I am looking for quantitative information regarding misdiagnosis of
mental illness.  Information regarding any or all of depression,
bipolar disorder, Alzheimer's, and ADHD would be most useful.

In particular, I would like to know if there are studies that give
information along the following lines:
1.) Rate and nature of misdiagnosis (false positive/false negative) 
   (e.g. "Depression goes unrecognized XX percent of the time" or "XX
% of people who are treated for depression actually do not have it")

2.) Cross-correlation diagnosis studies 
   (i.e. % agreement of diagnosis between two different doctors or
groups of doctors for any or all of depression, bipolar, Alzh., and
ADHD)

Information for misdiagnosis in both psychiatrist and general
practitioner doctor populations would be good.

The better-documented the answer the better (ie, I'd prefer refereed
journal articles to CNN articles)  I am willing to pay extra for
well-documented answers.

Everyone I know is a bit crazy, so misdiagnos is a matter of opinion.
Since the huge success President Bill Clinton had with lies, half
truths, and false inferrences, even the hard sciences are having
trouble identifing truth. Psychitry sociology and psycology are not
exact sciences and some of the leading figures in these fields are of
doubtful sanity.   Neil

To some degree, you're right. However, there are fundamental
biological bases behind some conditions, so it's not entirely a matter
of opinion that these things exist and that there are treatments for
them. I don't doubt that the diagnoses and the treatments are of
varying success and effectiveness; I was trying to get a handle on HOW
effective or ineffective the diagnoses are.  Thank you for your
comments.

Thanks, 
- Marco