DEMENTIA AS A COMORBIDITY OF DIABETES IN EVALUATING THE DIABETES PREVENTION PROGRAMME

Purpose: To update an existing Diabetes Prevention cost-effectiveness model to estimate the impact of including dementia in an evaluation of diabetes prevention policies in the United Kingdom. 

Method:

The School for Public Health Research (SPHR) Diabetes prevention model is an individual patient simulation model developed to evaluate the cost-effectiveness of a broad range of public health policies for Diabetes prevention. The SPHR Diabetes model described dynamic metabolic trajectories for BMI, systolic blood pressure, HbA1c, Total Cholesterol and HDL cholesterol to describe individuals risk of type 2 diabetes, microvascular outcomes, cardiovascular disease, congestive heart failure, cancer, osteoarthritis, depression and mortality. The model takes a lifetime perspective, estimating the impact of interventions on healthcare costs, clinical outcomes, survival and quality-adjusted life years.

The Diabetes Prevention Programme describes a programme of screening for individuals at high risk of type 2 diabetes and a lifestyle education programme. The lifestyle intervention is effective in reducing BMI, HbA1c, systolic blood pressure, and total cholesterol in individuals at high risk of type 2 diabetes. The programme has been shown to be effective and cost-effective in high risk groups. 

The SPHR Diabetes cost-effectiveness model was adapted to include Dementia as an outcome and an evaluation of the Diabetes Prevention Programme was updated. A conceptual model for dementia was developed following a systematic review of existing dementia models to review the methods of modelling in previous studies in dementia and in consultation with experts. A risk model that included diabetes, diabetes related comorbidities, BMI, systolic blood pressure and cholesterol was used to estimate individualised incidence of dementia. Dementia diagnosis and disease progression was modelled based on existing dementia models and in consultation with experts. 

Result: The model found that the Diabetes Prevention Programme was cost-saving. Including Dementia as an outcome of the model generated additional cost savings and QALY gains from reducing the incidence of dementia as a consequence of the intervention. However, there is evidence of competing risks between the main outcomes of the model.

Conclusion: The inclusion of dementia provides additional information in the evaluation of diabetes policies, and is particularly useful in evaluating policies targeting older age groups at greater risk of dementia.