Where are we at with model-based economic evaluations of interventions for dementia? a systematic review and quality assessment

Objective:To identify, review, and critically appraise model-based economic evaluations of all types of interventions for people with dementia and their carers.Design:A systematic literature search was undertaken to identify model-based evaluations of dementia interventions. A critical appraisal of included studies was carried out using guidance on good practice methods for decision-analytic models in health technology assessment, with a focus on model structure, data, and model consistency.Setting:Interventions for people with dementia and their carers, across prevention, diagnostic, treatment, and disease management.Results:We identified 67 studies, with 43 evaluating pharmacological products, 19 covering prevention or diagnostic strategies, and 5 studies reporting non-pharmacological interventions. The majority of studies use Markov models with a simple structure to represent dementia symptoms and disease progression. Half of all studies reported taking a societal perspective, with the other half adopting a third-party payer perspective. Most studies follow good practices in modeling, particularly related to the decision problem description, perspective, model structure, and data inputs. Many studies perform poorly in areas related to the reporting of pre-modeling analyses, justifying data inputs, evaluating data quality, considering alternative modeling options, validating models, and assessing uncertainty.Conclusions:There is a growing literature on the model-based evaluations of interventions for dementia. The literature predominantly reports on pharmaceutical interventions for Alzheimer's disease, but there is a growing literature for dementia prevention and non-pharmacological interventions. Our findings demonstrate that decision-makers need to critically appraise and understand the model-based evaluations and their limitations to ensure they are used, interpreted, and applied appropriately.