Embracing psychiatric uncertainty

I always get troubled looks from psychiatry residents when I point out that our field is the domain of the uncertain and the not-well-understood — and that it will always remain so.  As soon as the cause of a disease is known, it automatically leaves psychiatry for another specialty.  General paresis (advanced syphilis), once identified as an infectious disease, became the domain of internists.  Senility (dementia), multiple sclerosis, and many other apparently psychiatric conditions went to the neurologists.  Thyroid disorders belong to endocrinology.  Brain tumors and hemorrhages are surgical conditions.  And so forth.  I have little doubt that schizophrenia will someday be understood as due to a slow virus, a complex genetic error, or something else.  At that point it will no longer be a psychiatric condition.  It will join neurology, internal medicine, or some other specialty.

This makes my residents squirm in their seminar chairs, particularly when I point out that the closest analogy to psychiatry's status in medicine is philosophy's status among the humanities.  Philosophy consists of questions in the humanities that we don't yet know how to answer.  Once we do, that area is no longer considered philosophy.  "Natural philosophy" is what we now call science.  It isn't considered philosophy anymore.  Logic was one of the classic branches of philosophy; now it is better understood as a branch of mathematics.  In the same way, psychiatry consists of questions about human thoughts, feelings, and behavior that we don't yet know how to answer, not down to the level of mechanism anyway.  Once we do, that area is no longer considered part of psychiatry.

It's no mystery why the residents are uncomfortable.  They want and expect certainty.  Why did they study all that organic chemistry, memorize all the bones and muscles, spend years learning to diagnose and treat, if in the end they can't make definitive statements about their chosen specialty?  Many will cling to pseudo-certainties for reassurance.  Simple-minded factoids like "alcoholism is a disease" or "depression is due to a chemical imbalance" give them something to hang onto.  Unfortunately, we don't really know what causes depression, and alcoholism is disease-like in some respects, but not in others.  Most of our field is complicated, messy, and not well understood.  Moreover, this need for certainty in an uncertain field leads many psychiatrists, including and perhaps especially those well out of training, to convey unwarranted confidence regarding diagnosis and treatment recommendations.  We can come across as smugly self-assured.

Frankly, this very uncertainty — mystery, if you will — is one of the things I like about psychiatry.  It isn't a settled area.  It is endlessly debatable, much like an undergraduate philosophy course.  Yes, there are concepts and terms to learn, principles to refine and employ, scientific studies to evaluate.  There is a body of knowledge, a history, practice guidelines to teach and learn.  Most of all, there are real patients to help.  Yet as in philosophy, experts in psychiatry can and do disagree.  Our diagnostic categories are revised periodically. Treatments come and go.  Unscientific fads influence the field, as when American psychiatrists used to diagnose schizophrenia more liberally than our British counterparts, when multiple personality disorder suddenly became common in the 1980s and just as suddenly faded away, and in the way ADHD, PTSD, and bipolar diagnoses are so popular now.

Confident pronouncements of certainty have no place in psychiatry.  Humility is the only honest attitude to take to this work.  At the same time, the questions we face are fascinating, patients are suffering, and neither can wait for definitive knowledge.  We must do the best we can with imperfect knowledge, with limited data and educated guesses, with hunches and subtle impressions.  As in life generally, we cannot wait for certainty before acting.  As in life generally, this makes psychiatry risky, vibrant... alive.

Diagnostic alphabet soup

Earlier this year a reader asked me: "I would be very interested to hear your thoughts on patients becoming too focused on diagnoses. [...] While I was in an RTC as a teenager, and recently in the hospital as an adult, I have found that people almost treat their diagnoses as a competition. I was calling it the alphabet olympics. I also have a friend who will rattle off a bunch of abbreviations for his diagnoses. There is always something new popping up too. Sometimes I wonder if over diagnosing is a mistake some psychiatrists make."

I've seen this too.  Here's my take on the alphabet soup of diagnosis, and whether it's good for patients to focus on it.  First, a little history...

