A simple and accessible diabetic retinopathy risk prediction model: Establishment and validation in a hospital-based cohort of type 2 diabetes patients
Authors' conclusion
The proposed model shows promise for guiding early interventions and improving outcomes. Further external validation is needed to confirm its applicability across diverse populations.
Study context
- Population
- Patients with T2D admitted 2013–2019 (70,073 screened); mean age 59 yearsof 70,073 screened
- Participants
- 2,585 patients
- Exposure
- Derivation vs Validation
- Outcome
- Diabetic retinopathy (DR) prediction (retinopathy-free survival) by nomogram: discrimination and calibration
- Follow-up
- 1, 3 and 5 years
Methodological detailshospital-based, prediction model development and validation, random derivation (2/3) and validation (1/3) sets, Cox proportional hazards regression, nomogram
- Hospital-based
- Prediction model development and validation
- Random derivation (2/3) and validation (1/3) sets
- Cox proportional hazards regression
- Nomogram
Main finding
As reported in the abstract
Discrimination (AUC) of nomogram for 1-, 3-, 5-year retinopathy-free survival
| Group | AUC 1-year | AUC 3-year | AUC 5-year |
|---|---|---|---|
| Derivation | 0.941 | 0.886 | 0.594 |
| Validation | 0.747 | 0.736 | 0.670 |
AUC 1-year
Derivation0.941
Validation0.747
AUC 3-year
Derivation0.886
Validation0.736
AUC 5-year
Derivation0.594
Validation0.670
Observational data: associations, not effects.
Further findings
Incident retinopathy · median follow-up 34 months
- Developed retinopathy
- 220 (8.5 %)
Original publication
A simple and accessible diabetic retinopathy risk prediction model: Establishment and validation in a hospital-based cohort of type 2 diabetes patients. Diabetes Res Clin Pract. Published 2 May 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →