I have data of an RCT study. It is a two arm study with 3 post-baseline time points. Based on the study design and study objectives, I have ran a MMRM on a continuous outcome variables. The model contains the following variables: treatment + time + baseline + randomisation stratification factors + treatment*time. At each post-baseline time point, I have obtained three mean differences between the treatment arms from the model. My question is - which standard deviations do I use to convert the mean differences into standardised mean differences?