This year’s Cape Cod summer symposia were the occasion for much discussion about important aspects of clinical practice. In this issue, we will describe and summarize some highlights of that discussion. One symposium was broad, and aimed at discussing diagnosis in detail, as well as the existential approach to psychiatry (“Becoming a master clinician: Drugs, diagnosis, and existential psychotherapy”). The second symposium was more focused on psychopharmacology, but aimed at the advanced clinician (“Psychopharmacology: A master class”). In this review, we’ll focus on a question that arose about how to conduct a diagnostic interview. The discussion on that question touches on important aspects of existential approaches to psychiatry as well.
The basic order in which one should assess a case is psychopathology, then diagnosis, then treatment. One cannot know the diagnoses until the symptoms are established; that’s why psychopathology goes first. So don't begin with diagnoses in mind; begin with a search for symptoms. Then later begin to organize your understanding of those symptoms into potential diagnostic groupings. If a diagnosis then fits the psychopathology, you can begin to think about and discuss treatment options. These days, the psychopathology step is skipped usually, as clinicians immediately seek to apply DSM diagnostic groups. The existential tradition always begins with trying to understand the patient’s experiences, to see if they represent psychopathological symptoms, and only then, in some cases, to proceed to a diagnosis.
This diagnostic interview discussion will focus on assessment of mood symptoms, but a similar approach would apply to anxiety or delusional or other symptom groups.
Usually, when you go to the waiting room to find your patient, someone else is there, a family member and friend. About half the time, that other person remains seated. Most clinicians don’t invite that person to join the interview. They should. Not only that, it is wise to inform patients before their first interview that they should seek to bring a family member or friend along with them. This is for two reasons: first, family members and friends can corroborate the history or actually provide accurate history whereas patients, due to lack of insight, often invalidly deny the presence of some features of their personal history, especially manic or hypomanic episodes. Second, when discussing treatment options, the presence of others in the office improves the likelihood that the treatment plan will be understood and implemented; frequently, patients are depressed and cannot clearly understand the complex treatment options being described – family members can hear and repeat what was said later to the patient.
Further, if family members aren’t present, the patient will later have to explain what is said to them; it is better for the family to hear what is said directly from the clinician, rather than secondhand.
If there are concerns about certain confidential topics (which usually do not directly impact on the diagnostic interview anyway), family can be asked to leave the room for a few minutes in the middle of the interview, and then invited to return toward the end when clinicians provide diagnostic impressions and treatment recommendations.
Usually, the presenting complaint is depression. As the psychotherapeutic saying goes, it is best to meet patients where they are; so clinicians should start by the non-controversial and straightforward determination of the presence of clinical depression.
At one level, one simply identifies depression to get beyond it. In other words, it often is easy to know that a patient has a current clinical depression – it may only take five minutes to quickly review current neurovegetative symptoms – but that is not the end of the evaluation, only the beginning. This is because depression is not a diagnosis, but only a constellation of signs and symptoms. Diagnoses are bipolar depression, or secondary depression, or unipolar depression – with unipolar depression signifying that the bipolar and secondary diagnoses have been ruled out. So, as soon as a clinical depressive episode is identified, especially currently, the clinician should stop talking about depression, and shift the focus to identifying past mania or hypomania and assessing possible secondary causes (mainly medical).
It is irrelevant, for instance, to spend much time assessing how depressed the patient is, whether she is hopeless or helpless, whether her symptoms are atypical or typical, and so on. All those features are important perhaps prognostically and therapeutically, but they are unimportant diagnostically. They do not differentially diagnose bipolar as opposed to secondary or unipolar depression.
Obviously an assessment of concurrent psychotic symptoms can be diagnostically and therapeutically relevant, and an evaluation of suicidality is clinically necessary; but soon after identifying depression, the clinician should shift the focus to the more onerous task of looking for past mania or hypomania, and possible secondary causes.
