probably these files are existing PDF files which you imported to DocuWare, so not scanned or printed, right?
Probably the header in the PDF file is an image and the rest is text. DocuWare extracts the text and is not doing an OCR on the file as there is already text extracted. You can check this in any PDF Viewer like Acrobat Reader. You can try there if you can select the different types there.
There is currently one workaround to use OCR for all types of documents even if they contain text already.