The prevalence and risk factors associated with diabetic retinopathy in patient with type 2 diabetes mellitus in Primary Care Unit, Chiangrai Prachanukroh Hospital using an artificial intelligence device

Abstract

BACKGROUND: Diabetic retinopathy (DR) is a major complication of diabetes mellitus and can leads to visual impairment. Proper glycemic management with appropriate DR screenings can help prevent this condition.

OBJECTIVE: To study the prevalence and factors associated with DR in patients with type 2 diabetes at the Primary Care Unit, Chiangrai Prachanukroh Hospital using an artificial intelligence (AI) screening device.

METHODS: This was a retrospective cross-sectional study. Patients with type 2 diabetes who underwent DR screening using AI device at the Primary Care Unit, Chiangrai Prachanukroh Hospital from January 1st to December 31st, 2024. Data were analyzed using descriptive statistics and logistic regression analysis.

RESULTS: Data were analyzed from 971 patients. The prevalence of DR was 43.7% The distribution of DR severity was as follows: moderate non-proliferative DR (NPDR) 29.3%, mild NPDR 12.2%, severe NPDR 1.6% and proliferative DR (PDR) 0.7%. Factors significantly associated with DR included duration of diabetes ≥10 years (aOR 1.8, 95% CI: 1.36-2.33, p<0.001), chronic kidney disease stage 3 or higher (aOR 1.4, 95% CI: 1.05-1.91, p=0.021), glycated hemoglobin (HbA1c) ≥7% (aOR 1.5, 95% CI: 1.14-1.96, p=0.003), and LDL ≥100 mg/dl (aOR 0.7, 95% CI: 0.50-0.87, p=0.004).

CONCLUSIONS AND RECOMMENDATIONS: The prevalence of DR detected by AI screening devices was 43.7%, higher than that reported in previous studies, because AI screening has high sensitivity and specificity and can detect DR at an early stage. Therefore, patients with a duration of diabetes ≥10 years, chronic kidney disease stage ≥3, and HbA1c ≥7% should undergo screening to allow diagnosis of DR and appropriate treatment, thereby reducing the risk of future vision loss.