A Type 2 Diabetes Policy Model That Predicts Remaining Life Expectancy, Quality-Adjusted Life Expectancy and...

To illustrate a T2DM policy model incorporating socio-economic status.

Using the Scottish Diabetes Research Network (SDRN) national diabetes cohort, we identified individuals newly diagnosed with T2DM between 1 January 2004 and 31 December 2020 and followed from diagnosis until death or end of follow-up. Covariates included sex, Scottish Index of Multiple Deprivation (SIMD) quintiles, ethnicity, HbA1c, systolic blood pressure (SBP), eGFR, total cholesterol, high-density lipoprotein, BMI, smoking status and cardiovascular disease history. Remaining life expectancy was modelled using Gompertz regression, stratified by sex and SIMD. Probabilities of complications were estimated using repeated measure logistic regression, and healthcare costs were modelled using repeated measure linear regression. A parameterised Kaplan-Meier sample-average estimator was used to obtain life expectancy, quality-adjusted life years (QALYs) and costs.

We identified 269 759 people newly diagnosed with T2DM. The mean age at diagnosis was 63 years, 56% were male and 24% lived in the most deprived areas. Total follow-up was 2 090 108 patient-years (median 7 years per patient). Approximately one in four patients died during follow-up. The most common T2DM-related complications were coronary heart disease and heart failure. For a man diagnosed at age 50 in the most deprived quintile, with all other covariates at average values, the model predicted a remaining life expectancy of 26.7 years, 21.1 QALYs and healthcare costs of £273 k (undiscounted). For the same profile, the incremental cost-effectiveness ratio for a hypothetical intervention reducing HbA1c by 1% was £7832 per QALY.

We developed a T2DM policy model incorporating socio-economic deprivation, providing estimates of life expectancy, QALYs and healthcare costs to inform health technology assessment.