Prior to 1980, before the revolutionary 3rd edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-III), psychiatry tended to lump disorders into a few broad categories.  Schizophrenia covered a wide range of presentations, from relatively minor symptoms to devastatingly severe ones.  Depression could be brief, prolonged, triggered by obvious stressors or losses, or appear out of nowhere.  Neurosis referred to any presumed unconscious conflicts that interfered with life.

DSM-III changed all that.  (An excellent historical review article, in pdf format, is available here.)  This was the first effort by the American Psychiatric Association (APA) to publish an atheoretical, phenomenological psychiatric nosology.  What do these $10 words mean?  The idea was to create diagnoses that could be used regardless of one's school of thought or theory.  For example, some psychiatrists thought depression was biological, others considered it psychological.  Either way, if a patient had a low mood for two weeks, along with poor sleep, appetite, concentration, and libido, he or she had Major Depressive Disorder according to DSM-III.  It didn't matter why.

This scheme encouraged multiple diagnoses.  A given patient could fulfill criteria for Major Depressive Disorder, an Anxiety Disorder, a Personality Disorder, and other disorders, all at the same time.  This reflects a drawback of atheoretical diagnosis.  An underlying theory, such as Freudian psychoanalytic theory, or a systematic biological or learning theory, can pull together apparently disparate symptoms into a coherent diagnostic formulation.  Without such a theory to guide diagnosis, each set of symptoms stands on its own.  While some DSM diagnoses had exclusion criteria — they could not be listed in the presence of other diagnoses — this still left plenty of opportunity to list multiple disorders in the same person.

Each edition of the DSM grows in size.  One reason is that scientists can't stand to leave a good category alone — if it can be turned into two good categories.  Thus, anorexia and bulimia, which used to be one disorder, are now divided.  Depression is divided into major depression, dysthymia, seasonal affective disorder, adjustment disorder with depressed mood, and so forth.  Bipolar disorder comes in Type I and Type II, as well as lesser versions.  I am not against making these distinctions when there is good reason to do so, and there often is.  But one consequence is diagnostic alphabet soup: a growing set of arcane labels usually shortened to three- or four-letter abbreviations.  And the nature of atheoretical diagnosis means that any given patient may qualify for several.

Many psychiatrists feel they "understand" a patient better if they can establish one or more DSM diagnoses — although, being atheoretical,  such diagnoses don't actually explain anything.  They do, however, point reassuringly to recommended treatments, usually pharmaceutical.  Moreover, medications are FDA-approved for each of these indications separately.  This has marketing advantages for drug manufacturers.  Shyness doesn't sound like a psychiatric problem to be treated with medication, but "Social Anxiety Disorder," essentially a synonym for shyness, does.  Dividing anxiety into Generalized Anxiety Disorder, Social Anxiety Disorder, and many other types created markets for various medications.  In a parallel fashion, health insurers demanded more specific diagnoses in order to pay for psychiatric treatments.  There is money, and therefore politics, behind dividing human misery in these particular ways.

Perhaps the most interesting part of my reader's question is why some patients are attracted to these labels.  Her experience with teens and young adults may, in part, reflect embracing these labels in an ironic or mocking way:  "Now I have MDD, OCD, and PTSD.  Isn't that a kick?"  Probably more relevant is the concrete way a diagnosis seems to account for one's frightening instability.  Better to be "ADHD" than merely a scattered teen who can't study.  The former confers scientific legitimacy, promises specific treatments, and even justifies entitlements such as extra testing time in school.  These labels can also ease personal responsibility and humiliation, as when outrageous social behavior can later be attributed to Bipolar Affective Disorder or some other "chemical imbalance."  Despite the persistent stigma of psychiatric diagnosis, these labels have enough psychological and practical advantages that some patients wear them proudly.

The downside to all of this is that individuals can become known, even to themselves, by impersonal diagnostic labels.  Knowing oneself as PTSD, ADHD, and/or OCD can dehumanize.  It can prematurely close off inquiry and self-reflection.  And DSM diagnoses do not actually explain anything; they are better conceptualized as statistical categories.  Such diagnoses are useful tools, but like all tools they can be misused.

