Skip to content
DraconiticResearch
Search Sources API
201,683 studies31 sourcesUpdated 20 min ago
Peer-reviewedCC BYnpj Digital Medicine

Clinical setting-dependent diagnostic accuracy of artificial intelligence and store-and-forward diabetic retinopathy screening: a systematic review and meta-analysis

  • Kai-Yang Chen1
  • Hoi-Chun Chan2
  • Chi-Ming Chan3,4
4 affiliations
  1. 1Department of General Medicine, Chang Gung Memorial Hospital (Linkou branch), Taoyuan, Taiwan
  2. 2School of Pharmacy, China Medical University, Taichung, Taiwan
  3. 3Department of Ophthalmology, Cardinal Tien Hospital, New Taipei City, Taiwan. [email protected]
  4. 4School of Medicine, Fu Jen Catholic University, New Taipei City, Taiwan. [email protected]
Open at sourceDOICite

Abstract

Population-based diabetic retinopathy (DR) screening requires diagnostic strategies that optimize clinical utility by balancing missed disease against referral burden. We performed a Preferred Reporting Items for Systematic Reviews and Meta-Analyses of Diagnostic Test Accuracy studies (PRISMA-DTA)-guided systematic review and meta-analysis comparing autonomous artificial intelligence (AI) screening with store-and-forward (SAF) or conventional image-based teleophthalmology pathways, using manual, expert, or reading-center grading as the reference standard, across any DR, referable DR (RDR), vision-threatening DR (VTDR), and diabetic macular edema (DME). Twenty-eight diagnostic accuracy studies were included. AI showed higher pooled sensitivity than SAF for any DR (86.9% vs 80.9%), RDR (96.2% vs 88.6%), VTDR (96.2% vs 84.2%), and DME (97.2% vs 87.4%). AI also showed higher pooled specificity for any DR, RDR, and VTDR, whereas DME specificity was similar between pathways. Translating operating characteristics into decision consequences demonstrated that pathway preference depends on prevalence, decision thresholds, and misclassification weighting: at 15% prevalence, AI yielded higher net benefit (140.7 vs 120.8 net true-positive decisions per 1000 screened at pₜ = 0.10). These findings support pathway-specific deployment strategies rather than direct superiority claims.

Related

  1. Performance Assessment of an Artificial Intelligence System Used for Case Detection in a Large Community-based Diabetic Retinopathy Screening Program: A Retrospective Diagnostic Accuracy Validation Study.
    Maria Kristina Dg Eduardo, Junn R Pajarillo
    Acta medica PhilippinaPeer-reviewedCC BY-NC-NDFull textMedical imaging & radiologyEurope PMCPMC
  2. Artificial Intelligence-Based Burn Assessment Demonstrates High Diagnostic Accuracy: Systematic Review and Meta-Analysis
    Emily Uyen Thai, Mina Bahramand, Jordyn Carroll +5
    Plastic and Reconstructive Surgery, Global OpenPeer-reviewedCC BY-NC-NDFull textEpidemiology & public healthDOAJ
  3. From underdiagnosis to scalable screening: Applications of artificial intelligence for vertebral compression fracture detection – a systematic review and meta-analysis of diagnostic performance
    Hsuan-Yu Chen, Linjun Yang, Gwen Wilson +4
    Osteoporosis and sarcopeniaPeer-reviewedCC BY-NC-NDFull textGeneral medicineEurope PMC
  4. Artificial Intelligence Versus Human Interpretation in Diagnostic Radiology: A Systematic Review of Diagnostic Accuracy and Operational Efficiency.
    Santiago Diaz-Perdomo, María J Losada-Peña, Viviana Gutiérrez-Riobamba +3
    CureusPeer-reviewedCC BYFull textMedical imaging & radiologyEurope PMCPMC
  5. Artificial intelligence-supported diabetic retinopathy screening in Tanzania: a randomised controlled trial
    Charles R Cleland, Covadonga Bascaran, William U Makupa +16
    The Lancet Global HealthPeer-reviewedCC BY 4.0Medical imaging & radiologyCrossref

Cite

BibTeXRIS
BibTeX
@article{chen2026clinical,
  title = {{Clinical setting-dependent diagnostic accuracy of artificial intelligence and store-and-forward diabetic retinopathy screening: a systematic review and meta-analysis}},
  author = {Kai-Yang Chen and Hoi-Chun Chan and Chi-Ming Chan},
  journal = {npj Digital Medicine},
  year = {2026},
  month = may,
  doi = {10.1038/s41746-026-02627-0},
  url = {https://doi.org/10.1038/s41746-026-02627-0},
  note = {CC BY},
}
RIS
TY  - JOUR
TI  - Clinical setting-dependent diagnostic accuracy of artificial intelligence and store-and-forward diabetic retinopathy screening: a systematic review and meta-analysis
AU  - Kai-Yang Chen
AU  - Hoi-Chun Chan
AU  - Chi-Ming Chan
PY  - 2026
DA  - 2026/05/15
JO  - npj Digital Medicine
DO  - 10.1038/s41746-026-02627-0
UR  - https://doi.org/10.1038/s41746-026-02627-0
AB  - Population-based diabetic retinopathy (DR) screening requires diagnostic strategies that optimize clinical utility by balancing missed disease against referral burden. We performed a Preferred Reporting Items for Systematic Reviews and Meta-Analyses of Diagnostic Test Accuracy studies (PRISMA-DTA)-guided systematic review and meta-analysis comparing autonomous artificial intelligence (AI) screening with store-and-forward (SAF) or conventional image-based teleophthalmology pathways, using manual, expert, or reading-center grading as the reference standard, across any DR, referable DR (RDR), vision-threatening DR (VTDR), and diabetic macular edema (DME). Twenty-eight diagnostic accuracy studies were included. AI showed higher pooled sensitivity than SAF for any DR (86.9% vs 80.9%), RDR (96.2% vs 88.6%), VTDR (96.2% vs 84.2%), and DME (97.2% vs 87.4%). AI also showed higher pooled specificity for any DR, RDR, and VTDR, whereas DME specificity was similar between pathways. Translating operating characteristics into decision consequences demonstrated that pathway preference depends on prevalence, decision thresholds, and misclassification weighting: at 15% prevalence, AI yielded higher net benefit (140.7 vs 120.8 net true-positive decisions per 1000 screened at pₜ = 0.10). These findings support pathway-specific deployment strategies rather than direct superiority claims.
SN  - 2398-6352
ER  -