Thought Broadcast

I have a confession to make.  I don’t think what I do each day makes any sense.

Perhaps I should explain myself.  Six months ago, I started my own private psychiatry practice (one of the reasons why I haven’t posted much to this blog, but I hope to pick up the pace again!).  I made this decision after working for several years in various community clinics, county mental health systems, and three academic institutions.  I figured that an independent practice would permit me to be a more effective psychiatrist, as I wouldn’t be encumbered by the restrictions and regulations of most of today’s practice settings.

My experience has strengthened my long-held belief that people are far more complicated than diagnoses or “chemical imbalances”—something I’ve written about on this blog and with which most psychiatrists would agree.  But I’ve also made an observation that seems incompatible with one of the central dogmas of psychiatry.  To put it bluntly, I’m not sure that psychiatric medications work.

Before you jump to the conclusion that I’m just another disgruntled, anti-medication psychiatrist who thinks we’ve all been bought and misled by the pharmaceutical industry, please wait.  The issue here is, to me, a deeper one than saying that we drug people who request a pill for every ill.  In fact, it might even be a stretch to say that medications never work.  I’ve seen antidepressants, antipsychotics, mood stabilizers, and even interventions like ECT give results that are actually quite miraculous.

But here’s my concern: For the vast majority of my patients, when a medication “works,” there are numerous other potential explanations, and a simple discussion may reveal multiple other hypotheses for the clinical response.  And when you consider the fact that no two people “benefit” in quite the same way from the same drug, it becomes even harder to say what’s really going on. There’s nothing scientific about this process whatsoever.

And then, of course, there are the patients who just don’t respond at all.  This happens so frequently I sometimes wonder whether I’m practicing psychiatry wrong, or whether my patients are playing a joke on me.  But no, as far as I can tell, I’m doing things right: I prescribe appropriately, I use proper doses, and I wait long enough to see a response.  My training is up-to-date; I’ve even been invited to lecture at national conferences about psychiatric meds.  I can’t be that bad at psychiatry, can I?

Probably not.  So if I assume that I’m not a complete nitwit, and that I’m using my tools correctly, I’m left to ask a question I never thought I’d ask:  is psychopharmacology just one big charade?  **

Maybe I feel this way because I’m not necessarily looking for medications to have an effect in the first place.  I want my patients to get better, no matter what that entails.  I believe that treatment is a process, one in which the patient (not just his or her chemistry) is central.  When drugs “work,” several factors might explain why, and by the same token, when drugs don’t work, it might mean that something else needs to be treated instead—rather than simply switching to a different drug or changing the dose.  Indeed, over the course of several sessions with a patient, many details inevitably emerge:  persistent anxiety, secretive substance abuse, a history of trauma, an ongoing conflict with a spouse, or a medical illness.  These often deserve just as much attention as the initial concern, if not more.

Although our understanding of the pathophysiology of mental illness is pure conjecture, prescribing a medication (at least at present) is an acceptable intervention.  What happens next is much more important.  I believe that prescribers should continue to collect evidence and adjust their hypotheses accordingly.  Unfortunately, most psychopharmacologists rarely take the time to discuss issues that can’t be explained by neurochemistry (even worse, they often try to explain all issues in terms of unproven neurochemistry), and dwindling appointment times mean that those who actually want to explore other causes don’t have the chance to do so.

So what’s a solution?  This may sound extreme, but maybe psychiatry should reject the “biochemical model” until it’s truly “biochemical”—i.e., until we have ways of diagnosing, treating, and following illnesses as we do in most of the rest of medicine.  In psychiatry, the use of medications and other “somatic” treatments is based on interview, gut feeling, and guesswork—not biology.  That doesn’t mean we can’t treat people, but we shouldn’t profess to offer a biological solution when we don’t know the nature of the problem.  We should admit our ignorance.

It would also help to allow (if not require) more time with psychiatric patients.  This is important.  If I only have 15-20 minutes with a patient, I don’t have time to ask about her persistent back pain, her intrusive brother-in-law, or her cocaine habit.  Instead, I must restrict my questions to those that pertain to the drug(s) I prescribed at the last visit.  This, of course, creates the perfect opportunity for confirmation bias—where I see what I expect to see.

