Computer-based Screening for Diabetic Retinopathy

The study is an Observational, Phase 0 designed to establish that the risk for diabetic retinopathy assigned by the RiskAnalyzer improves the reading accuracy and consistency of any reader and it decreases the inter-reader variability.

Objectives: Objectives one, two, and three are arranged chronologically and in an increasing level of complexity as a three tiered approach to support the primary and secondary endpoint of the trial. Objective one is to test fully each system components of the study limited to a single site. Objective two is to evaluate the efficacy of the RiskAnalyzer to assign the risk of Diabetic Retinopathy in comparison to the gold standard. Objective three is to demonstrate that the reader's accuracy in grading images is improved when risk levels assigned by the RiskAnalyzer are made available to the reader while performing the grading of the images which is the primary endpoint of the trial Methods and Research Design: A network of clinical study sites will be established to meet the required number of cases needed as calculated by statistical analysis. Male and Female Subjects between the ages of 18-65 who are either pre-diabetic or diabetics will be eligible for participation in this study. Subjects will be recruited, consented, photographed and their images graded by two trained readers and analyzed by the RiskAnalyzer . The risk levels that are obtained from the RiskAnalyzer will be compared to the current gold standard practice, manual grading of each case by a reader. Data collected during this clinical trial will be reported to the referring physician in the form of a retinal screening report completed and signed by a licensed Ophthalmic professional and delivered to their attending physician. Risk levels for diabetic retinopathy obtained by use of the RiskAnalyzer will not be given to the attending physician under any circumstances in order to preserve standard of care for the patient.

The sensitivity, specificity, ROC, and data flow process of the RiskAnalyzer, retinal image reading system will be analyzed and based on the current gold standard of a human reader. This study is a three-aims study, 24 month in length, prospective, case-only study of the performance of the RiskAnalyzer. The risk levels obtained from the RiskAnalyzer are not made visible, i.e. are not unmasked to either of the two readers. In year two, the risk levels obtained from the RiskAnalyzer for half of the studies are unmasked to the two readers while grading the image.

Access to all study data and processes follows a role-based design. The clinical staff will have access only to clinical data but not the technical data. The technical team will have access to the technical data only but not the clinical data. The study coordinator will have access to all data. The use of computers will adhere to the Guidance for Industry Computerized Systems Used in Clinical Investigations and applicable sections of 21 CFRs part 11.