Crystallise Ltd., East Tilbury, UK
METHODS: We searched the heoro.com database (www.heoro.com) for cost-benefit, cost-effectiveness or cost-utility modelling studies in dementia that were published between 1960 and 30th May 2017. We analysed the abstracts identified by the search to determine the different types of interventions modelled across the range of geographical locations, by date, perspective and type of dementia. We presented the findings as an evidence map.RESULTS: We found a total of 131 abstracts. Of these, 76 were cost-utility models, 10 cost-effectiveness, 5 cost-benefit and 42 did not specify the modelling approach; 62 used a societal and 27 a healthcare perspective and this was unclear in 49. Most reported models for Alzheimer’s disease (66 abstracts) or any type of dementia (59). Drug therapies were the most common interventions, in particular donepezil (26), memantine (25), rivastigmine (13), galantamine (12) or any anticholinesterase (15), but screening or diagnostic approaches (16) and caregiver support interventions (14) were also frequently modelled. The jurisdiction of 42 models was the United Kingdom, 16 the United States, 14 the Netherlands, 6 each Canada and Sweden, and 26 were reviews of the international modelling literature. Of the 131 abstracts, 70 were published between 2010 and 2017 and 13 were published before 2000.
CONCLUSIONS: The focus on cost-utility models from a societal perspective reflects the high impact of dementia on quality of life of patients and caregivers. This is also shown by the substantial number of models on caregiver support and patient care interventions, as well as drug therapies. The preponderance of studies from the UK may reflect the burden on state-funded social care organisations in this country.