Cost-effectiveness models estimate both the financial impact of a decision and any health outcomes resulting from this decision. Cost-effectiveness models are often referred to as cost-utility models and cost-consequence models. When a new treatment is evaluated with a cost-effectiveness model, it is normally compared to another treatment, known as a comparator. This is usually the current standard treatment, or combination of treatments, that patients with a certain condition would receive.
Health outcomes are generally measured in terms of quality-adjusted life years (QALYs), which incorporates both the quality and length of life that patients experience when on a particular treatment. Usually, the final result of a cost-effectiveness model is expressed as the incremental cost-effectiveness ratio (ICER), which is equal to the total cost difference caused by a new intervention divided by the number of additional QALYs that have been gained as a result. There are a variety of cost-effectiveness model types, each with different methodologies – common examples include Markov models or decision trees.
Cost of illness model
A cost of illness (COI) model estimates the overall economic burden of a disease, and can include both the direct (costs that can completely be attributed to the disease, such as hospital admission costs) and indirect (costs that cannot be completely attributed to a disease, such as productivity losses) financial costs associated with an illness. COI models can take a variety of perspectives, such as the financial cost to society, or just the NHS. Typically, COI models do not include any comparators as these models do not assess the impact of new treatments. Compared to other model types, a COI model does not estimate costs in future years, based on extrapolations. Instead these models are focused on identifying all of the current costs incurred due to the illness, based on existing treatment approaches and the number of people affected by the disease.
Budget impact model
Budget impact models (BIMs) calculate the financial impact caused by a decision, for example, the introduction of a new treatment option. This would ignore any changes to patient outcomes, such as quality-of-life, and also any costs that would be unaffected by the decision. These models do contain comparators, and the total budget impact is the difference between the costs in the two scenarios. The scenarios will be predictions, because the model is concerned with costs that occur in the future.
What is the purpose of the model?
The first step of building any economic model is to clearly define the question the model should answer and the type of model you need (e.g. cost of illness, cost-effectiveness, or budget impact). This decision should take into consideration what data are available, because in the absence of certain information it may not be possible to build all types of model.