Comparing Paths: Teleretinal vs. AI for Effective Diabetic Retinopathy Screening

Diabetic retinopathy (DR) is a severe complication of diabetes that can lead to vision loss if not detected and treated promptly. Annual diabetic eye exams are vital for early DR detection, treatment, and for preventing potential vision loss. However, adherence to annual screening recommendations remains low, ranging from 15-50% of patients. The integration of teleretinal screenings and AI-powered diagnostic tools has revolutionized DR screening by bringing these examinations directly to the patient's point of care. This article explores the benefits introduced by teleretinal screening versus the benefits introduced by AI-based screenings.

Teleretinal screenings involve capturing retinal images at patients' point-of-care and transmitting them for remote evaluation by human readers. In contrast, AI-based screening utilizes advanced algorithms for instant image quality analysis and fully autonomous on-the-spot diagnosis. While both teleretinal and AI screenings contribute to advancing patient care, they differ significantly.

Teleretinal exams rely on human readers remotely interpreting captured images, which means that there are delays in feedback and diagnosis. Images are uploaded to an evaluation queue and will remain in the queue until an expert reader is available to analyze them. If an image captured is determined by the expert to be of insufficient quality, then that determination is made long after the patient has left the clinic. As a result, the patient needs to be called back for a retake, undoing the convenience of point-of-care screening. Furthermore, the time lag between the patient visit and the availability of the diagnostic results requires the primary care providers to merge the exam result into the patient record in the EMR and schedule a follow-up visit with the physician to discuss the result. This is an administrative burden for the primary care clinic. 

AI diagnosis stands out for its on-the-spot results. The technology offers instant image quality feedback to the retinal camera operator, allowing patients to be reimaged immediately if the images are not of diagnostic quality. The diagnostic result is received on the spot, which allows the patient to discuss the result with the physician on the same visit. In contrast, teleretinal screenings do not provide results on the spot, which means the results can be discussed only during a later visit. 

Reimbursement structures play a pivotal role in the adoption of these technologies. AI-based screenings, coded under 92229, offer a simple reimbursement process, directly benefiting primary care providers and simplifying billing. Conversely, teleretinal screenings, coded under 92227, have a technical and professional component (-TC and -26 modifiers) which are split between the clinic obtaining the images and the reading center providing the professional diagnosis. This creates administrative challenges, potentially impeding the adoption of teleretinal exams. 

In summary, the integration of teleretinal screenings and AI-powered diagnostics has advanced diabetic retinopathy screening, providing accessible and convenient solutions directly at the patient's point of care. While both methods contribute to improving adherence rates, AI-based screening introduces unbeatable benefits . The immediate results promote higher completion rates, prompt follow-up (and referrals when needed) and a reduction in the administrative burden on PCP clinics.