Aim

To describe the methodology to incorporate patient preferences in health technology assessment (HTA) at an individual level using simulation modelling, and the associated advantages and disadvantages.

Background

Patient preferences relate to patients’ experiences of healthcare delivery and include elements such as waiting times (e.g. short or no delay vs long delays). Patient preferences can be measured using a range of techniques, including methods such as Discrete Choice Experiments (DCEs). However, these preferences can be estimated at population level or individual level and the issues around added value of estimating and incorporating individual level preferences in HTA is not clear.

Methodology

Conceptual architecture describing the linkage of patient choice modelling (i.e. DCE results) and simulation modelling including the software choice and type of outputs was developed first. Extensive statistical analysis was needed to estimate the individual level preferences from the results of the patient preference survey. In addition, microsimulation modelling expertise was also necessary to incorporate the impact of these individual patient preferences on patient experience (patient utilities), outcomes (e.g. QALYs), resource use (e.g. physician visits and hospitalizations) and costs.

Conclusions

Incorporating individual patient choices in simulation modelling is a complex task involving a series of interlinked work packages and the key to success is to ensure that the format of the outputs for individual patient choices from DCE results are in coherence with the inputs for the simulation model. As such, substantial time and effort is required to incorporate individual level preferences and their impact on costs and QALYs in the modelling efforts. However, the modelling of preferences and their impact at an individual level allow a more granular evaluation of the impact of patient choices. For example, individual level modelling allows the tailoring of the interventions based on the preferences and the expected outcomes for an individual patient.