It doesn't really matter whether the 'problem name' is the same for all - you fix them one at a time; there really is no more practical way. Having done OCR corrections numerous times, I'm aware the OCR process might mis-spell the same name a number of different ways; you simply need to handle each instance as it arises. The only thing saving any of the data might achieve is to provide you with a selection of replacement names to pick from. That's a fair coding effort for relatively little gain.

__________________
Cheers,
Paul Edstein
[Fmr MS MVP - Word]