Monitoring and Assessment Framework for the European Innovation Partnership on Active and Healthy Ageing

What is MAFEIP


The MAFEIP monitoring framework comprises a web-based tool which rests on the principles of Decision Analytic Modelling (DAM). More specifically, it is based on a traditional Markov model, an approach that is commonly used in health economic evaluations to assess the impact of healthcare innovations in terms of health outcomes and resource use.

Based on data introduced into the tool, which may be (preliminary) data from clinical studies, expert opinions and your own views, this model performs an incremental analysis of the impact of your innovation. This means that it estimates the changes in healthcare resource use, societal resource use and health related quality of life that result from using your innovation instead of current care. As a result you would need data on both the current care situation for your target population, as well as the situation in which your intervention is used.

The Model


The outcomes for the intervention and the standard care are calculated by simulating the health status of the target population. This is done by simulating the transition of the target population between the states defined in the Markov model.

The MAFEIP tool allows you to select the number of states (3, 4 or 5), that better adapts your intervention.

In all the Markov models, the baseline state represents the general health status of the target population. The disease/impairment states reflect the health status of people who experience the condition of interest (the condition that the intervention aims to prevent, relieve or cure). Depending on the number of states selected, you will be able to define different stages of the disease or health condition (e.g. if you select 5 states you can define three stages of progressive disease severity, as the patient’s health condition worsens).

The Markov model is completely defined by the initial distribution among the states, the transition probabilities and the parameters associated with each state.

With respect to the transition probabilities, taking the 3-state model as an example, a transition from the baseline to the disease/impairment state represents a patient becoming ill (i.e., the incidence of the health condition). When a patient experiences an improvement in his/her clinical conditions, as in the case of disease remission, it is defined as the transition from the disease to the baseline state (i.e., the rate of recovery).

At any time in the simulation patients can die. This is represented by a transition to the dead state from the baseline state (i.e., baseline mortality in the target population) or the disease state (i.e., excess mortality in the population with deteriorated health).

On the other hand, each state of the model is defined by an amount of resource use (costs) and quality of life (utility or health outcomes). This represents the average resource use and quality of life of a patient in that health state.

The simulation considers a hypothetical cohort of patients beginning the process with some distribution among the starting states. At each cycle of the model the appropriate transition probabilities are applied and the distribution of patients in each state is adjusted. As the model runs over a specified time horizon (i.e., several cycles), it is possible to estimate the incremental costs and outcomes associated with a particular intervention.


The impact evaluation methodology


In order to estimate the impact of your intervention, the defined model runs twice: a) with parameter estimates for the respective intervention under assessment; and b) with parameters corresponding to the standard care scenario.

These two scenarios can differ in terms of the transition probabilities, as well as the utility weight, healthcare costs, and societal costs attached to the states. Therefore, the impact of the intervention on the quality of life (health outcomes) and resource use can be represented though multiple parameters in the model.

When the model simulates a hypothetical cohort of patients moving between the states over time, the differences in survival, utility and costs will accumulate to an estimate of the incremental costs (ΔC) and health effects (ΔE) that can be expected from the intervention under evaluation. As a result, the tool estimates the cost-effectiveness of an innovation versus its respective standard care alternative.

In order to easily grasp the evaluation outcome, the overall impact of the intervention is shown using a cost-effectiveness plane: the Incremental Cost Effectiveness Ratio (ICER) of the intervention under assessment is displayed in comparison with the Willingness to Pay (WTP) threshold in order to facilitate decision making.