In search for more confidence in outcomes of health economic models.
In this paper published in top-journal Medical Decision Making, we presented an original and innovative approach to handling uncertainty in health economic models.
The paper was based on collaboration with the wide acclaimed University of Technology in Delft, The Netherlands. read more
ABSTRACT IN PAPER
Objective: Sensitivity analyses are often performed on only a limited number of variables without justification of the choice of variables and range of each variable. External parties such as health authorities are increasingly requiring submission of the actual model, often in order to test the robustness of the outcomes of the model by performing additional sensitivity analyses.
The objective of this work was to develop an alternative method to capture the critical issues of a sensitivity analysis in a health economic model, especially regarding the selection of variables and determining the range for each variable. Apart from external parties such as health authorities, journal readers who want to perform their own sensitivity analysis but do not have access to the model may find this useful. Methods and Results: Statistical methods based on Markov chain modeling and
regression analysis, using the framework of the Taylor series expansion around a point, are used to derive an equation for 1-way sensitivity analyses. In particular, equations for costs and effects are being developed, from which the cost-effectiveness ratio is built. The article shows the feasibility of such equations for the execution of 1-way sensitivity analyses. Conclusion: An equation that can be derived in the manner described in this article provides a substantial amount of information. The inclusion of such an equation in a report may increase transparency of the reporting of outcomes of health economic models.