Politics, religion, and ADHD meds

Cross-posted from Technorati with permission.

At a dinner meeting a couple of weeks ago I met two psychiatrists who work at Kaiser Permanente, the large HMO system that boasts a 24% health insurance market share in California.  (This has nothing to do with my story really.  I just think it's amazing that a quarter of all insured Californians are Kaiser patients.)  As we described our practices, I mentioned that I recently helped a patient stop his Adderall, the amphetamine combination drug given for Attention Deficit Hyperactivity Disorder or ADHD.  The patient had come to me on a very high dose and was complaining of many side-effects: anxiety, muscle twitching, severe insomnia, weight loss.  I gradually tapered and eventually discontinued the stimulant over several weeks, with resolution of most of these symptoms.  He thanked me and said he felt much better.  I related this story with some pride, and mentioned to my dinner companions that I've had more success stopping high-dose medication, especially for ADHD, than I've had starting ADHD medication.  Moreover, I opined that ADHD is too readily diagnosed in adults, resulting in a lot of unnecessary amphetamine being prescribed.

I had not anticipated how odd these statements sounded coming from a psychiatrist.  One of the Kaiser docs, a child psychiatrist, quickly noted how many kids she'd helped by identifying and treating their ADHD.  It's an under-recognized problem, she assured me.  The other psychiatrist only saw adults, yet he too underscored how Adderall, Ritalin, and other stimulants helped his patients.  Someone mercifully changed the subject, and we let the matter drop.

It got me thinking though.  First, could we all be correct?  I have no reason to doubt the experience of child psychiatrists who see their young patients perform better, achieve more, and get along better with others when treated for ADHD.  I don't see children myself, and am basically a bystander in the debates over medicating children for ADHD and behavioral problems.  Moreover, even in adult psychiatry I believe that prescribed stimulants can sometimes help, not only for ADHD but also for depression in the severely medically ill, and in some other situations.

But my own experience has led me to be cautious.  "Adult ADHD" is a fad.  Its rate of diagnosis and treatment have skyrocketed in recent years, for no good scientific reason.  I get calls all the time from people who have diagnosed themselves using a simple online checklist and are seeking an MD to rubber-stamp an amphetamine prescription.  Since amphetamines are performance-enhancing even in normals and have street value as drugs of abuse, these potential patients put me in the uncomfortable position of second-guessing their request.  I'm not saying adult ADHD doesn't exist — in fact, I'm sure it does — but this isn't the kind of relationship I want with patients.  So I tell callers I don't do ADHD evaluations, leaving me with lingering regrets about thwarting the subset — I don't know how big it is — who have a legitimate need for this treatment.

And frankly, I've seen these medications over- or mis-prescribed by my fellow psychiatrists on a number of occasions.  In addition to the patient mentioned above, for nearly a decade I've seen a distinguished senior academic for medication treatment of anxiety and depression.  He's never had an ADHD diagnosis.  Nonetheless, he asked his psychiatrist back in 1993 to add Ritalin because a relative with ADHD benefited by it.  He's been on it ever since, 16 years.  I tapered the dose down by more than half, but my patient resists using less, even though it likely worsens his anxiety and he needs medication for sleep.  At this point I expect he's on it for life.  Another patient of mine, a young woman without an ADHD diagnosis but with a history of anorexia, had been prescribed 40 mg of Ritalin daily by her prior psychiatrist.  It helped her concentrate, but also suppressed her appetite, which was a major psychological issue for her.  We tapered down the Ritalin and discontinued it over about a year and a half, at which point she was doing well and reconciling with her mother — who, my patient said, had abused Ritalin herself for a long time.

Well, you get the idea.  I'm not opposed to psychostimulants, honest.  And I do believe ADHD is a serious problem and that it responds well to medication, along with other treatment.  However, I also believe that, in adults anyway, inattention and hyperactivity can mean lots of things.  I believe stimulant medications that cause anxiety, insomnia, loss of appetite, teeth grinding, high blood pressure, and other side-effects ought to be used judiciously.  (And I also believe that a patient needs to have had symptoms by age 7 to fulfill DSM-IV diagnostic criteria for ADHD.)

