Tender published to find artificial intelligence solutions for the prevention of retinopathy

The Health/AI Programme takes a step forward in the search for artificial intelligence (AI) solutions to solve the needs of the Catalan Health System, with the publication of the tender to deploy AI solutions to support the diabetic retinopathy care process in primary care. The Health Quality and Assessment Agency of Catalonia (AQuAS) has published the technical specifications and all the documentation associated with the Public Procurement Services Platform.

Companies interested in participating can submit their offers until 11 November at 1:00 p.m. The tender has an estimated contract value of approximately 340,000 euros, and has been made possible thanks to Catalonia’s Recovery, Transformation and Resilience Plan (NextGeneration EU Fund).

Purpose of the tender

The tender comes about after analysing the Preliminary Market Consultation, which was done as part of the challenge, and which concluded that most AI-based solutions have the ability to detect and classify the degree of severity of diabetic retinopathy.

The purpose of the tender is the procurement of the services required to implement, deploy and maintain an AI-based technological solution that facilitates the screening of diabetic retinopathy in Catalonia, and that it is integrated into the primary care flow of the Catalan Health System. The AI-based solution must be able to analyse the quality of the image and to identify, detect and classify the lesion and the degree of the disease.

Specifically, as detailed in the technical specifications, the services covered by the contract will consist of the automatic evaluation of the quality of the captured retinography; the clinical decision support system for screening the automatic analysis of retinographies; the messaging and alert systems, and the update service for the entire solution.

The integrated solution will apply to all the primary care centres of the Catalan Health System.

Benefits of the solution

It is foreseen that the future incorporation of an AI solution for the prevention of diabetic retinopathy will reduce screening and diagnosis time, lower the number of unnecessary transfers to repeat imaging, improve access to specialised ophthalmology services, have greater control of the disease and reduce the number of treatments related to injuries caused.

In addition to the benefits for people with retinopathy, it is also expected that healthcare personnel can have greater precision in reading the images, reduce interconsultations between primary care and ophthalmology for the prioritisation of cases, and access to evidence-based information, among others.

Finally, for healthcare providers and the Health System, the solution should allow improving adherence to the screening process of diabetic retinopathy, achieve greater fluidity in the professional and patient circuit, and reduce expenses related to the burden of the disease and the variability between regions and professionals, among others.

About diabetic retinopathy

Diabetic retinopathy is a neurovascular complication of diabetic patients that is asymptomatic in the initial stages and can lead to reduced or complete vision loss if not detected in time. In Catalonia, it is estimated that it affects 8% of the population over the age of 15 and 20% over the age of 65, which translates into nearly 630,000 people affected. People with severe diabetic retinopathy have reduced quality of life and physical, emotional and social well-being, and consume more healthcare resources.

The treatments available for diabetic retinopathy can only stop the progression of the disease if started in its early stages, so early detection is key. Currently, professionals do not have a method to stratify patients based on the risk of developing the disease. The screening process is based on a retinography that the family doctor orders and evaluates for patients with Diabetes Mellitus Type 1 within five years of diagnosis, and those with Diabetes Mellitus Type 2, who are most likely to develop the disease, at the time of diagnosis. The degree of expertise of primary care professionals in this area is highly variable, and this often creates the need to issue an interconsultation with ophthalmology to evaluate uncertain cases.

All in all, as set out in the tender, having an AI solution to support the process of care and detection of diabetic retinopathy in primary care would make it possible to address the disease earlier and avoid the serious harm that leads to those affected.

[Link to tender publication]