fundus cameras (which can cost $4000-$5000) in an emergency department (ED). It's hard for ophthalmologists to make it into the ED during a busy clinic, and sometimes, in patients with vision-threatening or severe neurologic conditions, important findings in their eyes are missed and this can be very important in patient care. This study basically validated that these cameras can be used to find significant eye disease in patients in whom it would otherwise be missed, because most doctors are not skilled in direct ophthalmoscopy.

After that study, we learned that a simple smartphone with a camera mode and a $200 lens that most ophthalmologists carry in their office or call bag can be used to take high-quality images of the ocular fundus. So, we wanted to find out whether these high-quality images could be produced and be diagnostically useful. We also evaluated whether people who were previously untrained in direct ophthalmoscopy could learn this technique and take useful photos.

We used the Wills Eye ED, which is a unique place; there are only 3 EDs like it in the country. We enrolled 79 patients and proved that with a skilled operator, you can take images of the same quality as a conventional fundus camera, and those who are not skilled learn relatively quickly that with this technique, they can achieve about a 50% hit rate of critical findings, which is much better than most doctors can achieve with a direct ophthalmoscope.

Dr. Pyfer: You were able to take impressive images of pathology. Were the patients dilated for this?

Dr. Adam: Yes, and that is one downside of this technique. To get an adequate view of the fundus, you need to dilate the eyes. Next we will see whether we can get an image with an undilated view, using different lenses and maybe a different light source on a smartphone. We are working with Drexel University now on that project.

Oral Eplerenone for Central Serous Chorioretinopathy

Dr. Pyfer: Next, I'm joined by Dr. Margaret Greven, who presented her work using oral eplerenone to treat central serous chorioretinopathy.[4] Margaret, can you tell us about your study?

Margaret A. Greven, MD: We did a retrospective chart review of patients who had received eplerenone, which is a selective aldosterone antagonist for the treatment of chronic central serous chorioretinopathy. We reviewed the charts of 12 patients who had received this medication to see the effect on subfoveal fluid and vision over time.

Dr. Pyfer: This is usually a self-limiting disease that we typically don't treat but instead use watchful waiting. How does this treatment have an advantage over traditional observation?

Dr. Greven: In our series of patients, all had subretinal fluid for at least 3 months, but the average was 6 months (range, 3-24 months). All patients had subretinal fluid for a long period of time, and some had undergone other treatments to try to resolve the subretinal fluid. Eplerenone was effective in reducing subretinal fluid and improving vision in these patients.

Dr. Pyfer: Were you able to treat acute central serous as well?

Dr. Greven: One arm of the study looked at patients with acute central serous chorioretinopathy, and they also showed some benefit from the eplerenone.[5]

Dr. Pyfer: Were there any side effects from the medication?

Dr. Greven: Not in our series of patients, although eplerenone is known to potentially cause hyperkalemia, especially in patients with renal failure or compromise.

Femtosecond Complications: They're Aren't Many

Dr. Pyfer: Next, we are joined by Dr. Alessandra Intili, who presented her work reviewing the outcomes of femtosecond laser-assisted cataract surgery performed by residents here at Wills Eye Hospital.[6] Can you tell us a little bit about your study?

Alessandra K. Intili, MD: We reviewed the first 100 cases performed by residents here at Wills Eye and found a very low incidence of postoperative and intraoperative complications. In fact, we didn’t have any posterior capsule ruptures. We had no dropped nuclei or anterior capsule tears.

Dr. Pyfer: What does this tell us about incorporating femtosecond laser-assisted surgery in a teaching program?

Dr. Intili: It tells us that it is safe to do and that the outcomes are comparable to the manual cataract surgeries that the residents perform here on a daily basis.

Dr. Pyfer: Is there any advantage to having residents start with femtosecond-assisted surgery or is it too soon to tell?

Dr. Intili: The most complicated part for a new resident is performing the anterior capsulotomy. So, if we could begin residents on a procedure that performs the capsulotomy for them, they may be able to improve their manual dexterity with phacoemulsification and then move on to the capsulotomy once their dexterity has improved.

Dr. Pyfer: Which laser did you use in this trial?

Dr. Intili: We used the OptiMedica (Sunnyvale, California) laser.

No Talking During Intravitreal Injections

Dr. Pyfer: Next, we are joined by Dr. Michael Dollin, one of the retina fellows here at Wills Eye Hospital, who presented his work on instituting a "no talking" policy during intravitreal injections.[7] Michael won the McDonald Award for the best fellow paper presented at the conference. Michael, can you tell us about your work?

