Back to the Top
The following message was posted to: PharmPK
Can someone provide the explicit equations (or a reference) for calculating the lower and higher 95% CI for a sigmoid Emax/EC50 model? I have 95% CI values for EC50 and Hill factor, respectively, and I want to obtain 95% CI for any given concentration. Thanks in advance.
Jack
Back to the Top
The following message was posted to: PharmPK
Jack,
It is not that simple to provide CI for the effect given CI for EC50 and Hill. The simplest way would be to do simulations: assume a distribution that corresponds to the 95% CI that you have (e.g., normal with SD=(CIupper-CIlower)/3.92 if your CIs are symmetric, or somethng similar for the log-normal) and simulate the effect at a given concentration, sampling EC50 and Hill from those simulated distributions. Then, compute CI of the resulting distribution of the effect. If you have not only CIs but the full variance-covariance matrix of the parameter estimates, it is better to simulate taking into account correlations. If you have bootstrap CIs (from the bootstrap analysis of the cocnetration-effect model) you may sample from the bootsrap EC50-Hill distribution.
Leonid
--
Leonid Gibiansky, Ph.D.
President, QuantPharm LLC
web: www.quantpharm.com
e-mail: LGibiansky at quantpharm.com
Back to the Top
The following message was posted to: PharmPK
Leonid,
I now realize this is something likely not doable, because two sets of data could yield a same set of parameters. Appreciate you take time to point it out.
Jack
Back to the Top
The following message was posted to: PharmPK
Leonid,
I now realize this is something likely not doable, because two sets of data could yield a same set of parameters. Appreciate you take time to point it out.
Jack
Want to post a follow-up message on this topic?
If this link does not work with your browser send a follow-up message to PharmPK@boomer.org with "Confidence Intervals for Emax Model" as the subject | Support PharmPK by using the |
Copyright 1995-2011 David W. A. Bourne (david@boomer.org)