Bipolar Disorder

Death by DSM

When DSM sets doctors and patients up to err.

Posted Dec 14, 2009

A few years ago, a middle-aged man had been newly diagnosed with depression (major depressive disorder, MDD; in contrast to bipolar disorder). His doctor treated him for a few years with different antidepressants, mostly without success. Finally, the man wrote his wife a note; they had no children. It's not your fault, he wrote; the pain is just too much. He killed himself.

The husband consulted a lawyer who called me. There is a strange fact in the history, the lawyer noted, after sending me some records. I looked at the chart. There, about a year previously, after starting an antidepressant, the patient had become hyperactive, overtalkative, spent a lot of money, did not sleep and was not tired, and was giddy - all this lasted for 2 weeks. It had never happened before; it never happened again, at least not that way.

The doctor diagnosed a manic episode, and added a neuroleptic to the antidepressant. The patient calmed down, and felt better for a while. Then the patient wanted to come off all medications for a while to see how he was; the doctor concurred. A few months later, on nothing, the doctor's noted documented racing thoughts and poor concentration. Another antidepressant, a few months of increasing depression, and then the end.

The doctor's expert witnesses, senior psychiatrists at prominent universities, argued that there was no legal malpractice. The doctor had followed the text of DSM-IV: mania, if induced by an antidepressant, does not count as bipolar disorder. Besides, they argued, the patient was 40 years old, and had no prior documented depression or mania, and bipolar disorder usually begins around age 20. The doctor's record had noted a period of sad mood and trouble sleeping at age 25, for which the patient received antidepressant treatment. Was that a first major depressive episode? No, said the doctor's experts: just understandable "demoralization" related to work stress.

The legal standards are high for malpractice, which is understandable. Negligence means not doing what is the competent standard of care in one's community. DSM-IV represents the standard of care, and the doctor followed it. There was no legal case.

But the patient still died. The doctor's defense was that DSM-IV was followed. Perhaps then DSM-IV was at fault.

Some researchers are skeptical about the reality of antidepressant-induced mania because some small studies do not find a difference between antidepressant and placebo (but a statistical law is that failure to find a difference is not the same as no difference), and because other studies find low rates (about 10% in one recent large study, equivalent to placebo, but that study used mood stabilizers, which are known to decrease risk of antidepressant induced mania; thus, one cannot conclude that antidepressants alone do not cause mania). Here is the stronger research evidence, I believe, that antidepressant-induced mania occurs, and that it almost always means the patient has bipolar disorder. Depending on the study population (e.g., type I vs type II) antidepressant-induced mania happens in about 5-50% of patients with bipolar disorder. According to almost all studies of carefully diagnosed MDD (i.e., with standardized interviews to rule out type I and type II bipolar disorder), the rate is much less than 1%. For instance, in a standard large recent study of MDD (STARD,with about 3000 patients), which used 4 different cycles of antidepressants one after another, mania cases only happened in less than ten persons, or about 0.3%. The relative difference is huge: 5-50% divided by 0.3% means that if antidepressant induced mania occurs, the patient is at least 166 times more likely to have bipolar disorder than MDD. (That is about 10 times the likelihood of getting lung cancer if one smokes cigarettes).

In 1994, when DSM-IV was written, the judgment was made to exclude antidepressant-induced mania from the definition of bipolar disorder, mainly to avoid the horrible bogeyman of overdiagnosing bipolar disorder.

When the current psychiatric system of diagnosis was set up with DSM-III in 1980, the authors clearly stated that they hoped that future revisions would change the system for the better, that research would guide judgments, and that new ideas would not be avoided simply because they were new. Instead, what has happened is that the DSM system has become a Bible, and some psychiatric leaders, like Allen Frances who headed DSM-IV, are explicit in their opposition to any changes unless overwhelming reasons exist to make those changes. This is nothing by a mental law of inertia, an opposition to change for the sake of opposing change.

I do not agree with those who bash DSM, partly as a means of bashing the profession of psychiatry. DSM itself is not the problem; all medical professions have diagnostic criteria (DSM like criteria exist for migraine and epilepsy and rheumatoid arthritis). The problem is not the method of diagnostic criteria; the problem is not paying attention to science.

Unfortunately, this kind of ideological conservatism led to a loophole in DSM-IV which appears to have led to the demise of this person. The doctor was not to blame, to the extent that he followed the rules. Perhaps, then, those rules need to be called into question.