Exploiting the Placebo Effect:  Legitimate Practice or Chicanery?

On June 13, 2014, Psychiatric Times published an article by Steve Balt, MD.  The article is titled Assessing and Enhancing the Effectiveness of Antidepressants, and is a good deal more candid as to the efficacy of antidepressants than one normally encounters from psychiatry.  Dr. Balt is a private practice psychiatrist in California.  He is the editor-in-chief of The Carlat Psychiatry Report.

The article opens with the observation that despite the large number of antidepressants on the market, and decades of clinical experience, no particular product seems to have emerged as substantially better than the rest.

“As a result, most psychiatrists choose antidepressants not on the basis of efficacy, but rather on the basis of insurance coverage, adverse-effect profiles, or particular clinical features of depression (eg, melancholic, atypical, anxious features), for which some differences in efficacy do exist.”

The author refers to a number of studies, including NIMH’s STAR*D trial, and states:

“Despite the scope and initial aims of the study, no single antidepressant strategy or combination appeared more advantageous than any other.”


“Effectiveness trials, therefore, seem to confirm the conventional wisdom that no single antidepressant works better than—or worse than—any other.”


“Findings indicate that there were no significant differences in effectiveness of antidepressants, although individual drugs did differ in terms of onset of action and ease of dosing.”

In studies that did detect differences between products, Dr. Balt expresses the view that they “…were slight and of questionable clinical significance.”

Then the author makes this surprisingly candid observation:

“Because a meta-analysis is only as good as the data on which it is based, these meta-analyses must be considered in light of the very real problem of selective publication. This is the tendency for favorable results to be published, while negative or neutral results are not. In an analysis of 74 antidepressant trials registered with the FDA between 1987 and 2004, Turner and colleagues found that nearly half (36, or 48.6%) were negative, and the vast majority of these were either not published or were published in a way that made the drug seem favorable. Likewise, industry-sponsored studies are more likely to favor the manufacturer’s drug, often because of nuances in experimental design. While most researchers make every effort to include unpublished results in their meta-analyses, the “file-drawer” phenomenon of unpublished negative results may bias the conclusions of analyses that exclude the inaccessible data.”

Dr. Balt summarizes all of this with what he calls:  two “stark truths about antidepressants.”

“First, there seem to be no significant differences among them; although future research may uncover patient-specific biomarkers that favor one medication over another, none has yet done so. Second, and somewhat surprisingly, antidepressant effectiveness is quite low.”

And he concludes:

“Thus, in the absence of data that can predict the best antidepressant regimen for a patient, enhancing the effectiveness of an antidepressant seems to be the best strategy.”

Dr. Balt mentions two possible strategies for enhancing the effectiveness of an antidepressant.  The first is to recognize that depression is not a single condition, but rather  “…multiple conditions, each deserving of its own unique treatment approach.”  He states that  “…much of the antidepressant response in mild to moderate depression may be due to the placebo effect”, and that anxious individuals may do better with psychotherapy.

Dr. Balt’s second suggestion for effectiveness-enhancing might be to combine antidepressants, or to supplement them, with what he calls “augmentation agents”.  But he acknowledges that while there might be some “intuitive appeal” to these strategies, there is little in the way of supporting evidence for their effectiveness.


“The more important question may be more about whom we are treating rather than what we treat with.”

Which, of course, is close to what we, on this side of the debate, have been saying all along – that effective help must be tailor-made to the personal characteristics and context of the individual, and that the facile labels which psychiatrists optimistically and misleadingly call “diagnoses” are counterproductive.

For the sake of thoroughness, Dr. Balt dismisses the relevance of genetics:

“In reality, however, the genetic contribution likely involves an impractically large number of variants, each having a very small effect, that together contribute to the very complex phenotype of antidepressant response. Indeed, 2 meta-analyses, using genome-wide analysis to identify polymorphisms to predict treatment response, found only a 1.2% contribution or no contribution at all.”

And metabolism:

“With few possible exceptions, no evidence exists that blood concentrations influence antidepressant outcomes, and there are multiple nongenetic factors that influence drug metabolism, such as diet, other medications, and adherence.”

