When To Not Prescribe An Antidepressant?

I encountered that interesting question just last week.  Antidepressant medications have been around for a long time at this point and they have an increasing number of indications.  Everywhere around the Internet there are algorithms that make prescribing these drugs seem easy and automatic.  Qualify for the vague diagnosis and follow the line to the correct antidepressant. At the same time there is plenty of evidence that their use is becoming less discriminate than in the past, largely due to the use of checklists rather than more thoughtful diagnostic processes.  It is common for me to encounter people who have been put on an antidepressant based on a "test".  When I ask them what that test was it is almost always the PHQ-9 or GAD-7, checklist adaptations of DSM criteria for depression and anxiety that can be completed in a couple of minutes.  There is a significant difference between the checklists and the diagnostic process as I pointed out in a previous post about the sleep question on the checklist compared with more detailed questions about sleep.  The list that follows contains a number of scenarios that will not be accessible by a checklist.  In those cases a more thorough diagnostic assessment may be indicated.

1.  Intolerance of antidepressants - Every FDA package insert for medications includes this warning, usually referring to an allergy or a medical complication from previous use.  In addition to allergic reactions (which are generally rare with antidepressants), there a number of significant problems that preclude their use.  Serotonin syndrome can occur with low doses and initial doses in sensitive individuals.  In the case of the more potent classes of serotonergic medications - the SSRIs as many as 20% of patients will experience agitation, nausea, headaches, and other GI symptoms.  By the time that I see them, they will tell me the list of antidepressants that made them ill and that they cannot take.  It is an easy decision to avoid medications that are known to make the patient ill.

2.  Behavioral intolerance of antidepressants - SSRIs in particular can have the effect of restricting a person's emotional range to a narrow margin.  They will typically say: ""I don't get low anymore but I also don't get as happy as I used to get."  A person who is affected in that way finds that to be a very uncomfortable existence.  Many have been told that they will "get used to it" - a frequently used statement about these medication related side effects.  I have never seen anyone get used to a restricted range of emotion and I tell them to stop it an not resume it.  I avoid prescribing antidepressants form that class and that class is typically SSRIs.

While I am on the topic, I frequently use the following vignette when discussing the concept of "getting used to" side effects:

"Many years ago I treated a man who came to me who had been taking a standard antidepressant for about 7 years.  He was not sure that he was depressed anymore.  He was sure that he had frequent headaches and very low energy.  I recommended that we taper him off the antidepressant and see how he felt.  He came back two weeks later and said: 'Doc - I feel great.  For about the last 6 years I felt like I had the flu every day and that feeling is gone.'  That is my concern about 'getting used to a medication'.  It may mean that what you really get used to is feeling sick.  That is why I encourage everyone that I treat to self monitor for side effects, and if they happen we stop the medicine and try something else."

That advice sounds straightforward but it is not.  I still get people who think that they need to "get used to" a medication and will only tell me in a face-to-face interview.

3.  Lack of a clear diagnosis - many of the people I see were started on an antidepressant during an acute crisis situation like the sudden loss of a significant person in their life or a job or their financial status.  There is no real evidence that antidepressants work for acute crisis situations, but some doctors feel compelled to prescribe a medication because it makes it seem like they are trying to help the patient.  I have also heard the placebo response rationalized for these prescriptions.  A similar cluster of symptoms can be observed along with the associated anxiety, but in the short term the main benefits to be gained will be from medication side effects like sedation rather then any specific therapeutic effect.   The real problem is that the medications don't get stopped when the crisis has passed.  I may be seeing a person who has been taking an antidepressants for ten years because they had an employment crisis or divorce at that time and have been taking the medication ever since.  They have been tolerating the medication well for that time, but it now takes a lot of effort to convince them that they don't need the medication and taper them off of it.

I try to prevent those problems on the front end by not prescribing antidepressants for vague, poorly defined emotional problems or crisis situations where they are not indicated.  In my experience, psychotherapy is a more effective approach and it helps the affected person make sense of what has been happening to them.

4.   An unstable physical illness is present - that can mean a number of things.  The commonest unstable physical illness that I routinely deal with is hypertension with or without tachycardia.  Patients and their doctors will often go to extraordinary lengths to avoid treating hypertension even hypertension that is outside of the most current and most liberal guidelines.  I am told that the person has "white coat hypertension".  How do they know that is all that they have?  Have they ever had a normal blood pressure reading outside of a physician's office?  Would they be willing to purchase their own blood pressure device, monitor their blood pressures at home and bring me the readings?  I have had people become angry at me because of these suggestions, even after a thorough explanation of the rationale.  It is almost like patients expect a psychiatrist to hand them a magical pill that takes care of all of their problems.  As an example the following warning if from the FDA package insert for milnacipran but most antidepressants don't include this warning - even when they might affect blood pressure:

"Elevated Blood Pressure and Heart Rate: Measure heart rate and blood pressure prior to initiating treatment and periodically throughout treatment. Control pre-existing hypertension before initiating therapy with FETZIMA"

There are a number of conditions ranging from glaucoma to angina that need treatment before antidepressants can be safely prescribed.  In some cases I am not happy with the pharmacotherapy for associated medical conditions.  Desiccated thyroid rather than levothyroxine for hypothyroidism is a good example.  Why is desiccated animal thyroid gland being used in the 21st century instead of the specific molecule?  In many cases, I will refer the patient to see a specialist and they will never come back because their real problem has been solved.  I posted about cervical spine disease some time ago after I had a number of patients come in for treatment of depression.  What they really had was insomnia from cervical spine disease and when that chronic pain was addressed their depression resolved completely.