Before doing so, however, PL would emphasize something that is rarely done: evaluate the course of the depressive illness. The reader will recall that the course of depression, unlike the details of current depressive symptoms, can differentiate between bipolar and unipolar conditions, as well as help identify secondary depression. Too often clinicians simply say: “The patient has major depression,” as if that is enough to make a diagnosis. They have no idea when the depressive episodes began, how many there have been, how long they have lasted, precipitating factors, interepisode symptoms, and so on.
Here is what is diagnostically important: age of onset, number of episodes, duration of major depressive episodes, and interepisode status. Ask patients how far back they can remember depression for the first time, refreshing their memory as to the definition of a major depressive episode (daily depressed mood or anhedonia with multiple neurovegetative symptoms, day in and day out, most of the day nearly every day, for weeks on end or more).
Then ask how long their depressive episodes have lasted in the past. Here patients usually throw up their hands and claim ignorance. The clinician can reply, “I could make it up, but your guess is better than mine.” Patients need to know that this is important; force them to think about it. Usually they can give an average duration; it need not be precise – if they are especially exasperated, give them options: more than a month, less than a month, about six months, over a year. These are the time-frames that are diagnostically relevant, since unipolar depression lasts 6 months to a year or longer while bipolar depression is shorter, usually 3-6 months or less. Also assess the durations ideally in untreated periods to understand the natural unmedicated history, but if a patient has been non-responsive to medications, then treated periods should reflect the natural history of the illness.
Then ask how many episodes the patient has had: “How many times in your life have you felt very depressed like that?” Usually, if currently depressed, patients overestimate their past depression: “Forever” is the common desperate answer. “Really?” one could reply, “All your life, every day, day in and day out, without every having one day of being different, forever?” Usually, they back off at that point. “So how many times?” “I don’t know, doc.” Again one can offer multiple choice answers: “Just once? A few times? More than five times? More than 10 or 20 times?” Sometimes it is obvious in the history that the patient has had many episodes, more than 10 or 20; in that case the exact number matters little. What does matter diagnostically is that if a patient has one or two or three episodes, this is common in unipolar depression and uncommon in bipolar disorder. Many episodes is more common in bipolar illness, especially if they are brief (less than 3 months in duration).
Finally, once you have identified the ballpark number of episodes, determine if there are periods of wellness between episodes. This is usually not difficult; either patients will claim they are always depressed, which may reflect interepisode dysthymia or subclinical depression, or they have periods of euthymia, which they or their family can clearly describe: “Were there ever times when you were not depressed and were your normal self, or normal like everyone else, in your mood and energy, for weeks or months on end or longer?”
In the typical interview, it might have taken five minutes to establish that the patient has a current major depressive. Another 5-10 minutes should establish the depressive course of illness. Next, one should take as much time as needed (up to 15 minutes or more) to examine the ins and outs of possible past mania or hypomania. Unfortunately this part of the interview, the most important clinical aspect diagnostically and therapeutically, is often conducted with only a single question or hurriedly. The clinician should take his or her time, and come at the question slowly and in a roundabout fashion, so as to avoid patients’ natural defensiveness about the stigma of bipolar disorder.
PL suggests beginning by an open-ended question, especially if you have established a period of normal or euthymic mood in the past in the assessment of the course of depressive illness: “Did you ever feel the opposite of depressed, where you were not sad and down and depressed, but you also weren’t just your normal self?” With equivocal responses, one might get more specific: “Did you ever have times where you were more energetic than normal, compared to when you were not depressed, or more energetic than most people around you, so that you were doing lots of things or not sleeping much and not feeling tired?” Or perhaps: “Did you ever have periods where you were angry and irritable but not depressed, and full of energy, doing lots of things?”
If a somewhat positive response is elicited, or if the patient comes to the appointment with possible past mania as a clinical question, clinicians can ask an open-ended question, so as not to direct the patient toward manic criteria, but seeking to get their own words about it: “Tell me about how you felt, and how you behaved, or what people told you about how you were, during that time (when you felt hyper or more energetic than usual or where you or your doctor or others said you might be manic or hypomanic)?”
