Enterprise AI Analysis
Diabetic Retinopathy Screening in Primary Care Real Practice: Study Procedures and Baseline Characteristics from the RETINAvalid Project
The RETINAvalid Project aims to evaluate diabetic retinopathy (DR) screening in primary care, focusing on concordance between primary care physicians (PCPs), ophthalmologists, and an AI system. This brief report details the study protocol and baseline characteristics of 301 diabetes patients, from whom 1517 retinographies were collected. PCPs referred 34.5% of patients to ophthalmology due to opacification, suspected DR, or other retinal issues. Overall, 13.62% of participants were suspected of having DR, with 9.63% receiving a definitive diagnosis. Future phases will analyze concordance to assess the reliability of DR screening across different evaluators and support the integration of effective DR screening strategies into real-world clinical practice.
Key Metrics from the Research
Actionable insights demonstrating the potential impact of advanced AI in healthcare.
301
Patients Included
1517
Retinographies Collected
34.5%
PCP Referral Rate
9.63%
Patients with Definitive DR Diagnosis
Deep Analysis & Enterprise Applications
Select a topic to dive deeper, then explore the specific findings from the research, rebuilt as interactive, enterprise-focused modules.
Context of the Study
Diabetic retinopathy (DR) is a major complication of diabetes and a leading cause of blindness. Early detection through screening is crucial. Integrating DR screening into primary care, using retinography, and leveraging AI holds significant potential to improve diagnosis and patient management while reducing healthcare workload.
Study Methodology
The RETINAvalid Study is a retrospective study conducted in Northern Metropolitan Area of Barcelona, Spain. It evaluates inter-observer concordance in DR screening using retinography among primary care physicians, ophthalmology specialists, and an AI software. Participants were adults (≥18 years old) with T1D or T2D, who underwent screening retinography within the previous six years. Exclusion criteria included conditions affecting decision-making capacity or retinography quality (e.g., severe cataracts, high myopia). Data collected included sociodemographic, clinical (diabetes type, duration, HbA1c, treatments, comorbidities), and retinography reports (PCP assessment, referral reasons, ophthalmology diagnosis). Retinographies were anonymized. Ophthalmology specialists will independently evaluate images for quality and DR/DME classification, with a third specialist resolving disagreements to establish a gold standard. An AI system (UMI DR v1.0.0) will also assess images for moderate NPDR or more severe stages. The study requires 625 retinographies for planned concordance analyses.
Key Study Results
From 329 individuals, 301 (91.5%) subjects were included, yielding 1517 retinographies (average 4.67 per patient) over six years. The study population was predominantly older adults (mean age 70.3 years) with type 2 diabetes (99.3%), and a mean disease duration of 12.8 years. Mean HbA1c was 7.4%. Hypertension was highly prevalent (74.7%), and 12.3% had coronary disease. PCPs referred 34.5% of patients to ophthalmology. Reasons for referral included opacification (14.7%), suspicion of DR (5.51% of those not diagnosed with DR, but 58.6% of those with a DR diagnosis), and other retinal diseases. Overall, DR was suspected in 13.62% of patients, with 9.63% receiving a definitive diagnosis. 82.7% of patients diagnosed with DR were referred to ophthalmology by PCPs.
Discussion and Implications
The study's preliminary results align with previous Spanish studies regarding DR prevalence and referral rates. The mean age and high prevalence of T2D reflect demographic trends. The observed HbA1c (7.4%) is acceptable for older patients with individualized glycemic goals. High prevalence of hypertension (74.7%) and coronary disease (12.3%) aligns with known diabetes comorbidities, emphasizing integrated care. The average abdominal perimeter (103.8 cm) suggests elevated cardiovascular risk. Diet (71.4% receiving recommendations) and non-smoking (55.8%) indicate lifestyle focus. The 34.5% referral rate is consistent with other studies, highlighting retinography's value beyond DR detection for other retinal pathologies. The effective identification and referral of 82.7% of DR-diagnosed patients by PCPs is positive. The study emphasizes the future analysis of concordance between PCPs, ophthalmologists, and AI to improve diagnostic consistency, streamline workflows, and enhance patient outcomes, leveraging telemedicine and AI's potential in DR screening.
34.5%
of patients were referred to ophthalmology by primary care physicians.
Enterprise Process Flow
Primary Care Retinography Screening
→
PCP Assessment & Referral Decision
→
Ophthalmology Specialist Reassessment (Gold Standard)
→
AI Software Assessment
→
Concordance Analysis & Strategy Enhancement
| Reason | No DR Diagnosis (n=272) | DR Diagnosis (n=29) |
|---|---|---|
| No referral | 192 (70.6%) | 5 (17.2%) |
| Suspicion of Diabetic Retinopathy | 15 (5.51%) | 17 (58.6%) |
| Opacification | 40 (14.7%) | 3 (1.10%) |
| Other retinal diseases | 14 (5.15%) | 0 (0.00%) |
| More than one reason | 9 (3.31%) | 4 (1.47%) |
9.63%
of participants had a definitive diagnosis of diabetic retinopathy.
Leveraging AI for Enhanced DR Screening
The study aims to evaluate an AI-based system (UMI DR v1.0.0) for automated retinography reading. This system has shown high diagnostic performance in previous validation studies, achieving 95.69% sensitivity and 94.44% specificity for detecting more than mild DR. Its integration into screening workflows could standardize diagnostic processes, enhance efficiency, and reduce workload for both primary care physicians and ophthalmologists. This will improve patient care by ensuring timely detection and appropriate referrals.
Advanced ROI Calculator
Estimate your potential efficiency gains and cost savings by integrating AI into your enterprise operations.
Annual Cost Savings
$0
Annual Hours Reclaimed
0
Your AI Implementation Roadmap
A typical journey to integrating enterprise AI, tailored for maximum impact.
Phase 1: Discovery & Strategy
In-depth analysis of current workflows, identification of AI opportunities, and development of a tailored implementation strategy. Define clear objectives and success metrics.
Phase 2: Pilot Program Development
Design and deploy a proof-of-concept AI solution in a controlled environment. Gather initial data, refine algorithms, and demonstrate tangible ROI on a small scale.
Phase 3: Scaled Integration & Training
Expand the AI solution across relevant departments, integrate with existing systems, and provide comprehensive training for your team to ensure seamless adoption.
Phase 4: Optimization & Future-Proofing
Continuous monitoring, performance optimization, and iterative improvements. Explore new AI capabilities and strategic enhancements to maintain competitive advantage.
Ready to Transform Your Enterprise with AI?
Schedule a personalized strategy session with our AI experts to explore how these insights can be applied to your specific business challenges and opportunities. Unlock unparalleled efficiency and innovation.
Ready to Get Started?
Book Your Free Consultation.
Let's Discuss Your AI Strategy!
Lets Discuss Your Needs
Select a Date
Sun
Mon
Tue
Wed
Thu
Fri
Sat