I will run into some situations where I insist the patient see a specialist (generally a Cardiologist) to get an opinion on safety of treatment.  This used to be called "clearance" by the Cardiologist but for some reason that term has fallen out of favor.  I think the "clearing" specialists don't want the designation, but from my perspective the patient is not going to get the antidepressant that we discussed unless the Cardiologist agrees.

5.  The patient prefers not to take the medication - I think that patients are often surprised at how easily they can convince me to not prescribe a medication.  Many expect an argument.  I will supply them with the information they want and direct them to reputable sites on the Internet where they can read as much as they want about the medication.  I am very willing to discuss their realistic and unrealistic concerns.  I will attempt to correct their misconceptions  and also provide them with my real life estimate of how many people tolerate the medication and the common reasons why people stop it.  I fully acknowledge that I cannot predict if a medication will work for them or give them side effects.  At the end of that discussion, if they don't want to try the medicine that is fine with me.  I have absolutely no investment in prescribing medication for a person who does not want it.  If the person has clear reservations, I let them know they don't have to come to a decision right in the office - they can go home and think about it and call me with their decision.  I am never more invested in the medication than the person who is taking it.  I will also provide them with feedback on whether or not their decision seems reasonable or not.

6.   Additional patient preferences -  Many people will talk with me about antidepressants and say that they want to solve their problems with psychotherapy, exercise,  or some other non-medical option.  Many people will also talk about drugs, alcohol, cannabis,  hallucinogens, psychedelics, and other drug based treatments for depression.  I can offer people what is known about the scientific basis of treatments for depression and encourage effective non-medical treatment where it is indicated.  I do not endorse the use of the use of alcohol or street drugs for treatment and let people know that I cannot prescribe antidepressant medication if those other substances are being used.  That includes "medical marijuana".  There is a risk for serotonin syndrome with various combinations of stimulants, hallucinogens, and/or psychedelics in combination with antidepressants.  Some web sites that profess to provide neutral advice to people who want to experiment will often have some posts on how to mix these medications to get enhanced effects.  None of that advice should be considered safe or reliable.  It is an indication to me that the person cannot be expected to take the prescription reliably.

7.  Context - very important consideration.  Seeing a person who has just survived a suicide attempt in the intensive care unit is a much different context than seeing a long  line of people who are dissatisfied with life for one reason or another.  Twenty three years of acute care work taught me that medical interventions are much more likely to work for clear cut severe problems than vaguely defined problems.  There are many people who are looking for a fast solution to difficult problems.  When I suggest to them that environmental factors need to be addressed or that they may benefit from psychotherapy or even more explicitly that psychotherapy will work better for your problem than medications - I am often met with resistance.  Common replies are that they cannot commit that kind of time or energy to psychotherapy.  Since most managed care companies discriminate against psychotherapy - many will tell me that their copays are too expensive.  If I point out that their work schedule or job is the problem - they will give me many reasons why they can't change it.  Treatment becomes conditional - as in - I am hoping that this antidepressant will work because I cannot change my life in any reasonable way and I can't do psychotherapy.

8.  Commitment to treatment - too many people come in and expect the prescription of an antidepressant to not only solve the problems but that nothing else is required of them except to show up for an occasional appointment.  If I want to see medical records like exams, labs, imaging studies, ECGs. EEGs, pharmacy records or other information it is generally not an option.  I need that information before any prescription occurs.  The same is true if I need to order these tests and see the results.  I am quite capable of having a discussion of the costs of these orders and that is why I have a preference for not repeating tests and looking at existing results.  That does not prevent the occasional complaints about how I am interested in making money off the person by ordering basic tests, even though I do not get anything at all for ordering tests.

The other part of treatment does involve agreeing to take the medication reliably and following the other recommendations that can be very basic.  If someone tells me that they are drinking two pots of coffee per day and they are anxious and can't sleep but are unwilling to stop the coffee because: "I know that I can drink two cups of coffee and still fall asleep" - I am probably not going to be able to do much with an antidepressant.  The same is true for somebody binge drinking a 12-pack of beer every night after work.  The effects of common substances like caffeine and alcohol are contrary to the goals of treating anxiety and depression with or without medication.

9.  Mania - it is possible for people who have taken antidepressants for years to become manic either while taking the antidepressant regularly or when the antidepressant has been disrupted.  Even though the incidence of mania from antidepressants is low and the treatment of bipolar disorder depressed includes an antidepressant-atypical antipsychotic combination (olanzapine-fluoxetine combination or OFC) stopping the antidepressant acutely is the best idea.  Many people discover at that point that mood stabilizers seem to work much better for their periods of depression than antidepressants.                              
10.  Misunderstanding the treatment alliance - fortunately treating depression and anxiety is not like treating standard medical problems.  Most office visits for new general medical and surgical problems are one or two visits in duration.  A medication is prescribed and it either works or it doesn't.  When it doesn't the problem either resolves on its own or becomes a chronic problem.  One of the best examples anywhere is acute bronchitis.  Over the past decades - tons of antibiotics have been prescribed for no good reason.  Acute bronchitis generally resolves on its own in young healthy people.  I try to be very clear with people that their response (good or bad) to the medication is in no way guaranteed.  I let them know that these medications are moderately effective at best and then only in the hands of somebody who knows how to rapidly switch them up and in some cases augment them.  Even then there will be some people who do not respond.  The key to all of that treatment is communication and it may require significant patience on the part of the patient.  It may also require more frequent appointments then they anticipated especially is associated problems like suicidal thinking and psychosis are also being addressed.

Those are my thoughts about the question of who I would not prescribe an antidepressant to.  I hope to transform those thoughts into dimensions in a useful graphic.  Feel free to let me know if I missed anything.



George Dawson, MD, DFAPA