Scanned text in OCR-documents not found by DEVONthink

Sometimes happens this: I scan documents with OCR and import them to DEVONthink. But some text will be found by Spotlight but not by a query in DEVONthink.

What is wrong…?

Could you give an example of a query that worked in Spotlight but not in DEVONthink? And in such a case, in DEVONthink select the searchable PDF and choose Data > Convert > to (plain or rich) text. Examine the query term as it occurs in that text. Was there perhaps a run together of characters, so that one or more characters prefix the query term in the text? If so, DEVONthink wouldn’t find it, unless the Wildcard symbol “*” were prefixed to the query term.