How to find comma without following space in MS Word 2003?

You can use this pattern:

[!<]?,[!>]?

That means: find a comma character between any characters, just not at the beginning of the word and not on the end of the word.

So that in the example below, one would find only the comma at "3,a" (but NOT those at: ",1" and "c,")

,123,abc,

I can tell you a simpler way.
Just replace all , with ,<space> where <space> means a blank space and then replace ,<space><space> with ,<space> to account for places where users did put a space after comma.

WARNING: If you use commas in numbers (either as thousand separator or decimal separator), you need to go through the document to correct numbers such as "1,200" which will now be rendered as "1, 200" .

To search and correct for commas not followed by a space. Check 'Use wildcards' and use following:

Find what: (,)([! ])
Replace with: \1 \2

(Note: There is a space after '!' in Find what)

If your text contains CTL languages/scripts (like Hindi/Devanagari) above does not work (don't know why?), use this instead (will also work with Latin script languages):

Find what: (,)([! ]?)
Replace with: \1 \2

If your text contains poetry then comma may appear at the end of lines (in the stanza), it should not be followed by space, so condition should be added to avoid commas at the end of lines:

Find what: (,)([! ^l]?)
Replace with: \1 \2

If commas appearing at the end of paragraphs should also be avoided then add ^13 (paragraph mark) with ^l, however there is low possibility for this, it is only possible when the typist had given paragraph breaks (instead of line breaks) at the end of each line.

One more case, if your text contains numbers containing comma (to indicate hundred's, thousand's separator) use this to avoid unwanted such commas:

Find what: ([!0-9],)([! ^l][!0-9])
Replace with: \1 \2