Archive for October, 2010

Common scenario:

Patient has been doing well on meds. All of a sudden no longer doing well.

What’s happening?

Hypothesis: The medication(s) stopped working.

Possible explanations:

1. Medication(s) all of a sudden became ineffective. What’s the chance of that? Small.

2. Patient stopped taking the medication(s) as prescribed. What’s the chance of that? High.

Solution: Make sure compliance is optimal BEFORE considering medication changes.

Stages of Change

I find Prochaska and Di Clemente’s model of change a very useful guide for my first encounter with a new patient.

We see people who have issues.

The important question is: on the change continuum, where does your patient place himself regarding his ability to engage in change work? A practical question as the answer will inform your intervention as much as current Axis I symptoms and personality dimensions.

Case and point:

Approaching an already committed to change patient with an educational agenda is poorly timed at best, and, at worst, unproductive and even irritating to the patient. It is an equal waste of time to approach a precontemplator with a list of change promoting interventions (which would be much better suited for patients already are in a preparation or action stage).

Pathologize and Normalize

Common mistake: to take whatever you see at face value.

Most times, most things are more than it meets the eye. Not always (and that is where psychoanalysis went wrong) but most times. A degree of natural shyness compounded by basic socialization rules (including common courtesy) precludes one from washing one’s dirty laundry in public. You don’t need sophisticated psychological theories to accept that people don’t always mean what they say (or viceversa), or that, when it comes to self-disclosure, appearances might be deceiving.

In the specific case of a psychiatric patient this common problem is further compounded by the fact that the hidden issues are very private and, more times than not, either embarrassing or deeply troubling. Would the patient fear being judged? Yes. All reassurances to the contrary will be of little help when this is the case. Is the patient afraid of stigma or consequences following from his decision to disclose? Yes, and unfortunately, this perception is often times, troublingly accurate. Then there are the times when the hidden might lay outside the patient’s sphere of awareness. And then there are the times when the patient chooses to deceive.

A good psychiatrist needs to combine theoretical knowledge about patterns of manifestation of the mind/brain continuum with flair – the experience, “gut feeling” that tell him that there is more to the story.

But flair is neither easy to quantify nor model, and the novice is left with the problem of knowing there is a good chance he might be fooled, yet without a clear solution about how to spot the ‘dig further’ sign posts.

Let me illustrate:
Case scenario #1:

You see a man in his mid 20s, with no prior psychiatric history, with a chief complaint of “paranoia” x4 days, committed for involuntary hospitalization on a 72 hrs hold. You note that the patient spends a long time reviewing the commitment paper. The patient states that the commitment paper provisions are too limiting to the point of being legally abusive. You conclude that the patient’s behavior is consistent with the patient’s ongoing paranoid delusions and think the hold is justified.

Case scenario #2:

You see a man in his mid 20s, with no prior psychiatric history, with a chief complaint of paranoia x4 days, committed for involuntary hospitalization on a 72 hrs hold. You note that the patient spends a long time reviewing the commitment paper. The patient states that the commitment paper provisions are too limiting to the point of being legally abusive. You ask how so. The patient explains that he just took his bar exam, and while he realizes that his view of commitment is a minority view, his take on it is in sync with his libertarian moral and ethical choices. He further explains that the “so called paranoia” is based on his belief that his safety might be at risk following some strong anti-establishment statements that he’d chosen to host on his blog about a week prior to the evaluation. He has since received “a ton of hate mail”. You conclude that the patient’s behavior is normal and break the hold.

Before reading futher, please cast your vote for case 1 or 2.

Now that you made your mind please go on reading:

Case scenario #1:
You contact the patient’s family and find out that the patient is a lawyer who just took his bar exam. You also learn from the family that the patient’s view on mental illness commitment has always been a minority view, in sync with his libertarian moral and ethical choices. The family further explains that the “so called paranoia” is based on his belief that his safety might be at risk following some strong anti-establishment statements that he’d chosen to host on his fairly popular blog about a week prior to the evaluation. He has since received “a ton of hate mail”. The patient, with his family support, self referred for evaluation after experiencing ruminative thoughts and troubles sleeping. They were shocked when informed the patient was placed on a hold.

Case scenario #2:

You contact the patient’s family and find out that the patient used to be a law student, but never graduated from law school, as he became increasingly preoccupied with conspiracy theories. The family further reports a course significant for gradually increased social isolation from family and friends, and overall marginal meaningful functioning. The patient did not have a blog; even if he wanted to, he couldn’t, as his computer skills were not adequate. The family decided to bring the patient in following his repeatedly calling them and stating that the “Feds are out to get me” over the last few days prior to the evalution. They were relieved when informed the patient was placed on a hold.

