03 October 2013


If someone looks euphoric, can we say they are having a manic episode?

To practice psychiatry, one must become intimately familiar with the Diagnostic and Statistical Manual of Mental Disorders (DSM). This tome lays out the diagnostic criteria for various mental illnesses. It's helpful because it ensures that clinicians are speaking the same language with each other. For someone to be classified as having a manic episode, they have to exhibit a certain number of particular symptoms, such as decreased need for sleep, racing thoughts, or grandiosity. The diagnostic criteria also specify that the symptoms must last for a certain amount of time and in the absence of other potential causes such as drug intoxication.

For many years, the DSM-IV has been the bible of the field of mental health. After fierce debate and negotiations, a new edition came out earlier this year: DSM-5. On the whole, the changes strike me as improvements. For example, there now is a single entity, Autism Spectrum Disorder, that replaces the confusing and seemingly artifical amalgam of five autism-like diseases found in DSM-IV. Under the old criteria for anorexia nervosa, pubescent girls had to have problems with menstruation. This criterion did not seem useful and has been dropped in the newest DSM.

When we were taught psychiatry during our pre-clinical years, we were taught the DSM-IV. But during my clinical rotations now, my attending physicians request that we use DSM-5 criteria. I think this is good. Academic institutions rightfully pride themselves on abiding by best practices.

At the end of my psychiatry rotation, we have to take a national exam that substantially impacts our grade. This exam uses DSM-IV. Part of the reason (I presume) is that it takes years to write new test questions and test them for validity.

What results is an odd and sometimes frustrating contradiction. During the day, I get grilled on the new DSM-5 criteria. But at night when I study, I have to learn the minutae of the DSM-IV, carefully mulling over details that no longer matter.