Sensitivity analysis

[minji kwak]

I am trying to conduct a sensitivity analysis for my glm with gamma distribution model to estimate the impact of a disease on cost. I would like to conduct a sensitivity analysis to address the measurement error (misclassification) of the disease. So far, the best way that I can do is to use parametric bootstrapping and do monte carlo simulation. However, I can't find a good way to conduct a parametric bootstrapping with probability of the disease that i obtained from other literature.
Can anyone help me how I can approach this? Is there any good reference that I can look into? I do understand the concept, but need a very thorough technical assisstance.

Thanks.

[Marcos Almeida]

Welcome to the Stata Forum / Statalist.

Perhaps you could use a Bayesian approach with different sorts of priors

[minji kwak]

Hi, Marcos,
Thank you for your reply. It is a good suggestion.