Hi all,

My colleagues and I are working on a project (i.e., health economics model) where we have fitted parametric curves (using Weibull, Exponential, Gompertz, Log-logistic, Log-normal, and Gamma distributions) to Kaplan-Meier data. In the project, we have estimated various survival curves using independent sampling of the curve parameters for each of the distributions mentioned above.

Using the variance-covariance matrices that we have produced from the survival analyses (i.e., Kaplan-Meier analysis), does anyone know how we can use these variance-covariance matrices to incorporate the correlation between the curve parameters into our economic model (i.e., Markov model)?

Ideally, I would want to do this in Excel, but if anyone can even just provide a few sentences on a potential theoretical approach that would be very helpful to get me looking in the right direction for more information. After spending some time researching the issue, the best I've come up with is that if it were a normal distribution, a Cholesky decomposition might have been useful.

Thanks for any help!