Multi-use disease models:A blueprint for application in support of health care insurance coverage policy and a case study in Diabetes Mellitus

AbstractThe National Health Care Institute (hereinafter referred to as ZIN) advises the Minister of Health, Welfare and Sport (VWS) on a variety of topics, including reimbursements for medicines and other treatments in the mandatory health insurance package. For this purpose, among other things, ZIN uses dossiers from pharmaceutical companies, in which they estimate the health benefits and costs based on decision models. These decision models are used to examine whether medicines and other treatments improve the health of individuals in the long term. For example, they calculate how better blood glucose levels in people with diabetes translate to less complications, such as cardiovascular diseases and amputations. The models incorporate the costs of the medicine, but also the savings made because the medicine reduces the likelihood of complications and the resulting high treatment costs. ZIN is currently assessing decision models dossier by dossier, with a different decision model being used for almost every medicine or treatment. Therefore, it is difficult to compare the effects of a number of medicines for the same disease. ZIN also spends a lot of time testing the quality of each decision model. It would be better to have a single model for each disease, or multi-use models, as such models meant for repeated use might be called. This would enable those involved to make better and more consistent decisions. RIVM and the Universities of Twente, Maastricht, Groningen and Utrecht have carried out a study into how ZIN could work with multi-use models. This report will assist ZIN in making the decision of whether to switch to multi-use models and, if so, how. By way of case study, RIVM has also created a multi-use model for diabetes. RIVM and its partners have developed five business cases for working with multi-use models and described their advantages and disadvantages. The roles and responsibilities of the parties involved, i.e. ZIN, other research institutes, academic groups, and consultancy bureaus, vary in these options. Issues discussed include ownership of the model, who is responsible for the maintenance and storage of results and who is accountable if mistakes are made. Other issues discussed concern the model methods, among others what the model must be able to do and how flexible it has to be for adaptation