ARTICLE IN BRIEF
New evidence from a three-phase study shows that photography of the ocular fundus, without the need for pupil dilation, is an easier and more effective alternative that can be incorporated into routine emergency-department use.
BOSTON—Nancy J. Newman, MD, FAAN, began her Houston Merritt lecture here at the AAN Annual Meeting in April with a case that went wrong. A 28-year-old woman who was obese came into the emergency department (ED) with a severe headache and nausea and vomiting. Her examination was considered normal, and she was told she had migraine. Two weeks later, she had decreased vision in both eyes and went back to the ED, where a computer tomography scan was normal. She had an outpatient consultation with a neurologist within a week. At three weeks, she still had vision problems, her headaches became worse, and the neurologist referred her to an ophthalmologist.
Her ocular fundus was finally examined: She had severe bilateral papilledema with no light perception vision in her right eye, and light perception only in the left eye.
“It was a catastrophe,” said Dr. Newman, director of neuro-ophthalmology at the Emory University School of Medicine, who was featured in May in Neurology Today. The woman did not have her ocular fundus examined on her first ED visit, despite indications to do so, nor on subsequent visits as her symptoms worsened.
The examination can have huge stakes: Papilledema or optic disc pallor can be a sign of a tumor, she said, and Grade 3 or 4 hypertensive retinopathy can be a sign of end organ damage, for instance.
The sad fact, Dr. Newman said, is that ophthalmoscopes can be difficult to use and their use is sometimes even discouraged in the ED. So the scopes sit in exam rooms and rarely get picked up, sometimes with agonizing consequences.
But photography of the ocular fundus, without the need for pupil dilation, is an easier and more effective alternative that needs to be incorporated into routine emergency department use, Dr. Newman said. With persuasive results found in a three-phase study — Fundus Photography versus Ophthalmoscopy Trial Outcomes in the Emergency Department, or FOTOED — she predicted the use of this photography will be a fixture of future care.
“Technology has advanced to the point where these cameras can take pictures through a non-dilated pupil. And it's easy for non-ophthalmic-trained individuals to perform,” she said. “You do not need to dilate the pupil, and it's able to take quality photographs of the back of the eye.”
To make her point, she showed two pictures of the same patient's ocular fundus side-by-side: One was taken by a professional photographer, in a dark room with a pupil dilated, and the other by a nurse practitioner in a lighted room with no dilation after 10 minutes of training. The photographs were of similar quality.
The first phase of FOTOED showed that nurse practitioners in the ED could indeed be easily trained to take quality photographs with a non-mydriatic ocular fundus camera under routine conditions. All the adult patients presenting with symptoms requiring a funduscopic exam — headache, visual complaints, acute focal neurological deficits, or a high diastolic blood pressure — had photographs taken, and neuro-ophthalmologists read the photos within 24 hours. The ED physicians were not provided the photographs in this phase.
The median photography session took 1.9 minutes. The photographs led to the detection of relevant findings in 44 of 350 patients that should have changed their management in the emergency room. Findings included disc edema, hypertensive changes and other abnormalities.
Just 14 percent of the 350 patients had documentation that the ED physician had tried to use the ophthalmoscope to examine the ocular fundus, and none of the relevant findings from the photographs was detected by the ED physicians.
In the next phase, the photographs were provided to the ED doctors — a camera icon appeared in the electronic medical record — and 68 percent of patients had their photographs viewed by the ED physicians. Nearly half of the relevant findings were assessed correctly, after comparison with examination by the neuro-ophthalmologists. Importantly, the ED physicians identified 86 percent of the normal fundi as normal on their viewing.
Researchers also found that when neuro-ophthalmologists looked at the photographs on their smart phones — probably “not as good as the iPhones you have now,” Dr. Newman said — the quality was deemed equal in quality to those seen on desktop computers.
A looming question, she said, is: Should medical students continue to be taught ophthalmoscopy at all?
Dr. Newman and other colleagues considered this, too. First-year medical students were trained with an ophthalmoscope on human volunteers or simulators, or viewed fundus photographs. The medical students were significantly better at identifying abnormalities when looking at the photographs.
“And even when they're energetic, enthusiastic, first-year medical students, and they've just gone through this great training about what you can see in the back of the eye, already only 49 percent of them said they would attempt direct ophthalmoscopy in clinical encounters that they had over the next year,” Dr. Newman said.
The performance of the same students was assessed again a year later. Although they performed worse with both the ophthalmoscopes and the photographs than they had before, they were still more accurate with the photographs. The median frequency of reported ocular fundus examination with ophthalmoscopes over the prior year was just 10 percent. The students reported that they weren't comfortable doing so, that they were discouraged from doing so by their preceptors, and that they didn't have enough time.
Phase 3 of FOTOED involved an attempt to train emergency physicians to perform better on ocular fundus photography interpretation, with a web-based program. In the end, 12.6 percent of patients had relevant findings that would change management, and ED physicians reviewed the photographs more often than they did when they had not been trained. But the training did not result in further improvements in assessing the photographs for abnormalities.
“The intervention was probably not well-designed and we need to make it better,” Dr. Newman said.
Still, ocular fundus photography in the emergency department clearly could prevent medical errors from occurring, she said.
“I think there's no getting around the conclusion that looking at the ocular fundus in whatever way you can, is more important than the method used to get there,” she said. “Non-mydriatic cameras need to replace direct ophthalmoscopes and we need to implement interventions that work for recognition and management.... It's only a matter of time before these cameras are everywhere.”
The cameras now cost between $15,000 to $25,000, but the cost is quickly coming down. Besides, she said, “What is the cost of a diagnostic error? What is the human cost and what is the medical-legal cost of even missing one of these cases?”
Steven Galetta, MD, FAAN, chair of neurology and a neuro-ophthalmologist at New York University Langone Medical Center, said the research at Emory has shown fundus photography to have “tremendous value” because it picks up findings that are often otherwise missed in the ED.
“The photographs become a permanent part of the record, can be transmitted to an expert for an official reading and allow the examiner time to stare at the findings without causing the patient any discomfort. It may be the only way that a fundus exam is consistently done in some settings and I'm a strong advocate of the technique.
He said that optimally fundus photography and ophthalmoscopy would be used in a complementary way, since ophthalmoscopy is still more portable and can provide more information about lesion depth and can allow the examiner to see the pulsatile properties of the blood vessels.
Dr. Galetta said fundus photography has the limitations of cost and that there is some technical skill required to take good photos, and the quality of those photos can be limited by opacity and other factors.
But he added: “Even experts miss important findings on ophthalmoscopy, so fundus photographs are definitely an important additive method to examining the back of the eye. Trainees at all levels also prefer to look at fundus photographs over examining the back of the eye, but I encourage them to do both, because it gives you more practice if nothing else.”
He said a portable and inexpensive way to obtain fundus photographs is “likely on the way” and that phones will probably play a big role once technical glitches are overcome.
“I'm very optimistic about the future of non-mydriatic fundus photography and its ability to improve care,” he said. “The eye will always remain an important gateway to what is happening in the brain, but you have to see it to believe it.”