How Antidepressants Should be Prescribed

 

As psychiatrists are pressured by various business interests to see more and more patients in less and less time, more and more shortcuts have come into practice. Not a good thing. Especially when it comes to anti-depressant prescribing practices. The drugs have become way over-prescribed, and antipsychotic medications are added to them for "augmentation" far more often than necessary. I’ve spoken about the use of symptom checklists, which are screening tests and not diagnostic tests, being used to make diagnoses.

Another recent development has been companies that offer various “genetic testing” to try to predict which medications might work best in a given patient, or what side effects they are likely to get. For example, GeneSight Psychotropic offers a “pharmacogenomic” test which means that it analyzes how your genes may affect medication outcomes. They say: “The GeneSight test analyzes clinically important genetic variations in your DNA. Results can inform your healthcare provider about how you may break down or respond to certain medications commonly prescribed to treat depression, anxiety, ADHD, and other psychiatric conditions.”  

The problem with these tests is that their predictive validity is actually quite poor – even the company admits that “It is important to note that not all patients who received a GeneSight Psychotropic test experienced improved outcomes.”  According to two docs at the NYU School of Medicine, "The tools were developed using small sample sizes, focusing on specific patient populations...and were not tested in real world settings different from the one in which they were developed."

The tests are IMO a waste of money.

Because people are so complicated biologically, psychologically, and socially, I think a far better way for docs to decide which meds to use is to use pattern recognition: Doing a wide-ranging diagnostic evaluation and looking at what factors suggest certain courses of action.

Let’s start with the diagnosis of major depressive disorder. That diagnosis has to be based on a variety of considerations. Symptoms must be pervasive (almost all day every single day, even if you suddenly win the lottery. I exaggerate, but only slightly) and persistent (at least two straight weeks), and the patient’s functioning and stress responses must be significantly different that their norm. (Antidepressants do not work on chronic unhappiness). All symptoms must take place at the same time – having no appetite on Monday and poor sleep on Thursday does not cut it.

If it’s a patient’s first episode, another big issue is that there is no way for the doc to be absolutely certain that it is not a first episode of bipolar disorder rather than a unipolar depression. This is important because if the doc prescribes an antidepressant to a bipolar patient, it can trigger a manic episode – with disastrous consequences. So what is a doctor to do to make sure that doesn’t happen?

Family history is important, since bipolar disorder usually runs in families. If a patient has a family history, the doctor has to be more careful. The problem is of course that the patient might not know if he or she even has a relative who had a manic episode. So what does a good doctor do? In addition to asking about family history, the doc should carefully review the patient’s history for any symptoms suggestive of the disorder. These symptoms should not have occurred only when the patient was high on cocaine, or during a rage reaction upon having found out that the their spouse and best friend were having an affair!

Next, the doc warns the patient about the bipolar issue, and informs patients that in the event of suddenly feeling revved up, they need to STOP the antidepressant and call the office immediately. This instruction can be given at the same time as the doctor describes all possible major side effects with instructions on what to do should they occur. Then, the doctor needs to have the patient comeback for a scheduled follow up visit, preferably within two weeks.

Now, about picking an antidepressant with which to initiate treatment. As I mention, genetic testing is not particularly valuable. It’s always a bit of a crap shoot because a patient’s presentation may be somewhat atypical, but if the doc takes a complete history, it can suggest which agent to try first. If the patient has some obsessive-compulsive qualities, an SSRI like fluoxetine  or escitalopram is a good first choice. If in addition to that there is a lot of anxiety, the SSRI paroxetine is usually the best choice. If the patient has chronic pain, then duloxetine is usually a good starting point. If sexual side effects are a big concern, buproprion is usually the place to start.

Next, the doctor raises the dose of the medication every three to four weeks until there is a response, or it reaches the maximum dose, or until side effects become too big a problem. At this juncture, the doctor should NOT add another, different class of medicine to “augment” the antidepressant, as is suggested by a lot of Pharma commercials. If there has not been a good response to the first drug, the anti-depressant should be stopped, and a trial of a second antidepressant should come next. And again, if no response, a third one. There are diminishing returns here but eventually most patients will have a decent response. If they do not, then than and only then should an augmentation strategy be instituted. Or the diagnosis may need to be re-evaluated.

Close follow up. Close follow up. Close follow up.