The simple answer is to ask someone else, or better yet a group; if D is small, then D^2 or D^4 will be infinitesimal. However, delusions are "infectious" (see Mass hysteria), so this is not really a good method unless you're mostly isolated from the main population.

The more complicated answer is to track your beliefs and the evidence for each belief, and then when you get new evidence for a belief, add it to the old evidence and re-evaluate. For example, replacing an old wives' tale with a peer-reviewed study is (usually) a no-brainer. On the other hand, if you have conflicting peer-reviewed studies, then your confidence in both should decrease and you should go back to the old wives' tale (which, being old, is probably useful as a belief, regardless of truth value).

Finally, the defeatist answer is that you can't actually distinguish that you are delusional. With the film Shutter Island in mind, I hope you can see that almost nothing is going to shake delusions; you'll just rationalize them away regardless. If you keep notes on your beliefs, you'll dismiss them as being written by someone else. People will either pander to your fantasy or be dismissed as crooks. Every day will be a new one, starting over from your deluded beliefs. In such a situation there's not much hope for change.

For the record, I disagree with "delusional disorders being quite rare"; I believe D is somewhere between 0.5 and 0.8. Certainly, only 3% of these are "serious", but I could fill a book with all of the ways people believe something that isn't true.