Some of the most contentious, polarized arguments in psychiatry revolve around ADHD and its treatment.  Are we poisoning our children with stimulants?  Or leaving thousands to suffer unnecessarily?  So far, these arguments still generate more heat than light.   The main thing I learned at the dinner meeting I attended is that, like politics and religion, ADHD and its treatment remains a touchy topic in polite conversation.

Borderline personality disorder: parasuicide

In my last post, I highlighted diagnostic challenges related to borderline personality disorder (BPD): Sometimes dramatic, self-destructive behavior leads to reflexive, inaccurate use of this label, while other times eagerness to diagnose a medication-responsive illness such as bipolar disorder can lead to overlooking BPD.  Naturally, this barely scratches the surface.  Thousands of books have been written about BPD.  This editorial from the May 2009 issue of the American Journal of Psychiatry provides a concise summary of controversies surrounding the diagnosis.  Even the Wikipedia entry on BPD has extensive useful information.  Today I'll focus on another central feature of BPD that has proven challenging to residents I've supervised (and me): parasuicide. Parasuicide refers to self-harming behavior identified by the patient as suicidal but unlikely to actually result in death.  Sometimes termed a "suicide gesture," typical examples include taking a handful of pills, and cutting one's wrists to draw blood, but not deeply enough to damage veins or arteries.  Often the patient realizes later that suicide was not "really" the aim of the behavior.  (Aim and intent become complex philosophical issues once the idea of a dynamic unconscious comes into play.  Can one intend something without knowing it?  Can intent be discerned by a therapist over the patient's heartfelt disagreement?)  Many patients in therapy eventually describe a very unpleasant, difficult-to-name emotional state that is relieved by these activities — especially painful, self-mutilating actions such as cutting or burning oneself.  There is a sense of tension release.  A communicative aspect is also often apparent, as in showing one's anger or rage to important others, and eliciting an emotional reaction from them in return.

Parasuicide puts families and therapists, especially beginning therapists, in an uncomfortable position.  These actions must be taken seriously, as failure to do so can make the person feel (further) abandoned and even angrier, leading to a spiral of increasingly self-destructive behavior.  "Upping the ante" in this way can even lead to accidental death.  For example, it is not widely appreciated that even modest overdoses of acetaminophen (Tylenol) can cause lethal liver failure.  A seemingly minor overdose can unwittingly prove fatal.  On the other hand, parasuicide looks manipulative.  It is loudly claimed to be suicidal but isn't "serious."  Families and therapists become angry themselves, potentially resulting in isolation, retaliation, and further harm to the patient.  Patients brought to the emergency room after parasuicidal behavior challenge the on-call psychiatrist, often a resident, to walk a tightrope between dismissing the risk too casually, versus overreacting on the principle of "better safe than sorry."  Patients are sometimes hospitalized unnecessarily.  A fascinating theoretical paper on psychiatric risk assessment can be found here.

To its credit, dialectical behavior therapy (DBT) tackles parasuicide head-on, as its top priority.  This is wise not only from the perspective of patient safety, but also as a means to contain the anxiety of treatment providers.  It is very difficult to work collaboratively with a patient who both scares and angers the therapist.  While DBT addresses the problematic behavior itself, the manner in which a patient induces such negative feelings in the therapist is a direct focus of psychodynamic therapies.  Transference-focused psychotherapy (TFT), another empirically validated treatment for BPD, was developed by Dr. Otto Kernberg and colleagues at Cornell, and pays particular attention to the communicative aspect of parasuicidal acts.

Parasuicide may look and feel manipulative to observers, but to the person with BPD it is a desperate attempt to secure relief from painful overwhelming feelings.  It is both highly characteristic of the disorder, and one of its most challenging clinical features.

