The above description is a little sketchy, but an indexing scheme that includes both beginning and ending page numbers for examinations can work. I don't quite understand why one might need that, but nevertheless it can be implemented. In addition to codes at the beginning of each examination, you will need another set of codes in hidden text at the end of each examination.
If I'm following the above description correctly, the index table needs to be set up along these lines. I tested this with three and four examinations and multiple witnesses, and it works perfectly:
WITNESS: <Index Heading On A>
|
|
|
|
<Index On A>
|
<Index Cell 1>
|
-
|
<Index Cell 2>
|
In the body of the transcript, the witness' name needs to be coded thusly, preferably in a swear blurb so this same index can be used with multiple witnesses: <Index Heading On A>John Q. Witness<Index Heading Off A>.
At the beginning of each examination, the attorney's name needs to be coded thusly: <Index On A>Mr. Attorney<Index Cell 1><Index Off A>. If you are using fields for speakers, of course, it would be like this: <Index On A><Field><Index Cell 1><Index Off A>.
At the end of each examination, you need this coding, and presumably you want to put it into a hidden text so it doesn't print in the transcript: <Index On A>Mr. Attorney<Index Cell 2><Index Off A> or <Index On A><Field><Index Cell 2><Index Off A> IMPORTANT: Everything must be the same within this Index-On/-Off pair as at the beginning of the exam, with the exception of the cell number being different. If you put extra spaces anywhere or a different ordering of components, the index will fail to build correctly.
EDITED LATER: Be sure to read Liz's post that follows before getting excited that this is the perfect solution for all situations; it isn't.