Without more details, it is hard to say whether you can do this or not.  It depends on what you mean when you say you want to predict for new times.   Are you looking to predict at new survival times, that is, extrapolate beyond the event times already contained in your data?  If so, the proportional hazards model is not an appropriate model to do this.  It would simply extend the survival time beyond the last time in the model.   If you have a "new" time within the range of the event times used to build the model, then you could estimate the survival at these "new times" by adding these new observations to the DATA= dataset.  You can define a new variable on the DATA= dataset to be used in the FREQ statement of PHREG.  This variable would take on a value of zero (FREQ=0) for the new observations and also assigns the original times a value of one (FREQ=1).  Note that you will, of course, need to provide covariate values for the new observations.  The survival estimates for these new observations will be in the OUTPUT data set. ... View more

Yes, that is a reasonable approach. ... View more

Yes, the Pearson Chi-Square test will test the equality of the two proportions as detailed in this usage note. 22561 - Testing the equality of two or more proportions from independent samples If you want an exact test, then you can report Fisher's exact test. ... View more

It looks like to me that you have the steps correctly specified.  The only issue is that you did not standardize the data prior to running the GENMOD step.  This would effect the betas as well as the MSE.  I would recommend using REG instead of GENMOD as well so you can easily pick off the Sigma0_2 using the RMSE. ... View more

One option might be to make use of the WEIGHT statement in Proc CAUSALMED to run with one of the replication methods.  The general approach would be to use one of the SURVEYXXX procedures to compute the replicate weights and then to run CAUSALMED with each set of replicate weights.  There is an example of doing something similar with Proc GENMOD.  You could likely adapt this example by substituting the CAUSALMED step for the GENMOD step.  You might need to do some research to see if this would be statistically valid, but I have seen others use a similar approach with BRR for mediation models. SAS/STAT Poisson Regressions for Complex Surveys ... View more

Take a look at this usage note. 23087 - Odds ratio estimate is not equal to the exponentiated parameter estimate - exp(Beta) Essentially, the default parameterization for the CLASS variables involves effects coding and thus the odds ratios are no longer merely exp(beta) and interpretation of the parameter estimates is no longer simply the log odds comparing the level vs the reference. ... View more

The short answer is that no, FMM will not report the exponentiated estimates for the mixture probability model(s).  That being said, it isn't clear what exactly you mean by subgroup membership.  The mixture is a mixture of components and not different models based on some group membership.  I am not sure of what use an odds ratio would be in relation to the various components as I have  never seen anyone report it or clearly define it.  In some cases, you might want to predict the component membership based on some rule(usually the highest posterior probability), but this would not really be related to any kind of odds ratio calculation. ... View more

In general, the difference will be in the manner in which the variances are computed.  Complex survey data uses methodology that adjusts for the design (cluster and strata) effect.  Data that is merely weighted will use the usual formulations for weighted variances.   ... View more

To be clear then, the data sets are identical in all the variables--trt,baseline, week6, week12, week18, week24 and in the same order with missing values on the same observations? If you can add the SIMPLE option to the MI statement and post its output along with the Missing Data Patterns output, then that might be helpful. ... View more

It isn't clear what you mean when you say that the data sets are different, but the observations are the same.  Can you elaborate what exactly is different and what exactly is the same in the two cases? Also, could you post the two sets of codes you are using? ... View more

The long format is not usually compatible with performing  multiple imputation, thus data restructuring from long to wide or the reverse is often needed for multiple imputation.  There is a good discussion in Raghunathan's Missing Data in Practice text (2016) pages 121-126 and in the 2018 paper linked below. Using SAS for Multiple Imputation and Analysis of Longitudinal Data ... View more

No, you are interpreting it correctly.  I was wrong about what was contained in that data set. In any regard, hypothesis testing in finite mixture models is not very well defined because it is difficult if not impossible to derive the asymptotic distribution for the mixture likelihood.  There is a paper that discusses the problem and makes a few suggestions related to goodness of fit tests that might work.  Anything that they propose would not be available in SAS, but you might be able to program it yourself. Hypothesis testing for finite mixture models - ScienceDirect   ... View more