Estimating risk factor progression equations for the UKPDS Outcomes Model 2 (UKPDS 90)


Computer simulation models, such as the UKPDS Outcomes model 2 (UKPDS-OM2), allow users to predict disease progression, health outcomes and costs in individuals with type 2 diabetes.
Data from 5,102 UKPDS participants from the 20-year trial and 4,031 survivors with 10 years post-trial follow-up were used to derive time path equations for 13 clinical risk factors: HbA1c, systolic blood pressure, LDL-cholesterol, HDL-cholesterol, BMI, micro- or macro-albuminuria, creatinine, heart rate, white blood cell count, haemoglobin, eGFR, atrial fibrillation and peripheral vascular disease. Using these new equations significantly improved the predictive accuracy of UKPDS-OM2 when compared with holding the risk factors constant over 25 years.
The ability to project risk factor time paths beyond observed data should improve modelling of long-term health outcomes for people with type 2 diabetes when using UKPDS-OM2 or other models.

[Read the manuscript here]