Then, importantly, clinicians should write down what the patient says verbatim. It is very important to do this. Bipolar disorder is such a controversial topic, with patients getting different opinions from different doctors, that it is important to avoid miscommunication by letting patients speak for themselves. Consider if you write: “The patient had elevated mood, with decreased need for sleeping, flight of ideas, distractibility, and increased goal-directed activity for five days.” The patient may disagree and go to another clinician, who is skeptical about the bipolar diagnosis, and that clinician might simply disbelieve my interpretation of what the patient had said. But no one could deny mania if you write: “The patient stated that he would feel ‘hyped up and like I could do anything, I was a tyrant, felt I was smarter than everyone else, like there was nothing I couldn’t do, I didn’t need to sleep for days on end yet I was full of energy, I was giddy at times, my thoughts were all over the place, I couldn’t keep up with them, I would wake up in the middle of the night and clean the house five times over, then the next day I would paint the house inside and outside even though it was perfectly fine, and a week later I would do it again with a different paint color.”
Once a manic or hypomanic episode is identified, the diagnostic process is over: the diagnosis of bipolar disorder has been made. If mania or hypomania cannot be identified, the interview is still not over; then the absence of mania/hypomania needs to be confirmed by third parties. This is done most efficiently if family or friends are present at the interview; if not present, the clinician should call, or ask the family to call, to quickly review mania criteria over the phone. Sometimes you can make this phone call during the patient interview, sometimes later.
These causes are most often medical, though they can also be psychosocial. It is important to distinguish between a trigger and a cause. A trigger is the final event that leads to the episode, but it is not the sole or even the main cause of the episode. This is similar to what Aristotle called the efficient cause. Sometimes there is a certain trigger, sometimes another trigger, and sometimes no trigger. Don’t focus on triggers, though they may be somewhat relevant later especially in psychotherapy; they are diagnostically irrelevant.
Unfortunately most patients, and many clinicians, focus on the most recent psychosocial triggers as if they were absolute causes of episodes. This is a big mistake. The best way to think about this problem, in the PL viewpoint, is to recognize that the brain is a rationalizing machine. The classic split brain experiments in epilepsy show us how: In patients with intractable seizures requiring corpus callosotomy (severing the two hemispheres of the brain from each other), one creates a circumstance where the right hemisphere might note something and yet not be able to transfer that information to the left hemisphere, where verbal control is located. Since the left visual field is transmitted to the right hemisphere, researchers can place the right visual field behind a blindfold, and show pictures in the left visual field, such as violent and anxiety-provoking images. The right hemisphere sees these pictures, and the patient feels scared and nervous. When asked why, the patient says: “Well, my neighbor had a car accident last week and I was thinking about that?” or “I was thinking about the recent war in the Middle East.” In other words, the patient cannot verbalize why he suddenly feels anxious or scared, yet he comes up with an explanation, any explanation, as long as it is plausible, even though it is wrong. We do this all the time: so when patients say there are depressed because of x, y, and z happening in their lives – they may be right, and they may be wrong: we cannot take those explanations as true at face value.
True psychosocial causation, secondary depression due to a psychosocial cause, should be relatively obvious and should be placed in the context of a non-recurrent course of illness: The patient may have one episode after a horrible psychosocial trauma; or maybe two episodes after two horrible psychosocial traumas; but most people, fortunately, do not have many isolated psychosocial traumas (or if they are the victim of recurrent abuse, they usually experience prominent post-traumatic symptoms rather than simply major depressive episodes alternating with euthymia). If recurrent major depressive episodes occur, the psychosocial factors should be seen as triggers, not causes.
The role of substance abuse should be seen in the same way as psychosocial stressors. If a patient never had mania, then takes cocaine for the first time in her life, and then has a manic episode, one can call that condition cocaine-induced mania. But, if she has experienced many manic episodes and many periods of cocaine use, it would seem difficult to demonstrate a one-to-one correlation so as to justify the diagnosis of mood disorder secondary to cocaine use. Often the situation is the reverse: cocaine use frequently occurs in those settings as the result, rather than the cause, of manic episodes.
