From time to time we are asked to review the care that people with cycling moods have received, trying to figure out, from a complicated story of medication changes and mood cycles, what to do to get someone out of a period of deep pain and dysfunction.
A few years ago I did such a review on a young man with bipolar moods and a history of severe childhood trauma which taught me a lesson about the importance of mood charting, and why it is essential to look at long term medication response (not just the immediate reaction to a change).
Fortunately, this young man had been keeping a very careful mood chart, and all I had to do was download the information from his online spreadsheet and then add in the information about medication changes. 
The chart to the right shows his mood chart.
The hard to see blue line traced his mood (increasing depression) and the yellow line tracked his mental energy (speed of thought).
The medications were added in as colored bars showing when the medication was started and when it ended.
From this chart it was almost immediately obvious that the change that most clearly seemed to be associated with his depression was the addition of olanzapine (OLZ on this chart). There are several reasons why this association was missed. For one thing, studies find that olanzapine is usually associated with the opposite effect – a reduction in depression and anxiety. But the key issue was that right after starting the medication, this young man came in to see his psychiatrist and said “this medication is great, it is really helping…” And he said the same thing a week later… and so neither the psychiatrist nor the patient had any idea that, after that positive response, the medication gradually became the source of a pervasive depression. In other words, that early positive response was a Trojan Horse.