Detecting ROP: A Better Way?

Graham E. Quinn, MD, MSCE


August 04, 2014

Editorial Collaboration

Medscape &

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

Good morning. I'm Graham Quinn. I'm a pediatric ophthalmologist here at Children's Hospital of Philadelphia. I'd like to share information on a study we just published in JAMAOphthalmology about telemedicine and retinopathy of prematurity (ROP).[1]

The study was funded by the National Eye Institute and National Institutes of Health. There were 13 clinical centers that participated in this clinical study.

Premature babies are at risk for ROP, an eye disease that can cause blindness. However, access to necessary screening in the neonatal intensive care unit (NICU) is somewhat limited in some areas. There are about 40,000-60,000 babies born in the United States at less than 3.3 lb and less than 30 weeks of gestation who are at risk for ROP and need to have eye examinations.

So, how do we serve that population? Over the past 30 years, the National Eye Institute has put almost $100 million into studies of ROP. This is a national commitment to help prevent blindness from this disease.

In our study, there were about 1250 kids with birth weights of less than 3.3 lb and at less than 31 weeks of gestation, who were enrolled in a clinical study in which the eye exams were conducted in the routine clinical manner by ophthalmologists who were experienced with ROP. At the same examination, during the same period, an imager took retinal images from that baby's eyes with a retinal camera. These imagers were nonphysicians who were trained specifically to take on this task.

Those images were then uploaded to a central server, and the images were downloaded to a reading center here at the University of Pennsylvania, where trained nonphysician readers graded those images for certain characteristics that are of concern in ROP -- not necessarily those eyes that need treatment, but those eyes that need to be evaluated by an ophthalmologist to consider what treatment or follow-up is needed.

In our study, we found that the trained readers were very good -- very reliable in detecting serious ROP more than 90% of the time. Among the 160 or so eyes of babies that needed treatment for ROP, 98% of them were detected with this telemedicine use.

What do you do with that? I believe that what this validity study shows is that this technique could be used in the underserved remote areas of the United States to provide excellent ROP screening in those situations. It may also be very generalizable to other countries where they're saving babies who didn't used to survive and are at higher risk than our babies are for ROP.

One of the things I really did not expect from this study was that ROP awareness became higher in the nursery because the nurses and neonatologists could see the disease and the family could understand why we're concerned about these babies. I think this will, in the long run, raise the overall level of care for these children.

Another plus of using ROP telemedicine screening would be that babies wouldn't necessarily have to be transferred to an ROP expert from a remote nursery for evaluation. You could simply send images of that eye to decide whether that baby needs to be transferred. That will decrease hardship for families, and that's an excellent outcome of this.

Thank you very much for watching.