The same is the case with medical factors. One might have no past depression, then have a stroke, and then experience a major depressive episode. That is post-stroke depression. Most individuals do not experience repeated strokes (since most persons get treated) followed by depressive episodes after every stroke. On the other hand, mild hypothyroidism may be a factor in contributing to recurrent depressive episodes.
Unlike the above scenarios, there is a kind of secondary depression which, when present, is sui generis. Perhaps the most important secondary cause of clinical depression is “vascular” depression. As discussed on the PL website, this condition consists of small microvascular cerebral infarcts, visible on MRI, usually the consequence of hypertension and/or diabetes. This type of depression is common in middle and older age, and has a poor prognosis.
The rest of the history involves obtaining treatment history, family history, and other aspects of a standard medical evaluation. In the interests of space, PL will not examine these topics in detail except to say that treatment history and family history should be examined in more detail than is commonly the case. For treatment history, for instance, the relevant factors needed are not only the names of the medications, but the durations of treatment, doses if available, and concomitant medications used. For each medication, one should assess efficacy and side effects and the reason for discontinuation. Whatever the patient can recall should be documented as well as possible.
At times, patients who claim they cannot remember details are overconcerned with being exact. For instance, when you ask how long they took a certain drug, they become exasperated because they cannot remember if it was 2 months or 3.5 months or 4.25 months. Since they can’t be precise, they will just say they don’t remember. Clinicians can then give them multiple-choice options: “Was it less than one month, more than one month, more than one year, more than ten years?” Obviously this forces them to say something and we can get some valuable data, such as the idea that they took the drug somewhere between 1 to 6 months.
However long it was used, clinicians should ask about whether other medications were taken with it. Often patients can say that other medications were taken, but they do not remember which ones. It is just as important to know that they did not have monotherapy trials, especially with mood stabilizers in bipolar disorder, than to know the details of the other medications they received. Clinicians can then ask whether they thought the medication trial was effective or not, and whether they had side effects. People usually remember side effects more clearly than efficacy, though sometimes they can be clear when a drug was either definitely not effective or definitely extremely effective. It is still clinically very useful to see if one can elicit either extreme response of marked efficacy or marked inefficacy, or the absence of ever having such clear good or bad effects. Sometimes, if a patient has taken many drugs from the same class (such as ten different antidepressants), clinician can simplify the efficacy assessment by asking, “Was there any one of these that worked very well for you, that stands out as especially effective?” Also, if efficacy is reported, one should ask about whether the drug maintained its benefits if used long-term (to assess tolerance or “poop out”).
When a diagnosis is made, clinicians should describe their rationale for that diagnosis, based on symptoms, course, genetics, and treatment response - the four classic validators of diagnosis. Then treatment options can be discussed.
It is important, respectful, and humane to explicitly state the diagnosis to the patient before discussing treatments in any way, and further to elicit a two-way dialogue about the patient’s feelings about the diagnosis.
The clinician should keep in mind, and the patient should be told, that working diagnoses are just that – working; so they are liable to change, and in fact should change sometimes, in the course of months to years of follow-up. In psychiatry, the course of the illness is the final arbiter of diagnosis: “Time will tell whether this diagnosis is right, or whether another one turns out to be the case,” one can say. Time will tell; both clinicians and patients need to be open-minded and revisit the diagnosis over time. A major mistake, often seen in public mental health settings, is that a diagnosis (often schizophrenia or “depression”) is parroted over years, often from clinician to clinician, without ever being re-evaluated, even though the course of illness proves it wrong.
Suggested times will vary based on complexity of a patient’s history. In a newly diagnosed patient, no time is needed for past treatment history, and more time can be spent on course of illness, secondary triggers, and discussion of diagnostic rationale. In a patient with extensive past treatment, the other sections might be somewhat briefer to allow for more extensive examination of past treatment history.