I hope that by now it is clear that there are dangers in both normalizing and pathologizing. That is, of course, why collateral information matters.

At the same time normalizing and pathologizing can be some of your most helpful guides toward improved data validity. How so?
Examine the patient alternatively though “normalizing” and “pathologizing” lenses.

“This is normal” and “this is clinically significant” are your working hypotheses, equally weighted until proven otherwise.

Your mission is to gather as much data as you can for and against the cases of normal and pathological.

You will weight the evidence at the end. More times than not you will be surprised. When you do a good job, the balance tilts so sharply you might not even need collateral info. When you do get the collateral it will only confirm what you already know.
One last note:

Are you a normalizer or a pathologizer? This is yet another instance when it helps to know who you are. Find out and make an effort to consciously err in the opposite direction. It’s well worth it – think of it as the effort of balancing an uneven coin by adding extra weight to the lighter side.

On the lighter side: balance is the name, and building is the game.

Personality Dimensions

Categorical clarifications have the advantage of clarity. No shades of gray; it’s either black or white. Of course,  this is a quite efficient approach when things are strictly black and white. No ambiguity, no confusion. But these very advantages become weaknesses when the reality actually presents the observer with shades of gray. As there is no in-between, the nuances will be stripped of their defining characteristics and rigidly forced in one of the existing categories.

Such approach remains simple but it is also simplistic in the worst sense. Reality is butchered with gray turned into white or black depending on observer preference or bias. Furthermore, there is a risk that the gray will simply be missed by someone whose conceptual foundation, when it comes to colors, includes only black and white. It’s been shown that our perception of realty is concept based, i.e. no concepts equals no perception. When this is the case the gray will no longer be misclassified (that’s the positive) as black or white as it will no longer be perceived at all (a negative that I believe outweighs the positive with several orders of magnitude).

Case and point: my residents see Cluster A or B or C personality disorders. The better ones even see traits. But not even the best see in-between. When someone does not conform to the DSM typology they could get a “the patient does not have any Cluster A or B or C personality traits” formulation of sorts. In other words, there is black, and there is white, but there is no gray.

My suggestion: think (and look) for dimensions instead of categories. We all have personality dimensions (tones of grey).

The model I use is Costa and McCrae’s Five-Factor Model where the dimensions are:

O – Openness to experience (novelty seeking) versus Closeness

C – Conscientiousness versus Laissez-faire

EExtraversion versus Introversion

AAgreeableness versus Dis…

N – Neuroticism versus Emotional Stability

There are a few advantages of using a dimensional trait (rather than categorical) approach:

1. The factors are based on laborious statistics (factor analysis) and not opinion based constructs.

2. There is evidence about the factors genetic (albeit complex) genetic transmission. Which also helps to nicely bridge traditionally separated domains of assessment/inquiry such as Axis 1 (biological in origin, i.e. amendable through drug treatment) versus Axis 2 (psychological in origin, therefore amendable through psychotherapy).

3. The factors are measurable by testing (also public domain, free of charge online testing is available at a click of a button).

4. The factors have been shown to be universal, culturally independent construct (i.e. can be measured in China as they can be measured in Germany) (McCrae & Costa, 1997).

5. Understanding the dimensions is useful for insight and improvement through therapy (and these personality “observations’ can be shared with the patient right away) (Costa & McCrae, 1992).

6. The factors are stable over decades (average 45 years) (Soldz & Vaillant, 1999).

But even without former testing, keeping the traits in mind while sitting down with your patient, is a more straightforward model that you can use to learn about important aspects of their persona. Rather than basing your learning about the patient on complicated (and only partially proven) models about how the persona and the mind work, it’s just so much more straightforward and easier to have a trait-informed assessment.


Costa, P. T., Jr., & McCrae, R. R. (1992). Normal personality assessment in clinical practice: The NEO Personality Inventory. Psychological Assessment, 4, 5-13.

McCrae, R. R., & Costa, P. T., Jr. (1997). Personality trait structure as a human universal. American Psychologist, 52, 509-516.

Soldz, S., & Vaillant, G. E. (1999). The Big Five personality traits and the life course: A 45-year longitudinal study. Journal of Research in Personality, 33, 208-232.