We should also make an effort to educate doctors and patients alike about how little we actually know.  The subjects in trials to obtain FDA approval do NOT resemble real-world patients and are not evaluated or treated like real-world patients (and this is unlikely to change anytime soon because it works so well for the drug companies).  Patients should know this.  They should also know that the reliability of psychiatric diagnosis is poor in the first place, and that psychiatric illnesses have no established biochemical basis with which to guide treatment.

Finally, I should say that even though I call myself a psychiatrist and I prescribe drugs, I do not believe I’m taking advantage of my patients by doing so.  All of my patients are suffering, and they deserve treatment.  For some, drugs may play a key role in their care.  But when I see my entire profession move towards a biochemical approach—without any good evidence for such a strategy, and without a fair assessment of alternative explanations for behavior—and see, in my own practice, how medications provide no real benefit (or, frequently, harm) compared with other treatments, I have to wonder whether we’ve gone WAY beyond what psychopharmacology can truly offer, and whether there’s any way to put some logic back into what we call psychiatric treatment.

** There are areas in which psychopharmacology is most definitely not a “charade.”  These would include the uses of benzodiazepines, psychostimulants, and opioids like methadone and Suboxone.  With each of these agents, the expected effect is quite predictable, and they can be very effective drugs for many people.  Unfortunately, each of these can have an effect even in the absence of a diagnosis, and—probably not coincidentally—each has a potential for abuse.

Consider the following two clinical scenarios:

A.  William, a 62 year-old accountant, has been feeling “depressed” since his divorce 5 years ago.  His practice, he says, is “falling apart,” as he has lost several high-profile clients and he believes it’s “too late” for his business to recover.  His adult son and daughter admire him greatly, but his ex-wife denigrates him and does everything she can to keep their children from seeing him.  William spends most of his days at his elderly parents’ house, a two-hour drive away, where he sleeps in the room (and bed) he occupied in his childhood.

William has been seeing Dr Moore every 1-2 weeks for the last 2 years.  Dr Moore has tried to support William’s ill-fated attempts to build up his practice, spend more time with his children, and engage in more productive activities, including dating and other social endeavors.  But William persistently complains that it’s “of no use,” he’ll “never meet anyone,” and his practice is “doomed to fail.”  At times, Dr Moore has feared that William may in fact attempt suicide, although to this point no attempt has been made.

B.  Claudia is a 68 year-old Medicare recipient with a history of major depression, asthma, diabetes, peripheral neuropathy, chronic renal failure, low back pain, and—for the last year—unexplained urinary incontinence.  She sees Dr Smith approximately every four weeks.  In each visit (which typically lasts about 20 minutes), Dr Smith must manage all of Claudia’s complaints and concerns, and while Dr Smith has made referrals to the appropriate medical specialists, Claudia’s condition has not improved.  In fact, Claudia now worries that she’s a “burden” on everyone else, especially her family, and “just wants to die.”  She and her daughter ask Dr Smith to “do something” to help.

Each of these scenarios is an actual case from my practice (with details changed to maintain anonymity).  Both William and Claudia are in emotional distress, and a case could be made for a trial of a psychiatric medication in each of them.

The problem, however, lies in the fact that only one of these “doctors” is a medical doctor: in this case, Dr Smith.  As a result, despite whatever experience or insight Dr Moore may have in the diagnosis of mental illness, he’s forbidden from prescribing a drug to treat it.

I recently gave a presentation to a Continuing Education program sponsored by the California School of Professional Psychology.  My audience was a group of “prescribing psychologists”—licensed psychologists who have taken over 500 hours of psychopharmacology course work in addition to the years to obtain their psychology PhDs.  By virtue of their core training, these psychologists do not see patients as “diseases” or as targets for drugs.  Although they do receive training in psychiatric diagnosis (and use the same DSM as psychiatrists), neuroanatomy, and testing/assessment, their interventions are decidedly not biological.  Most of them see psychotherapy as a primary intervention, and, more importantly, they are well versed in determining when and how medications can be introduced as a complement to the work done in therapy.  Most states, however (including my own, California) do not permit psychologists to obtain prescribing privileges, resulting in a division of labor that ultimately affects patient care.

