The Harvard Bipolar Clinic's Dr. Gary Sachs:
A 2005 Interview (re: Bipolarity)

Although not discussed specifically, this interview effectively introduces Harvard's "Bipolarity Index". Here's a link to the whole interview (requires registering with Medscape, pretty painless; register then take the link).  Three excerpts below illustrate Sachs' style, and his three main points:  we should not ask simply "do you have bipolar disorder, or not?";  the answer will not likely come in the first interview; and treatment requires collaboration between doctor and patient.  Notice his sense of humor at the end of the second comment below. 

Point One:  The question in bipolar diagnosis should not be "Do I have it, or not?" but rather "How much of it might I have?"

Medscape: ... It's not a categorical yes or no, you're bipolar or not, but rather, "To what extent are you bipolar?"

Dr. Sachs: Yes, and we're not looking to replace the Diagnostic and Statistical Manual with it. What we're trying to do is answer the patient's biggest question, "How likely is it that I have this disorder?" And it isn't that we're infallible, either; it's just that we are able to approach it more as a continuous issue, rather than as a black-and-white, yes-no.

Point Two:  Making a diagnosis will take a while.  It is not likely to be firm at the end of a first interview. 

Dr. Sachs: ... So we're less concerned about saying, "Oh, you definitely have bipolar I," than saying, "You know, you look a lot like you might have classic bipolar disorder." ...  [It] is much more important to establish a therapeutic relationship and make sure that we approach treatment in much the same way. So we do not tell people that they have to accept our recommendations for the rest of their life. This is, I think, a kind of practice that can be demoralizing to patients and lead to... [less] agreement between patient and care provider about what the appropriate treatment is. So if I ask somebody to accept lifelong treatment with what I recommend when I first meet them, that's like asking somebody on their first date to get married and have a family. It's both unnecessary and unproductive, in my opinion.

Point Three:  Choosing treatment is a collaborative process involving both the patient and the doctor. 

Dr. Sachs: ...So what we'd like to do is -- analogous to getting to a second date if things are comfortable for patients -- establish their individual responses to treatment. We'll begin with the evidence-based recommendations. We'll teach the patients what the profiles of those treatments are like and give them some choice. And then what we do is integrate measurement into the management so that we're not betting that, "Oh, this is definitely going to work." All we're really doing is starting out on a kind of journey with them and having these measurements in which they participate. -- They have a hand in the assessment process as well as in the decision-making about treatment process.