TY  - JOUR
PY  - 2011//
TI  - Uncertainty analysis is inherently Bayesian
JO  - Value in health
A1  - Stevens, John W.
SP  - 202
EP  - 203
VL  - 14
IS  - 1
N2  - The recent article by Barendregt [1] raised several statistical issues and failed to cite relevant articles on the analysis of uncertainty in economic models.   The author failed to cite the work of Claxton et al. [2], who state that the choice of distribution for uncertain inputs should be guided by the form of the data, the type of parameter and the estimation process, and a discussion that details the issue of reflecting uncertainty in economic models.   The author suggests that parametric bootstrapping is used to propagate parameter uncertainty through an economic model whereas the most common approach is to use Monte Carlo simulation [2], although other approaches are available when economic models are computationally expensive [3, 4].   The author failed to distinguish adequately between epistemic and aleatory uncertainty. Epistemic uncertainty arises from a lack of knowledge and may be reduced or even eliminated by obtaining more information, whereas aleatory uncertainty is due to randomness and is irreducible. The methods for estimating a confidence interval for an incremental cost-effectiveness ratio described by the author are frequentist methods for patient-level analyses and are not relevant to a discussion on the analysis of economic models...<p /> <p>Language: en</p>
LA  - en
SN  - 1098-3015
UR  - http://dx.doi.org/10.1016/j.jval.2010.10.002
ID  - ref1
ER  -