Let’s return to the scenarios: in scenario “A,” Dr Moore could not prescribe William any medication, although he followed William through two brief antidepressant trials prescribed by William’s primary care physician (with whom, incidentally, Dr Moore never spoke).  When Dr Moore referred William to me, I was happy to see him but didn’t want to see myself as just a “prescriber.”  Thus, I had two long phone conversations with Dr Moore to hear his assessment, and decided to prescribe one of the drugs that he recommended.  William still sees both Dr Moore and me.  It’s arguably a waste of time (and money), since each visit is followed by a telephone call to Dr Moore to make sure I’m on the right track.

Claudia’s case was a very different story.  Because Claudia complained of being a “burden” and “wanting to die”—complaints also found in major depression—Dr Smith, her primary care physician, decided to prescribe an antidepressant.  He prescribed Celexa, and about one month later, when it had had no obvious effect, he gave Claudia some samples of Abilify, an antipsychotic sometimes used for augmentation of antidepressants.  (In fact, Dr Smith told Claudia to take Abilify three times daily, with the admonishment “if you want to stop crying, you need to take this Abilify three times a day, but if you stop taking it, you’ll start crying again.”)  Like it or not, this counts as “mental health care” for lots of patients.

Some would argue that the only ones qualified to prescribe medications are medical doctors.  They would claim that Dr Moore, a psychologist, might have crossed a professional boundary by “suggesting” an antidepressant for William, while Dr Smith, a physician, has the full ability to assess interactions among medications and to manage complex polypharmacy, even without consulting a psychiatrist.  In reality, however, Dr Smith’s “training” in psychotropic drugs most likely came from a drug rep (and his use of samples was a telltale sign), not from advanced training in psychopharmacology.  When one considers that the majority of psychotropic medication is prescribed by non-psychiatrists like Dr Smith, it’s fairly safe to say that much use of psychiatric drugs is motivated by drug ads, free samples, and “educational dinners” by “key opinion leaders,” and provided without much follow-up.

Furthermore, Dr Smith’s training in mental health most likely pales in comparison to that of Dr Moore.  Psychologists like Dr Moore have five or more years of postgraduate training, 3000 or more hours of clinical supervision, research experience, and have passed a national licensing exam.  But they’re forbidden from using medications that have been FDA-approved for precisely the conditions that they are extraordinarily well-equipped to evaluate, diagnose, and treat.

A satisfactory alternative would be an integrated behavioral health/primary care clinic in which professionals like Dr Moore can consult with a psychiatrist (or another “psychiatric prescriber”) to prescribe.  This arrangement has been shown to work in many settings.  It also allows for proper follow-up and limits the number of prescribers.  Indeed, pharmaceutical companies salivate at the prospect of more people with prescribing authority—it directly expands the market for their drugs—but the fact is that most of them simply don’t work as well as advertised and cause unwanted side effects.  (More about that in a future post.)

The bottom line is that there are ways of delivering mental health care in a more rational fashion, by people who know what they’re doing. As it currently stands, however, anyone with an MD (or DO, or NP) can prescribe a drug, even if others may possess greater experience or knowledge, or provide higher-quality care.  As an MD, I’m technically licensed to perform surgery, but trust me, you don’t want me to remove your appendix.  By the same token, overworked primary care docs whose idea of treating depression is handing out Pristiq samples every few months are probably not the best ones to treat depression in the medically ill.  But they do, and maybe it’s time for that to change.

Last Saturday’s early-morning massacre at a crowded movie theater in Aurora, Colorado, stands as one of the most horrific rampages in American history.  But before anyone had any clue as to James Holmes’ motive for such a heinous act—even before the bodies had been removed from the site of the carnage—websites and social media were abuzz with suspicions that the gunman was either mentally ill, or was under the effect of psychotropic medications—one (or both) of which may have contributed to his crime.

As of this writing, any statement about Holmes’ psychiatric history is pure conjecture (although as I post this, I see a Fox News report claiming that Holmes mailed a notebook to “a psychiatrist” detailing his plan—more will surely be revealed).  Acquaintances quoted in the media have described him as a shy person, but have reported no erratic or unusual behaviors to arouse suspicion of an underlying mental illness.  Until recently, he was even enrolled in a graduate neuroscience program.  Some reports suggest that Holmes had spent weeks engineering an elaborate and complex scheme, hinting at some highly organized—albeit deadly—motive.  Nevertheless, the fact remains that we simply don’t know about any diagnosis, medication, or other psychiatric condition or treatment, which might shed light on Holmes’ frame of mind.