Here is a rather common scenario. The inexperienced trainee is exposed in his didactics to a variety of methods: short and long-term psychodynamic therapy (with its own separated chapters for drives, ego, self psychology, and object relations theories), cognitive, behavioral, interpersonal, supportive types of therapies. In some places motivational, adlerian, existential and rogerian approaches are also discussed. And then “integrative”, which to the unprepared mind, appears to be a hotchpotch of all the above.

Rather confusing, isn’t it?

The trainee is then taught that all therapies have similar rates of success and outcomes (which, by the way, it is not true) and that the single, most important predictor of success in therapy is neither theory, nor method, but the quality of the patient-doctor relationship (and implicitly rapport).

After all this, the trainee rightfully concludes that therapy is mainly about: 1. “getting the patient to like me” – as a resident once told me (as one needs good rapport, remember?) and then 2. merrily mix and match . Can you blame him?

I can’t. I actually get his point.

That would be just fine if the patient gets better. Unfortunately, hastily, non-systematic, on the spur of the moment interventions rarely result in great outcomes. More times than not the patient will not improve or will get worse. Furthermore, following a practice of  mixing and matching in a non-systematic way, one is at a loss when it comes to deciding what to stop and what to continue.

There is a fundamental risk associated with the above, non-systematic approach: the potential of seriously impairing one’s ability to discern what’s doing what. In other words, whatever outcome one’s intervention produces, there is no way of ascertaining a cause and effect relationship. And without learning, there is no growth.

My recommendation: chose one approach. Study the theory. Study the method. Practice. Practice some more. Think critically. Master the method. Understand the theory. Move on to the next theory.

No big surprise here. A systematic approach and practice are the foundation of any type of understanding.

An interesting question. We live in a day and age when emphasis shifted from psychological to biological theories and interventions. However, it seems like a psychiatrist persona in the public eye, continues to be defined not by one’s pill-pushing abilities but first and foremost by his ability to understand the twists and swirls of another’s mind. If a prescription for a medication follows, that is fine as long as the psychiatrist bases his recommendation on a in-depth understanding of the patient’s mental and emotional workings. Despite all biological advances it remains anathema to pill push based on a straightforward symptom count. Counting is not a core competence of a psychiatrist, understanding is.

Symptoms checklists are barely acceptable in non-psych medical specialties; at least, non-psych doctors can realistically justify the long duration of training as necessary due to the high number of symptoms, syndromes, diagnoses, and treatments that one needs to master. The higher the quantity of data, the longer the training.

That is not the case in psychiatry.

An Aristotelian take on modern psychiatry finds that Axis I genu has only 10 “species”: affective and psychotic disorders, substance use, anxiety, eating, somatoform and factitious disorders, dissociative and impulse control disorders, and adjustment disorders. Of course, there are subspecies variations; nevertheless, this is a rather small number of definitions and categories.

Axis II genu with its 3 “species” (Clusters A, B, C) and N=12 subspecies is not much more complicated. Count this: Cluster A: Paranoid, Schizoid, and Schizotypal Personality Disorder; Cluster B: Antisocial, Borderline, Histrionic, and Narcissistic Personality Disorder; Cluster C: Avoidant, Dependent, and Obsessive-Compulsive Personality Disorder, and Personality Disorder Not Otherwise Specified. Not that hard, is it?

Does one really need 4 years of training (after 4 years of medical school, and 4 other pre-med years) to learn how to identify 23 categories?

The Accreditation Council for Graduate Medical Education (ACGME) and the American Board of Medical Specialties (ABMS) define physician general competences in terms of 6 organizing principles: 1. Patient care; 2. Medical Knowledge; 3. Interpersonal and communication skills; 4. Practice based learning and improvement; 4. Professionalism; 6. Systems based practice.

I am of the opinion that for Psychiatry this order needs to be changed.

#1 should be the interpersonal and communication skills domain. While important for any physician this area of competence is the foundation of a good psychiatrist. A psychiatrist who is not a master communicator faces the risk of gathering invalid and unreliable data – as psych data is mostly self reported and subjective. Thus, the quality of the data directly correlates with the quality of the relationship/rapport.

Poor rapport à poor data à poor assessment à poor intervention.

What are the requirements of a good communicator?

