Hi, I want to use MPLUS for analyzing safety data of randomized clinical trials. Adverse event data are often grouped into different body systems. For example, cadiovascular related adverse events, nervous system related AEs etc. I want to consider the individual body system as latent variable, and calculate the latent variable using the observed available AEs. Some AEs are recurrent events. The goal is to test whether the study drug has different safety profile as compared to the active drug/placebo.
Do you have any suggestion on how to apply the MPLUS to evaluate safety profile of a drug in randomized clinical trials?
Because you mention safety profiles, perhaps latent class analysis would be a useful approach, grouping individuals into different types of AE profiles (latent classes). The drug may influence latent class membership.
Thanks for the reply. I may need to talk with you about the statistical issues related to safety data analysis in clinical trials. For the pharmaceutical and FDA statisticians, the safety data evaluation is a challenging task. Your expertise in modeling may help on this issue. By reading the MPLUS manual, I think, MPLUS may help us on safety data analysis. We want some statistical approaches which are simple and understandable by the FDA's clinicians.
I am planning to attend the MPLUS training course at the end of August at JHU. Will you be the training instructor at JHU?