Those who focus on Holmes’ mental state at the time of the murders seem to fall in one of two camps.  Some argue that medications (if he were under the influence of any) may have enabled or facilitated this horrific act.  Others say that if Holmes had been diagnosed with a psychiatric illness, then this catastrophe serves as proof that we need more aggressive treatment—including medications—and services for the mentally ill.

It will be some time before we get answers in Holmes’ case.  And to me, that’s just as well.  Determining whether he has a mental illness, or was under the influence of psychotropic drugs last weekend, unfortunately reframes the question in such a way that further propagates the rift between these two factions, and fails to address how we should handle future cases like Holmes’ more humanely.  If, for example, Holmes is found to suffer from untreated schizophrenia, our society’s (and profession’s) reaction will be to lobby for more aggressive treatment, greater access to medication, more widespread screening programs, and, perhaps, a lower threshold by which to hospitalize psychotic individuals we deem as potentially “dangerous to others.”  If, on the other hand, a toxicology report reveals that Holmes had an antidepressant or antipsychotic in his bloodstream at the time of the murders, our inclination will be to restrict our use of these drugs, remove them from our formularies, and the outcries against psychopharmacology will grow ever louder.

Whether Holmes has a mental illness or not is irrelevant.  He was in crisis—probably for quite some time before last weekend—and that‘s what matters.  There was no one—and no way—to reach out to him and meet his needs in such a way to prevent this tragedy, and that, in my opinion, transcends whether he is “mentally ill” or not.

How can we fix this?  In his column in Monday’s New York Times, David Brooks almost provides a solution.  He writes, correctly, that prevention of such catastrophic events occurs through “relationships”—relatives or neighbors, for instance, who might recognize a change in someone’s behavior and encourage him to get help.  Admittedly, establishing a caring relationship with someone who suffers a history of trauma, grief over a recent loss, poor self-esteem, pathological narcissism, or acute psychosis may be difficult.  But relatives and neighbors are indeed often the first to notice questionable behavior and are well positioned to help those in need.  Perhaps in Holmes’ case, too, we’ll soon learn of some classmates or coworkers who felt something was amiss.

Brooks goes on to argue that it’s the responsibility of that neighbor or relative to “[get] that person treatment before the barbarism takes control.”  He doesn’t define what sort of “treatment” he has in mind.  But he does say that killers are “the product of psychological derangements, not sociological ones,” so the aggressive treatment options he endorses presumably include more aggressive psychological (or psychiatric) treatment.  But to expect a neighbor or relative to help an individual access treatment is precisely a sociological phenomenon.  It puts the onus on our culture at large to pay attention to how our neighbors think and act, and to offer a helping hand or a safe environment (or a locked psychiatric unit, if it has progressed that far) to those of us who think and behave differently or who are suffering a crisis.

That, unfortunately, is not what Brooks is arguing for.  (After all, the title of his essay is “More Treatment Programs.”)  If mass murderers suffer from psychological problems, which is what Brooks seems to believe, the solution “has to start with psychiatry.”  But this introduces the longstanding problem of defining that arbitrary border between “normal” and “abnormal”—a virtually impossible task.  And, of course, once we pathologize the “abnormal,” we’re then obligated to provide treatments (antipsychotic medication, involuntary hospitalization, assisted outpatient treatment, forced drugging) which, yes, might decrease the likelihood of further dangerousness, but which also compromise patients’ civil rights and do not always enable them to recover.

Brooks is right on one point.  Relationships are part of the answer.  Relationships can provide compassion and support in one’s most difficult times.  One take-home message from the Aurora tragedy should be that people like Holmes—regardless of whatever he is even “mentally ill” at all—need the security and comfort of safe, trustworthy individuals who are looking out for his (and society’s) best interests and who can intervene at a much earlier stage and in a much less aggressive way, perhaps even avoiding conventional psychiatric treatment altogether.

Getting to that point, unfortunately, requires a sea change in how we deal more compassionately with those in our society who are different from the rest of us—a change that our nation may be unwilling, or unable, to make.  If we fail to make it, we’ll be stuck with the never-ending debate over the validity of psychiatric diagnoses, the effectiveness of psychiatric drugs, the ethics of forced treatment, and the dilemma of defining when antisocial behavior becomes a “disease.”  In the meantime, troubled souls like James Holmes will continue to haunt us, left to their own devious plans until psychiatric treatment—or worse—is the only available option.