  1. Awareness: Be present. Notice everything. Verbal communication can (and will be) censored, i.e. only partly informative. Always notice the lesser controlled, i.e. more informative, non-verbal behavior. In therapy this is “going after the affect”.
  2. Focus: Give undivided attention. The most frequent reason for distraction is your own agenda.
  3. Curiosity: Let the patient (and not yourself) paint on the canvas.
  4. Empathy: Put oneself in the other’s shoes.
  5. Unconditional acceptance: withhold judgment (N.B. does not equal condoning or withholding action).
  6. Competence: Lead and support.

Everything else can be learnt in a couple of years (incidentally, an Internal Medicine residency takes 3 years).

But developing a “good communicator” skills-set is an entirely different story. This is a life time project and thus, an additional two years might not even be enough for anything else but scratching the surface.

The important point here is that 4 years is way too long for a training limited to descriptive approaches to diagnosis/symptomatic approaches to treatment, but not nearly enough if the goal is to master and use the complexities of social interaction to enhance mental health growth, and mental illness prevention and recovery.

A social brain perspective, emphasizing communication (see the Group for Advancement in Psychiatry 2009 report), is a useful model to bridge apparent opposites (bio versus psycho) and promote a training “climate” change.




What is the purpose of a formulation?

  1. To summarize the pertinent findings/data
  2. To find the common thread bringing together all the different historical (past, reported) and exam-based (present, observed) clinical data
  3. To inform treatment decisions
  4. To make a prognostic prediction

The above principles are valid regardless of the type [of the formulation]. Good biological, psychological (psychodynamic, behavioral etc.) or interpersonal/psychosocial formulations satisfy all the above requirements.

When you have questions.

Sounds obvious but I’ve seen neuropsych ordered because “it would be nice to get some neuropsych done…” Just like that. The more specific your question, the more helpful NP testing is. Not helpful to make a diagnosis for most of the major Axis I disorders, i.e. psychotic, mood and anxiety disorders. Helpful in quantifying cognitive deficits, including memory/dementias, attention/AD(H)D but the diagnosis is made on clinical grounds (i.e. NP provides additional info but it is not a deal breaker).

A list of usual NP applications:

1) Diagnosis clarification (see above)
2) Identification of patient’s strengths/weaknesses
3) Competency/disability
4) Research: baseline data or monitoring
5) Vocational planning
6) Patient care/planning

There is a lot of discussion about the BPS model. Trouble is that in today’s climate this is more a point of discussion of academical interest than a reality informing our day-to-day assessments and treatment plans. Excuses of why this is are plenty: from the severely limited time allowed in most treatment settings for the regular intake and follow-up,  to the dis-interest of managed care in anything that is not clearly linked to medical/biological outcomes and interventions, viz medications.

It turns out that when one compares a patient-centered with a system-centered approach, the former is both more efficient (i.e. allows for more complete data collection) and more effective (i.e. captures more data per unit of time). Counterintuitive, yes – nevertheless, true.

One way to self-monitor one’s exam focus is to jot down the info in a 3 column format: 1st: biological data; 2nd: psychological data; 3rd: social data. Each piece of data would be placed on a line; the more data, the higher their column. An ideal interview/history should produce 3 columns of fairly comparable sizes, while a biologically/system-centered interview will necessarily result in a pretty sizable biological column and rather small psycho-social columns.

Of course this way of organizing the information is cross sectional in terms of the classical categories of anamnesis, i.e. history of present illness (HPI), past psychiatric, substance and medical history, family history etc.

Instead, each of the above categories will be constructed across the 3 BPS axes. For example, an HPI of recent neuro-vegetative symptoms (B), poor self-esteem (P), and recent break-up with the significant other (S), in context of work stress due to a difficult boss (S), which brings up memories of parental emotional abuse (P), will be divided along the 3 axes. By organizing the information accordingly, one can have a visual cue about the overall balance of the BPS contributors. An obvious unbalance can thus be easily noted and corrected.

Last but not least, by organizing the info visually along the 3 axes, dominant themes become more easily noticeable. Finally, a comprehensive BPS case formulation naturally emerges after this simple act of BPS organizing.

A Manual of Psychiatry?

What will you find here?

Tips and pearls for psychiatric fellows, residents and medical students who have psychiatry at heart.  This blog is based on my rounds and questions that I am getting from my students and, at times, my patients.

Being a good psychiatrist is a journey. We all start with baby steps. But baby steps – when it comes to understanding another human being through the filters of modern psychiatry – is not necessarily an intuitive process. 

 The purpose of this is to share some of the things I have learnt and I continue to learn on a daily basis thanks to my patients and students.

This blog is dedicated to all of them.

%d bloggers like this: