My second post in this series on sloppy thinking in psychiatry is devoted to polypharmacy, the medical term for prescribing multiple medications at once, especially for the same problem. Polypharmacy is at best a risk thoughtfully taken because nothing simpler and safer will do. At worst it’s a dangerous error, exposing patients to unnecessary hazards purely as a result of laziness and sloppy thinking by their doctors. Unfortunately, the latter is all too common in psychiatry. Let’s look at why.
It has been said that the less we know about an illness, the more treatments we have for it. Instead of one definitive cure that attacks the root of the problem, various remedies ease symptoms — not the cause — often via different mechanisms. A good example of a definitive cure is a specific antibiotic to treat a bladder infection. We know how bacterial infections work, and we have antibiotics to attack the root of the problem. Ancillary treatments for fever or pain are sometimes used, but they are clearly secondary, and often optional. In contrast, the pathogenesis of psychiatric disorders is not known, thus we have no treatments to attack the roots of these problems. For example, antidepressants affect neurotransmitters that appear implicated in depression, but the exact way these neurotransmitters relate to the syndrome of depression is unknown. Thanks to our ignorance, we have medications that affect serotonin, and others that affect norepinephrine and/or dopamine. In recent years atypical neuroleptics (antipsychotics) have been approved as add-ons for treating depression, a worrisome development given their risks.
Since we don’t have a definitive cure for depression, many patients report partial (or minimal) improvement from any one medication. The prescriber may then add another on the theory that it may help via a different chemical mechanism — a theory that is difficult to confirm or refute, as we don’t know the mechanism in the first place. The original medication is not stopped: If the patient improves, why disrupt a winning combination? And if the patient doesn’t improve, we wouldn’t want to withhold an antidepressant from a depressed person, would we? Sloppy thinking all around, yet sadly common.
Similar arguments can be made for the treatment of bipolar disorder and schizophrenia. Lacking a true understanding of pathogenesis, we treat empirically. And empiric treatment, while often compassionate and necessary and helpful, invites the shaky logic of adding more medications hoping for more empiric benefit.
Compounding and worsening this situation is psychiatry’s abandonment of parsimony in diagnosis and clinical assessment over the past 30 years. Prior to the publication of DSM-III in 1980, psychiatric evaluation was an attempt to explain a patient’s seemingly unrelated complaints using a single theory (often psychoanalytic, but possibly biological or even behavioral). The introduction of phenomenological diagnosis in DSM-III encouraged multiple diagnoses in the same patient, say Major Depression and PTSD on Axis I, and a personality disorder on Axis II. There was no longer any attempt to tie it all together. This has encouraged a piecemeal approach to treatment: a medication for depression, a different one for PTSD, maybe something for sleep, and something else again for agitation due to the personality disorder. That’s four different psychiatric medications already, and we’ve hardly even started. Patients with personality disorders often complain of “mood swings,” so let’s add a mood stabilizer like lithium or Depakote. And they’re anxious, so we could add a benzodiazepine tranquilizer like Ativan, or a beta-blocker like propranolol, or an atypical neuroleptic. Or what the hell, all three! We’re up to seven or eight medications now, and we haven’t even considered a stimulant for their ADHD — because, after all, the patient is having trouble concentrating… funny how it was never diagnosed before. And we haven’t augmented the antidepressant with thyroid supplementation, nor have we added a second antidepressant…
While 10+ psychiatric medications is clearly over top, I’ve evaluated a number of patients who arrive on six, often an (1) antidepressant, (2) mood stabilizer, (3) tranquilizer, (4) sleep aid, (5) stimulant, and (6) another antidepressant or mood stabilizer. Almost without exception, I’ve been able to cut this list in half, and in some cases down to zero, or more often, one medication. It’s less a matter of expert medication choice, and more an aversion to sloppy thinking. According to one study, antipsychotic polypharmacy can be simplified without harm 2/3 of the time.
Psychiatric polypharmacy is often intellectually lazy. Needless to say, there are far more drug combinations than there are studies assessing the risks and benefits of these combinations. Polypharmacy is nearly always an educated guess, not “evidence based medicine.” It’s not even good single-case research, where one would ideally change a single variable at a time. All too often, medications are added to treat the side-effects of other medications, as with “ADHD” in the case above, a tail-chasing exercise that only gets worse over time. With every added medication there are added side-effects, and sometimes adverse interactions that can be more harmful than the original problem. In my experience, generic side-effects such as weight gain and cloudy thinking are more the rule than the exception in patients taking multiple psychiatric medications. It should happen a lot less than it does.
Once again, photo courtesy of Petr Kratochvil.