AUTHOR=Wang Minglan , Zhou Xiyuan , Liu Dan Ning , Chen Jieru , Zheng Zheng , Ling Saiguang 

TITLE=Development and validation of a predictive risk model based on retinal geometry for an early assessment of diabetic retinopathy

JOURNAL=Frontiers in Endocrinology

VOLUME=Volume 13 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/endocrinology/articles/10.3389/fendo.2022.1033611

DOI=10.3389/fendo.2022.1033611

ISSN=1664-2392

ABSTRACT=Aims: To develop and validate a risk nomogram prediction model based on retinal geometry of diabetic retinopathy (DR) in patients with type 2 diabetes mellitus (T2DM) and investigate its clinical application value.
Methods: In this retrospective study, we collected clinical data from 410 patients with T2DM from the Second Affiliated Hospital of Chongqing Medical University between October 2020 and March 2022. Firstly, the patients were divided into development cohort and validation cohort at a ratio of 7:3. Then, the modeling factors were selected by the least absolute shrinkage and selection operator (LASSO). After that, a nomogram prediction model was built with these identified risk factors. Another two models were constructed with only retinal vascular traits or only clinical traits, in order to confirm the performance advantage of this nomogram model. Finally, the model performances were assessed with the area under the receiver operatin characteristic curve (AUC), calibration plot, and decision curve analysis (DCA).
Results: Five predictive variables for DR among patients with T2DM were selected by LASSO regression from 33 variables, including fractal dimension, arterial tortuosity, venular caliber, duration of DM and insulin dosage (P<0.05). A predictive nomogram model based on these2 selected clinical and retinal vascular factors presented good discrimination with the AUC of 0.909 in training cohort and 0.876 in validation cohort, respectively. By comparing models, the retinal vascular parameters were proven to have predictive value, and combined with clinical characteristics could improve diagnostic sensitivity and specificity. The calibration curve displayed highly consistency between predicted and actual probability, both in training cohort and validation cohort. The decision curve analysis (DCA) demonstrated that this nomogram model would lead to net benefits in a wide range of threshold probability and could be adapted for clinical decision-making.
Conclusion: This study presented a predictive nomogram that may facilitate the risk stratification and early detection of DR among patients with T2DM.