Diabetes Risk Prediction Models: Which One Is Most Effective?
Three updated QDiabetes-2018 prediction algorithms effectively predict the risk for type 2 diabetes in patients with a higher risk for the condition, according to a recent study.
To validate these models, the researchers routinely collected data via the QResearch database of 11.5 million individuals without diabetes at baseline from 1457 general practices in England. Patient age ranged from 25 to 84 years. Patients were assigned to either the derivation cohort (n = 8.87 million) or the validation cohort (n = 2.63 million).
Diabetes Increases Fracture Risk by 22%
Diabetes Linked to 5-Fold Increased Mortality Risk in Young Adults
In the derivation cohort, separate risk equations were derived in men and women for evaluation at 10 years. Various risk factors including body mass index (BMI), family history of diabetes in a first-degree relative, and treated hypertension were considered. Fasting blood glucose and glycated hemoglobin A1c (HbA1c) were included in additional models. In the validation cohort, measures of calibration and discrimination were determined for men and women separately and for individual subgroups.
During follow-up, 178,314 incident cases of type 2 diabetes per 42.72 million person-years of observation had occurred in the derivation cohort, compared with 62,326 cases per 14.32 million person-years of observation in the validation cohort. Models A, B, and C included traditional risk factors for diabetes, and Models A and B also included new risk factors: statin use and learning disability, and gestational diabetes and polycystic ovary syndrome in women.
All 3 models had good calibration, as well as high levels of explained variation and discrimination. Model B identified and explained 63.3% of the variation in women and 58.4% in men in time to diagnosis of type 2 diabetes. Model B also had the highest sensitivity compared with current National Health Service recommended practice based on bands of either fasting blood glucose or HBA1c. However, only 16% of patients had complete available data for blood glucose measurements, smoking, and BMI.
“Three updated QDiabetes risk models to quantify the absolute risk of type 2 diabetes were developed and validated: model A does not require a blood test and can be used to identify patients for fasting blood glucose (model B) or HBA1c (model C) testing. Model B had the best performance for predicting 10-year risk of type 2 diabetes to identify those who need interventions and more intensive follow-up, improving on current approaches,” the researchers concluded.
Hippisley-Cox J, Coupland C. Development and validation of QDiabetes-2018 risk prediction algorithm to estimate future risk of type 2 diabetes: cohort study. BMJ. 2017;359:j5019. https://doi.org/10.1136/bmj.j5019.