In context, we're talking about removing information from documents, so that partial documents can be released without the removed information being released. The problem is that the PDF format was simply not designed to support that, and the great majority of PDF editing tools (or perhaps all of them??) do not allow for it.
The solution is to create an entirely new PDF, in a way that ensures that the targeted information is not the file. For example, one could take an existing PDF, cover the targeted information with black rectangles, print the resulting document, and then scan the hard copy to create a new PDF.
