Article Text

PDF
Telemedical diagnosis of retinopathy of prematurity: accuracy of expert versus non-expert graders
  1. Steven L Williams1,
  2. Lu Wang1,
  3. Steven A Kane1,
  4. Thomas C Lee2,
  5. David J Weissgold3,
  6. Audina M Berrocal4,
  7. Daniel Rabinowitz5,
  8. Justin Starren6,7,
  9. John T Flynn1,
  10. Michael F Chiang1,7
  1. 1Department of Ophthalmology, Columbia University College of Physicians and Surgeons, New York, USA
  2. 2The Vision Center, Childrens Hospital Los Angeles, Los Angeles, California, USA
  3. 3Retina Center of Vermont and University of Vermont, Burlington, Vermont, USA
  4. 4Department of Ophthalmology, Bascom Palmer Eye Institute, Miami, Florida, USA
  5. 5Department of Statistics, Columbia University, New York, New York, USA
  6. 6Biomedical Informatics Research Center, Marshfield Clinic, Marshfield, Wisconsin, USA
  7. 7Department of Biomedical Informatics, Columbia University College of Physicians and Surgeons, New York, USA
  1. Correspondence to Dr Michael F Chiang, Columbia University College of Physicians and Surgeons, 635 West 165th Street, Box 92, New York, NY 10032, USA; chiang{at}dbmi.columbia.edu

Abstract

Background/aims To assess accuracy of telemedical retinopathy of prematurity (ROP) diagnosis by trained non-expert graders compared with expert graders.

Methods Eye examinations (n=248) from 67 consecutive infants were captured using wide-angle retinal photography (RetCam-II, Clarity Medical Systems, Pleasanton, California, USA). Non-expert graders attended two 1-h training sessions on image-based ROP diagnosis. Using a web-based telemedicine system, 14 non-expert and three expert graders provided a diagnosis for each eye: no ROP, mild ROP, type 2 pre-threshold ROP or treatment-requiring ROP. All diagnoses were compared with a reference standard of dilated indirect ophthalmoscopy by an experienced paediatric ophthalmologist.

Results For detection of type 2 or worse ROP, the mean (range) sensitivities and specificities were 0.95 (0.94–0.97) and 0.93 (0.91–0.96) for experts, 0.87 (0.71–0.97) and 0.73 (0.39–0.95) for resident non-experts, and 0.73 (0.41–0.88) and 0.91 (0.84–0.96) for student non-experts, respectively. For detection of treatment-requiring ROP, the mean (range) sensitivities and specificities were 1.00 (1.00–1.00) and 0.93 (0.88–0.96) for experts, 0.88 (0.50–1.00) and 0.84 (0.71–0.98) for resident non-experts, and 0.82 (0.42–1.00) and 0.92 (0.83–0.97) for student non-experts, respectively.

Conclusions Mean sensitivity and specificity of trained non-experts were lower than that of experts, although several non-experts had high accuracy. Development of methods for training non-expert graders may help support telemedical ROP evaluation.

  • Child health (paediatrics), diagnostic tests/investigation
  • medical informatics
  • paediatric ophthalmology
  • retina
  • retinopathy of prematurity
  • telemedicine

Statistics from Altmetric.com

Footnotes

  • Funding This study was supported by the Starr Foundation, New York (SLW), the Doris Duke Charitable Foundation, New York (SLW), a Career Development Award from Research to Prevent Blindness, New York (MFC) and grant EY13972 from the National Institutes of Health, Bethesda, MD (MFC). MFC is an unpaid member of the Scientific Advisory Board for Clarity Medical Systems (Pleasanton, California, USA). The authors have no commercial, proprietary, or financial interest in any of the products or companies described in this article.

  • Competing interests MFC is an unpaid member of the Scientific Advisory Board for Clarity Medical Systems (Pleasanton, California, USA). The authors have no commercial, proprietary, or financial interest in any of the products or companies described in this article.

  • Ethics approval This study was approved by the Institutional Review Board at Columbia University Medical Center, and included waiver of consent for use of de-identified retinal images.

  • Patient consent Obtained.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • At a glance
    Harminder S Dua Arun D Singh