Skip to main content
  • Telemedicine May Improve Quality, Frequency of ROP Screening

    Telemedicine may improve surveillance for retinopathy of prematurity (ROP) by sidestepping local shortages of ophthalmologists who are trained for ROP screening, according to a presentation by Antonio Capone Jr., MD, at the Retina Subspecialty Meeting on Friday.

    Pediatric ophthalmologists are in short supply. Nearly half of level III NICU directors say there are too few screening ophthalmologists in their area to follow patients at risk for ROP. Only 1 in 5 at-risk infants undergo routine retinal imaging, according to a large 2017 survey.

    “There’s a significant manpower shortfall,” said Dr. Capone, a clinical professor of biomedical sciences at Oakland University, in Rochester, Michigan.

    Telemedicine technology could help bridge this gap—while enhancing the quality of ROP surveillance and reducing costs—thanks to advances in smartphone fundus photography, machine learning, and image analysis.

    Telemedicine workflows may improve quality and frequency of ROP screens. Recent studies suggest that teleophthalmology is 40% more cost-effective than ophthalmoscopy, and 3 times as fast. While traditional ROP screening is done at the infant’s bedside roughly every 2 weeks, telemedicine requires fewer physicians and can be performed weekly on an inpatient or outpatient basis.

    The digital format allows physicians to perform longitudinal image comparisons and high-throughput analyses and disseminate data easily. Smart software can perform semi-automated zone 1 determination, quantitate signs of disease, and minimize the likelihood of a missed diagnosis.

    What the future holds. Dr. Capone described up-and-coming technologies that rely on optical coherence tomography (OCT), including a 3-D system for detecting ROP from researchers at Duke University.

    As the field advances, teleophthalmology technologies will become even more powerful, Dr. Capone said. Machine learning will further improve ROP screening by enhancing feature segmentation and detecting the earliest signs of disease long before they might have been detected by a human examiner.

    “We won’t be replaced anytime soon, but machine learning will likely help us do what we do—just better,” said Dr. Capone Anni Griswold

    Financial disclosures: Alcon Laboratories, Inc.: S; Aura Biosciences: S; Broadspot: O; FocusROP: O; Genentech: S; Iconic Therapeutics: S; interVIEW: O,P; Novartis Pharmaceuticals Corporation: C; Ohr Pharmaceuticals: C,S; Otsuka Pharmaceutical Co.: S; Phoenix Technology Group: O; Retinal Solutions: O,P; Spark Therapeutics: C; ThromboGenics, Inc.: S

    Disclosure key. C = Consultant/Advisor; E = Employee; L = Speakers bureau; O = Equity owner; P = Patents/Royalty; S = Grant support.