I think it depends on what your question is? For example, if you're trying to compare differences in translation rates, it's unlikely that you will be interested in the UTR, and I think a few of the ribosomal profiling papers even exclude the first section of the CDS, because it will have an initiation peak that's not necessarily related to overall translation rates. However, it might be interesting to look at UTR associations to see whether there are peaks of binding associated there etc. But I wouldn't include it by default in a first-pass analysis unless there was a compelling reason to.