Michael L. Dollin MD: We treat a lot of macular degeneration, diabetic macular edema, and macular edema from other common causes here at the Wills Eye Hospital on the retina service, and most of these conditions are treated with intravitreal injections.

Over the past 4 years, we performed more than 120,000 intravitreal injections. During the past 2 years, we have instituted a "no talking" policy to minimize speech during intravitreal injections. The idea is that by minimizing speech, you minimize the aerosolization of respiratory droplets and you create a more sterile field with less risk for contamination and infection in the eye (endophthalmitis). We compared endophthalmitis rates from the past 2 years with those of the preceding 2 years (before implementing the "no talking" policy), and we found that during the "no talking" period not only were the overall cases of endophthalmitis approximately 50% lower but also the infections caused by oral flora were reduced by more than 50%.

Dr. Pyfer: These infections are very rare but can still be devastating when they occur. What is the underlying infection rate for most centers using intravitreal injections with standard sterile policy?

Dr. Dollin: Thankfully, it is very low. Depending on the study, it is in the range of 1 in 1000 to 1 in 3000.

Dr. Pyfer: Were you able to show that the flora that caused the infections were oral in origin?

Dr. Dollin: They were. Oral flora are typically streptococcal species, and during the "no talking" period, we had a significant reduction in cases of endophthalmitis caused by these oral flora species.

Dr. Pyfer: Do you recommend that retinal specialists in the community institute a "no talking" policy while they are giving injections?

Dr. Dollin: I certainly think that minimizing speech during intravitreal injections would be a positive thing for both the patient and the doctor.

Crowdsourcing Retinal Image Evaluation

Dr. Pyfer: We are joined by Dr. Chris Brady, who completed his residency here at Wills Eye and is currently involved in the Retina Fellowship Program. He presented work on using Web-based crowdsourcing to evaluate fundus photography in diabetic retinopathy.[8] Chris, can you tell us a little bit about your work?

Christopher J. Brady, MD: Our work was based on the observation that screening rates for diabetic retinopathy are dismal. We need to be screening twice the number of patients that we are screening now. If we were to do that, there would be a huge human resources crunch, particularly in emerging economies where rates of diabetes are going through the roof.

Dr. Pyfer: What is crowdsourcing?

Dr. Brady: Crowdsourcing is the division of a task into small bits, and sort of sprinkling it out to many different -- perhaps anonymous -- people on the Internet. It doesn't have to be on the Internet, but in general, crowdsourcing now involves people using the Internet to do tiny pieces of work for a small fee.

Dr. Pyfer: These people could be anywhere in the world as long as they have Internet access?

Dr. Brady: They could be absolutely anywhere in the world. On the Amazon interface that we used for this project, they were all over the world.

Dr. Pyfer: Are these novice interpreters with no formal training whatsoever?

Dr. Brady: That's right. When they are not sitting there doing my project, they could be doing anything. They could be looking at a photograph that has pricing information and writing down the pricing information. They might be looking at a picture and saying whether it has an element that is inappropriate for a family Website or for the public. We thought, "Let's see whether they can be trained to look at a fundus photo and say whether it's normal or abnormal."

Dr. Pyfer: How did you train them?

Dr. Brady: The key with all of this work is to make sure it's worth it for the person doing it. So it has to be very brief. We had a Web page where there were 6 training images that they could hover over, from normal fundus through a severe fundus, annotated with the different features of diabetic retinopathy. They could hover over them to learn about retinopathy. I'm assuming that they would probably do that more on the first or second picture, and then as they got more advanced, they wouldn't have to do that anymore.

Dr. Pyfer: Then they would grade a series of images and try to identify what stage of retinopathy each image represented?

Dr. Brady: That's right. We started out asking them whether they thought the image was normal or abnormal, and from the outset they did that fantastically. They identified about 90% of the first batch of pictures that we put up there correctly. When I asked them to grade them as mild, moderate, or severe, trying to translate it into plain language, they didn't do much better than chance. We then made some modifications to the training, and they did quite a bit better.

Dr. Pyfer: How did they compare to expert interpreters of fundus photography?

Dr. Brady: That's a great question. That trial hasn't been done yet. For this study, we compared their grading against 2 surgeons (here at Wills Eye) as the gold standard. I didn't pit them against other types of graders. There have been studies of retinopathy of prematurity using telemedicine in which they looked at medical students, who were at different stages of their training, to see how well they interpret retinopathy of prematurity photos. That would be interesting to do here. There is also the issue of the resolution of the interface. An expert might get a different score when looking at images on a Website.

