1987-P: Real-World Gains Evident for Eye Exams for Diabetes in Rural and Urban Clinics with Use of Autonomous AI



Introduction and Objective: Annual eye examinations for diabetes (EEDs) are essential for early detection of diabetic retinopathy, yet adherence remains suboptimal, particularly in rural communities where access to eye care is limited. Autonomous AI systems offer a scalable approach to expand patient access at the point of care. We evaluated the real-world impact of LumineticsCoreā„¢, an FDA-cleared autonomous AI diagnostic system for EEDs, deployed across rural and urban clinics within SSM Health (SSM).Methods: This retrospective observational study assessed adult patients with diabetes receiving care at SSM clinics before and after LumineticsCore implementation. Clinics were classified as rural or urban using Rural-Urban Commuting Area (RUCA) codes based on clinic ZIP codes. The primary outcome was change in adherence to annual EEDs, defined as completion of a documented eye exam by LumineticsCore or an eyecare provider within the measurement year. Adherence rates before and after implementation were stratified by rural versus urban clinic designation.Results: A total of 52 clinics (rural: 13; urban: 39) were included, representing 43,344 patients for the calendar year pre-implementation, and 48,607 for the calendar year post-implementation. Following implementation, adherence to annual EEDs increased significantly from 58.3% (25,288/43,344) in the pre-period, to 65.1% (31,652/48,607) in the post-period (p<0.001). Comparable improvements were observed across settings, with adherence increasing in urban clinics from 60.4% (18,941/31,360) to 67.3% (23,231/34,508) and in rural clinics from 53.0% (6,347/11,984) to 59.7% (8,421/14,099). The magnitude of improvement was similar between rural and urban clinics, with no significant difference between settings (p=0.40).Conclusion: Adoption of an autonomous AI diagnostic system resulted in equivalent improvements in EED adherence across rural and urban clinical settings. These findings suggest autonomous AI can support scalable access to EEDs and help close longstanding rural care gaps.

Disclosure

J. Eckelkamp: Other – possible speaker support and cost of poster submission.; Current; Digital Diagnostics. S.D. Vazquez: None. D. Vollman: Advisory Panel; Current; Barti EHR. D. Weitzman: Employee; Current; Digital Diagnostics. S.K. Katakam: None.



Source link