When somebody says: maybe it isn’t bipolar

The diagnosis here is not closed by one test or at one appointment — it gets refined over years. In a large survey of people with bipolar disorder in the UK (2,458 responses, people put themselves forward), the median time from first telling a clinician about it to being given the diagnosis was nine and a half years. So “maybe this isn’t it” is not a malfunction. It is what this road looks like for almost everyone.

Questionnaires do not settle it. In a study that took 98 people who came out positive on the commonest bipolar questionnaire, 23.5 per cent had bipolar disorder and 27.6 per cent had borderline personality disorder. A result like that, on its own, names neither one.

This app deliberately makes no feature out of any of it. Attempts to teach a machine to separate the two from data like this stop at around 58 per cent — a little better than a coin. So there is no test here and no “this looks more like…”, and there will not be: getting this question wrong costs a person years of their life.

A second opinion is not a complaint about a doctor, and not a search for somebody who will say something nicer. It is an ordinary part of the road. What weighs most in it is not a story told from memory but what was written down: when each thing started, how long it lasted, what sleep was doing then, what the medication did and how fast, what people nearby noticed. That is exactly what the report in this app prints — it can come along, so nothing has to be gathered at the appointment from scratch.

And two things at once, because together they are more honest than either alone. A changed diagnosis is not a demotion and not “all those years were a mistake”: it is the same history, read more carefully. And a changed diagnosis is not automatically the right one either — it goes on being refined too. There is deliberately no list here of the signs by which “one differs from the other”: in life they overlap so much that such a list adds confidence more often than accuracy.