When patients want feel-good medicine

The request for antidepressants by patients who don't meet the clinical diagnostic criteria is remarkably common. Patients may request these medications for improvement of mood symptoms, work performance, academic performance or social performance. What is the ethical response?

Reply:

The question about prescribing antidepressants for individuals who do not meet strict clinical criteria for a depressive illness comes at a time of close scrutiny of prescriptions for off-label use of medication in general, and psychotropic medications in particular.

On the one hand, there is concern about the influence of pharmaceutical-company money and marketing on prescribing practices, especially where evidence of medical efficacy is thin. On the other hand, there is decreasing stigma for antidepressant use, increasing rates of antidepressant prescribing and extended indications for certain antidepressants in the treatment of, for example, anxiety disorders, smoking cessation, and specific types of pain. Consequently, the number of people who have ever taken an antidepressant has grown. Many may have had exposure to an antidepressant for one indication -- for example, smoking cessation -- and found corollary benefits, such as an increase in energy and concentration. They may want to continue even after the initial indication for the medication has resolved.

The question is especially relevant in primary care. Here physicians often have only 10 minutes to sort through a large number of competing medical demands, of which only one may involve depressive symptoms. Primary care doctors write the bulk of antidepressant prescriptions in the country. Despite the increasing numbers of such prescriptions, undiagnosed and undertreated depression remains a large problem in primary care. As much as 40% of depressive illness in primary care settings may be undiagnosed.

As with so many questions that we frame as ethical concerns or matters of professional integrity, good answers start with good clinical information. One needs to consider requests for antidepressant medication within the context of four diagnostic groups for whom treatment approaches may differ.

Assessing the need

First, there is a group of patients whose depression is inaccurately diagnosed, reflected in false-positives -- patients who don't meet criteria for depression but are identified as such; and false-negatives -- patients who are clinically depressed but not diagnosed. From the mental health perspective, the greater risk is in failing to correctly identify and treat the false-negatives.

Because of the relatively safe profile of second-generation antidepressants, erring on the side of "overtreating" is preferred. The risk in treating a false-positive is low, unless the patient is in the depressed phase of a bipolar disorder, in which case antidepressant use produces an elevated risk of mania. Therefore, it is necessary to assess evidence of hypomania and mania before beginning treatment. You can do this by asking the patient: Has there ever been a period of at least four days when you were so happy or excited that you either got into trouble, or your family or friends worried about it, or a doctor said you were manic? A positive answer should prompt the clinician to follow up with questions clarifying presence of mania.

A second group of patients has subthreshold depressive diagnoses. That is, they do not meet criteria for major depression but present with clinically relevant symptoms. This presentation is often called "minor depression." It may be as frequent in primary care as major depression.

While symptoms are not as severe as in major depression, and patients are not as functionally impaired, minor depression is a persistently disabling condition that is qualitatively similar to major depression and requires careful assessment and monitoring. Bear in mind that differences between symptom counts that are threshold (five or more of the nine relevant DSM-IV depressive symptoms), versus subthreshold (between two and four symptoms), are arbitrary, to some degree.

This category covers a variety of disorders, including the early phase of an evolving major depressive disorder, a resolving-phase major depressive disorder, and chronic depression that doesn't meet any clinical criteria. Not surprisingly, the evidence base for antidepressant intervention in this heterogeneous group is mixed, with some clinical trials showing benefit and others not. The decision here is less straightforward, and the choice of treatment requires a clear discussion between physician and patient about the risks and benefits of antidepressant treatment. Some patients in the misdiagnosed group fall into this category.

Third, there are patients who meet criteria for a diagnosis. For example, there are patients with adjustment disorder with depressed mood, for which no antidepressant clinical trial evidence exists. Often, treatment involves counseling or short-term medication for symptomatic relief, including use of sleep or anxiolytic medications for a limited time.

Finally, there is a group that neither meets criteria for a depressive diagnosis nor has subthreshold symptoms, and that "just miss" a diagnosis. This group, too, is diverse. Bereavement, or grief reaction, a part of the normal human experience, may fall into this category.

The request here is an "ordinary" one -- patients are experiencing distress that is an acute worsening from their normal state. Its severity can vary greatly, and it can potentially evolve beyond grief into the subthreshold or threshold category.

However, there also can be patients who request antidepressants for "extraordinary" uses, in which they are not requesting relief from acute distress but, instead, wish for improved performance beyond their normal state. From this group come requests such as those for "a pill for shyness" or "something to help me work more efficiently." These patients can be quite distressed by their trait-like symptom(s). Unfortunately, though, there is no clinical trial evidence to guide therapeutic decisions.

Things to consider

When assessing whether a request for antidepressants should be granted, the approach should be:

• Clarify in which of the above categories the patient fits.
• Follow up any screening tools that indicate increased risk of depression with clinical assessment.
• Provide information on what is known about the chances of treatment response (and other potential benefits) and what is known about the potential risks, and document the patient's consent. Be sure to discuss treatment alternatives, such as cognitive behavioral therapy and interpersonal therapy, each of which appears as effective for mood and anxiety symptoms as medication.
• Make the decision to treat collaboratively with the patient.

Second-generation antidepressants are relatively safe: While close observation is recommended at first to monitor suicidality, there is no risk of abuse or dependence, the overdose potential is minimal, and serotonin syndrome is rare. Should the patient choose a trial of antidepressant medication, clearly identify target symptoms (self-reported depression instruments are handy to help clarify) and a timetable for monitoring response. Note that for symptoms that do not meet diagnostic criteria, there is no evidence addressing either correct dosing levels, optimal length of treatment, or risk of suicidality.

And be sure to monitor. No response by eight weeks indicates the intervention is most likely not working and should be changed or discontinued. If there is a beneficial response, document it and continue to monitor.

In my practice, nearly all of the patients who are not clinically depressed and present for treatment are those who have subthreshold symptoms that are clinically relevant and, sometimes, functionally impairing. For me, informed consent is the guiding ethical principle. We review what limited evidence exists regarding antidepressant treatment of their psychiatric presentation, then we discuss treatment alternatives, the symptoms we want to alter and a time frame for monitoring. We talk about "stopping rules" -- for example, how we will know that it's time to stop the medication trial. And we agree that this is a joint collaboration that requires clear and ongoing communication between patient and physician.

It is rare that patients ask for a pill to help them with a single, "extraordinary" symptom, such as shyness or low energy. In the absence of any direct clinical trial evidence for this use of antidepressants, I'm most informed by the literature that suggests greater long-term mood and anxiety benefit from cognitive behavioral psychotherapy than from antidepressant medications. This indirect support also reflects my clinical experience -- for a very discrete symptom target, the overall cost/benefit ratio of good psychotherapy beats that of an antidepressant. In these situations, I recommend a trial of psychotherapy.

Assuming that a person is able to give informed consent, I don't think a trial of an antidepressant in this situation is inherently unethical -- it's just not what I believe is the best clinical decision, and it's not an intervention I'd provide.

A primary care physician's comfort in managing these requests can vary. Some may be quite comfortable that they have the time, training and resources to pursue diagnosis and management on their own -- at least initially. Others may prefer to refer to a psychiatrist for a second opinion. Either decision is sound.

Bradley N. Gaynes, MD, MPH, associate professor of psychiatry at the University of North Carolina, Chapel Hill, School of Medicine; researcher on the assessment and management of psychiatric illness in primary care medical patients.