In addition to Steve's request for more detail to your question, I'll make some general observations. I infer you are working with clinical research data.
The request for easy is a nonstarter, it ain't gonna be easy. It's not particularly hard, but I would not call it easy.
Typically clinical research data will have a table with visit information, one row per visit per subject. Each visit will have some attributes (sometimes a hundred). The visits are identified by visit number and visit date. Sometimes there are visits that occur between scheduled visits (a real pain, I usually segregate them out and process them separately). You want to take these rows and line them up by sets of columns on visit number. TRANSPOSE will take the maximum number of visit and line up based on that. So, if you are interested in the first 5 visits, but somebody had 15 visits, a naive transpose will generate 15 times x variables. To make that come together, you need to first subset the data from access to get just the rows and columns of interest.
Similarly, if you have a diagnosis table. Every diagnosis is recorded, but you might only be interested in categories of diagnosis. Transform the dx to a category and transpose based on that variable.
There is some broad processes in common here. In the data warehousing domain, it's called "Extract, Transform, and Load" (ETL). You might get some good ideas on how to handle additional problems in the ETL literature.
Doc Muhlbaier
Duke