Skip to navigation Skip to main content Skip to footer

Approved Research

External validation of individual-level risk prediction models for prediabetes and type 2 diabetes complications

Principal Investigator: Dr Jianchao Quan
Approved Research ID: 76351
Approval date: January 13th 2022

Lay summary

Type 2 diabetes mellitus is a condition where the body has high blood sugar. It has become a major public health issue in recent decades as over time it leads to many complications including heart attack, stroke, kidney disease, vascular disease, amputations and blindness. Risk prediction models to simulate the risk of developing these complications can be used to assess prevention and treatment strategies to improve population health. However, existing risk prediction models have rarely been tested in populations outside of their training sample. We aim to externally validate risk prediction models for diabetes outcomes (including UKPDS OM2, RECODe and CHIME) by comparing model predictions to observed outcomes in other large datasets.

We aim to externally validate risk prediction models for diabetes (including UKPDS OM2, RECODe and CHIME) using individual-level participant data to compare model performance. External validation to assess the accuracy of these prediction models for diabetes and its complications helps to formulate population health policy.