Dr. Pyfer: Do you think that this may have application in such areas of medical imaging as radiology?

Dr. Brady: There is a role for it. There are major medicolegal and even ethical boundaries that would have to be worked out before it could be used clinically. Any diagnostic modality has to be approved by the US Food and Drug Administration. Any imaging modality needs to be tested carefully before it can be used clinically. The only other use that I saw when we did the literature search was thick blood smear interpretation for malaria. That was very interesting, but if I were to look at a chest x-ray now, I would probably do a little worse than I might have after a few years of medical school.

Dr. Pyfer: It certainly has a cost advantage.

Dr. Brady: We used 10 graders per image, at 10 cents apiece. Amazon gets a 10% commission, so the cost was $1.10 per image. There are papers in the literature saying that if you want to make a diabetic retinopathy project feasible in the developing world, you have to keep the cost at about a dollar or two. So that was our target and that is how we chose that number. We think it is a very interesting first step.

Bringing Glaucoma Care to Underserved Communities

Dr. Pyfer: Finally, welcome to Dr. Julia Haller, the Ophthalmologist-in-Chief here at Wills Eye Hospital. Julia, tell us what you found interesting at the conference this year.

Dr. Julia Haller: It was a great conference. There were a number of highlights. One was the media session where we talked about ethical issues and how medicine, specifically ophthalmology, can interface better with the media.

Another highlight would be the session with the resident and fellow papers. I know we have had many sound bites and clips in this session, where some of the speakers have highlighted the main points of their discussions. As a retina specialist, I thought the paper given by Mike Dollin was very interesting, in which he looked at a huge number of patients who were given intravitreal injections before and after instituting a "no talking" rule in the injection room, and showed a significant drop in the rate of endophthalmitis. The rate of endophthalmitis is low to begin with, but eliminating the chance of particulate spray and mouth flora in the room dropped it significantly. We were able to show that in part because such a huge number of intravitreal injections are done here.

Dr. Pyfer: That included keeping everything else the same -- standard infection precautions with povidone iodine prep and a sterile lid speculum -- but just not having the patient or the physician talk during the injection.

Dr. Haller: Exactly. The point was made that you can inexpensively increase the success of these injections and lower the risk. One thing that is always debated is mask vs no mask. This was low cost; no masks were used -- just a strict "no talking" policy.

I would also have to highlight a groundbreaking study that is ongoing with our glaucoma service through a grant from the Centers for Disease Control and Prevention. They are taking everything from the glaucoma clinic to community centers to underserved populations in the greater Philadelphia area.[9] I joke that they are taking the mountain to Mohammed because they are examining patients, doing visual field tests, and finding patients, for example, with narrow angles out in the field and then doing laser iridectomies right there. It is astonishing how much pathology they are finding. Patients who otherwise would not come in to see the physician are being found and treated prophylactically. It's very exciting, and the idea is that this could be scaled up as a new approach to reaching the underserved.

Dr. Pyfer: As part of that effort they presented bilateral simultaneous laser peripheral iridotomies that showed that they were just as safe as doing them separately.

Dr. Haller: That was a little bit controversial and occasioned some discussion. That was a very interesting point.

Dr. Pyfer: Is there anything about the program here at Wills Eye Hospital that supports this type of research and encourages our residents and fellows to get involved that leads to such quality work?

Dr. Haller: Overall there is an atmosphere of the "inquiring mind" and what we can do to give evidence-based information, which will push us into new clinical directions. You noticed that most of the papers were very clinically focused with clinical research. One of the key innovations has been establishment of a new program whereby seed projects are funded through our endowment. It's called the Innovations Grant Program, and many of the residents and fellows have to be involved in these projects for them to be funded. Many of them referenced the funding by the Innovations Grant Program, and we are seeing that come to fruition. It shows how a little bit of funding -- and the kindness of your philanthropists -- can make a big difference in resident education and in research.

Dr. Pyfer: I am very impressed by the amount of collaboration with other institutions that happens as well.

Dr. Haller: We are very lucky here in Philadelphia to have so many outstanding scientific institutions. Because of our structure as an eye hospital, that is "Switzerland," if you will, Wills Eye has the ability to do projects with so many of the top investigators of the world, and that is very exciting.

Dr. Pyfer: Thank you very much for joining us. This has been a collaborative effort between Wills Eye Hospital and Medscape Ophthalmology.