Borderline personality disorder: diagnosis

Just as I was formulating a few thoughts on borderline personality disorder (BPD), I see the NY Times beat me to it. Jane E. Brody's 6/15/09 "Personal Health" column, "An Emotional Hair Trigger, Often Misread," provides an evocative description of this vexing disorder. Brody's column seems informed largely by her consultant, Dr. Marsha M. Linehan, who devised the best known and best studied treatment for BPD, a combined individual and group therapy called Dialectical Behavior Therapy, or DBT. (Here are some links describing DBT: 1, 2, 3). Dr. Linehan also invited readers' questions about BPD on a related NY Times blog, garnering over 200 comments. She began to answer some of those questions here. In this post I'll offer some of my own views on diagnosing borderline personality disorder, and in the next I'll share some more reflections and thoughts on BPD.

The term "borderline" came from the impression of early clinicians that the disorder originates at the border of neurosis and psychosis: too severe to be the former, not severe enough to be the latter. Over the decades psychiatry has refined its understanding of this syndrome , yet much remains unclear. The May 2009 issue of the American Journal of Psychiatry was devoted in part to BPD. One article by psychiatric diagnostician John Gunderson MD reviews the history of the diagnosis and is well worth reading.

BPD is not as easily diagnosed as people, including many clinicians, think it is. Not all dramatic, irritable, self-destructive, and/or manipulative people have BPD. I currently see two patients in my practice who were referred to me by other well-trained psychiatrists as clearly having BPD. They don't. One is a woman who suffered repeated childhood sexual abuse, leaving her full of mistrust, anxiety, and anger. She hardly discussed her traumatic past with her former psychiatrist of many years, who saw her weekly and maintained her on several antidepressant, tranquilizing, and sedating medications. I confess that I, too, thought she had BPD when we first met: She was overwhelmed by affect and seemed unable to sustain relationships. This has all changed with therapy. Now, about two and a half years later, she takes no psychiatric medications, has several important relationships, and usually can tolerate her own strong emotions. Either I cured her BPD (I don't think so), or she never had it in the first place. My other patient has dramatic affective storms, identity diffusion, frequent limit-testing, "manipulative" suicidal threats, and so forth. But psychotherapy has revealed emotional conflicts, not borderline pathology, at the root of her distress. She too is improving.

I have no doubt that Dr. Linehan's DBT helps a great many patients suffering from BPD. But I can't help but worry about all those who do not really have BPD, and who could be helped in more fundamental ways by a nuanced understanding of their emotional dynamics. It is worth remembering that Dr. Linehan herself does not claim that improvement from DBT is diagnostic of any particular disorder. Who would not benefit by increased mindfulness, improved interpersonal effectiveness, and better emotion regulation and distress tolerance? (These are the four "modules" of DBT.) Perhaps some variant of these modules should be taught to all schoolchildren as a public health measure!

So there are people who "look" like they have BPD, but really have neurotic conflicts. Conversely, I have seen a number of patients who carry a diagnosis of bipolar disorder, usually qualified with terms like "atypical" or "rapid cycling," who really have BPD. The world of psychiatry is divided into those who believe bipolar disorder is under-diagnosed, and those who believe it is over-diagnosed. (The same is true of ADHD and other popular [trendy?] diagnoses.) I happen to believe it is over-diagnosed. Rapid-cycling bipolar is defined as four or more extreme mood states per year. These would be moods that last at least a week or two, usually considerably longer. Dramatic mood swings that occur hour to hour, or day to day, are most likely something else: a personality disorder, an organic brain condition, a drug or alcohol addiction. It's a waste and a risk to take unneeded bipolar medication for years and years, surely worse than undergoing DBT for a mistaken BPD diagnosis. Worst of all, I suppose, is to be given both diagnoses, bipolar and BPD, when neither is correct. I am very wary when patients tell me they have both disorders. While not impossible, it far more likely points to sloppy diagnosis than to a particularly unlucky patient.

The term "borderline" has seeped into public consciousness just enough to make it a powerful putdown, or pseudo-explanation to account for a socially difficult or antagonistic person. Moreover, the DSM-IV does a poor job with personality disorders, perhaps because it aims to be atheoretical, whereas personality assessment relies inherently on a theory of mind. I believe a psychodynamic framework is required to understand BPD, even if effective interventions need not themselves be psychodynamic.

More reflections to follow in the next post.