This article in Vision Monday last week caught my eye: “The ‘Comptometrist’—Is There a Better Way to Determine Myopic Power?” The opening paragraph presents a very noble concept:
“When 14-year old Harini Venkatesh’s younger sister visited the eye doctor for her first checkup and came home complaining that she hadn’t been able to tell what was blurry and what wasn’t when the doctor had asked, she thought to herself, there has to be a better way to do this. ‘That was pretty scary, and added to the fact that I knew multiple people who did not communicate verbally (i.e. children with special needs, people who face a language barrier, etc.) it was jarring to realize that their experience at an optometrist’s office would be very different than mine.'”
Were that the focus of this budding scientist’s work, the idea of creating more objective ways of determining refractive status would be welcome. But in the video that accompanies the article, Harini does not limit herself to that goal. In the two minute video above (setting a new speed record for expressive language), she makes several eye-opening statements:
- “The patient response-based system of eye exams allows room for a large error. In fact there are approximately 2.86 million people every year who receive grossly incorrect prescriptions.” Wow, that’s interesting. I’d like to see the reference for that “fact”, and to know what constitutes a “grossly incorrect prescription”.
- “By investigating the relationship between the extent of the elongation of the optic disc and the amount of myopia in an eye, I’ve created a prototype that when given an image of the traditional eye scan taken at an appointment, will accurately return the amount of myopia in a patient’s eye in seconds.” As you listen further, you’ll hear that the error range of this model is 0.5D, which Harini says mirrors the acceptable error of “manually created prescriptions”. It appears that she trained her model through AI based on the Optos image.
- “My prototype also eliminates the need for patient responses in an eye exam, which closes the massive window for error in this process.” While there is good agreement that myopic stretching is elliptical, this statement qualifies as hyperbolic.
There is no doubt that Harini is a bright young individual with a promising future in science, but the concepts presented reflect a rather myopic view of myopia, based entirely on objective parameters better suited to robots than humans. Another glaring omission is any mention of the fact that one has two eyes, and a binocularly balanced Rx is worth the time and effort taken to derive it. As clinicians who have a high rate of Rx acceptance learn, when the patient is able to communicate with the doctor sufficiently, there is great value in tempering the final Rx based on subjective responses at distance and near. Even when the patient in pre-verbal, or has special needs that make communication more challenging, there is utility in observing the patient’s response to a tentative Rx.
In contrast to this myopic view is an article in Review of Myopia Management a few months ago on “The Importance of Binocular Vision in Managing Myopia Progression” by Dr. Eric Chow, and brought to my attention by Dr. Carole Hong. It is a comprehensive approach that takes into account all the considerations and concerns above. Dr. Chow is co-owner of Miami Vision Therapy where his team offers vision therapy and myopia management services. He graduated from the SUNY College of Optometry in 2016 and completed a residency in pediatrics and binocular vision from NOVA Southeastern College of Optometry in 2017. He is an FAAO who is currently working toward completing his fellowship with COVD.
““It works by first, capturing an image of the back of the eye using a low-cost but efficient fundus camera based on Raspberry Pi (a low cost, credit-card sized computer that plugs into a computer monitor or TV, and uses a standard keyboard and mouse). Then, an algorithm analyzes the image and locates a region of the retina, after which dimensions of this region are regressed against prescription power. Then, new images can be analyzed and extrapolated onto this model to calculate power.
“This process minimizes time spent on a checkup, reduces effort on the doctor’s part for what is a routine visit, and brings down the chance of error when calculating power,” Harini said. ”
No, it doesn’t ‘work’ that way. This shows the futility of relying on inexperience for wisdom. Ok, enough time wasted on this one… next?
Thanks for sharing another futile attempt to replace the art of prescribing with the science of prescribing. I also assume that the author is unaware of the meaning of a subjective! On the other hand, I do agree that there are many instances of inappropriate prescribing. From overminusing in myopia to overcorrecting in hyperopia, patients have been challenged to adapt to the wrong prescription. As we know, these patients either adjust, stop wearing their glasses or seek another opinion.
And then, we have ASTIGMATISM, which tends to be ignored by many scientists because it doesn’t fit easily into any research protocol. I won’t even discuss the prescribing issues with this common refractive condition! Thanks again for sharing your knowledge and insights.
You’re welcome, Richard.