When you convert a PDF into an editable Word document, the conversion is often full of strange formatting because the 'text flow' is complicated by the need to replicate the 'layout' of the converted text. Sometimes it is worth trying different methods to convert the PDF because some conversions do a better job than others - this depends heavily on the layouts being converted.

It may be as simple as doing the find/replace as suggested by JeremyMarto or perhaps you need to flatten a number of floating text boxes.

Can you post a sample DOCX file to show what exactly you are looking at once the PDF has been converted?

__________________
Andrew Lockton
Chrysalis Design, Melbourne Australia