This last assertion – that antidepressant blood concentrations have no effect on outcome – is a truly extraordinary admission from a psychiatrist, and could even be interpreted as the (psychiatric) baby going out with the bathwater.

But then, just as Dr. Balt appears to be morphing into a psychotherapist, he slips back into psychiatric mode:

“Not surprisingly, we can take advantage of patient preferences to enhance treatment outcomes.” [From the context, it is clear that he means drug treatment]

“…the quality of the therapeutic alliance between prescriber and patient is sometimes a better predictor of patient outcome than which drugs are prescribed.”

The last quote sounds positive, but watch where Dr. Balt goes with it:

“One study found that ‘effective’ prescribers obtained better outcomes with placebos than ‘less effective’ prescribers with active antidepressants.  Asking ‘which’ medication may be less important than the ‘meaning’ of medication to both clinician and patient.”

So what Dr. Balt is describing and suggesting is to use the therapeutic alliance, not to connect to the individual, and to provide real support in the resolution of personal concerns and problems.  But rather, to enhance the effectiveness of psychiatric drugs.

“The recent emphasis on personalized antidepressant prescribing seems warranted, but rather than taking a combination or pharmacogenomic approach to medication selection, clinicians should focus more on a personalized approach, establish realistic (but hopeful) expectations, and use patient preferences and beliefs to optimize outcomes.” [Emphasis added]


“Even when patients’ preferences do not have any bearing on outcome, matching treatments with patients’ preferences increases their willingness to initiate and adhere to a treatment plan.” [Emphasis added]

So if a client believes that Zoloft, say, will be helpful, because it seemed to help a family member, perhaps, or even because the TV ad made the product seem attractive, the psychiatrist should prescribe Zoloft?  That feels a little bit like a con.  It’s an admission that although the drugs aren’t much good, we can sort of trick the client into enrolling, adhering, and perhaps even becoming happier.

And why is “adherence” so important, if the drugs aren’t much use to begin with, and often bring a host of additional problems?

I can appreciate Dr. Balt’s honesty concerning antidepressants, but, in my view, his recommendations entail a good measure of condescension and disrespect.  They also have the potential to encourage increased prescribing of drugs, despite his assertions of their limited efficacy.

. . . . . . . . . . . . . . . .

And incidentally, besides his recent critique of antidepressants, Dr. Balt had earlier gone on the record with some biting and revealing criticism of a psychiatric clinic where he worked from about 2009 to 2012.

In August 2014, Dr. Balt was taken to task on Twitter for being a top prescriber of Seroquel (quetiapine), a neuroleptic, in California in 2009 (1,000+ prescriptions).  In a subsequent interview with William Heisel of William Heisel’s Antidote, Dr. Balt is quoted as writing:

“More than 1,000 new prescriptions in one year is hard to fathom. Given my part-time schedule there, it would have meant six to seven new prescriptions each day, which is impossible. However, if it includes each month’s refills for patients who took the drug all year long, that does not seem at all unreasonable for that clinic. Given the volume of patients (and their specific requests for Seroquel, a medication in high demand), automatic refills were frequently given.”


“That was my first full year at a clinic where quick visits, busy schedules, and knee-jerk prescribing were the order of the day. Over time, I became quite disappointed — if not angry — at what seemed to pass for ‘treatment’ and the overwhelming emphasis (on the part of patients and doctors alike) on the use of psychotropics to manage what were, most of the time, sociocultural stressors in our largely indigent, population.”


“I’d estimate that 50% to 70% were for ‘schizophrenia.’  Keep in mind, however, that a schizophrenia diagnosis is freely given in these settings, e.g., for otherwise healthy people who say, ‘I talk to myself sometimes,’ or ‘I sometimes see shadows at night,’ or ‘I’m always looking behind my back when I’m outside,’ etc.  As for the rest, those prescriptions were probably given for bipolar disorder or off-label for sleep.  As I learned more about the metabolic effects of Seroquel I tried to use less of it.”


“I have worked at other clinics, and my experience tells me that overprescribing is rampant in the Medicaid setting and apparently accepted as the standard of care. I have seen the same over-reliance on psychotropics, over-diagnosis and waste of resources in other parts of the state, although the volumes were not nearly as high as what I experienced in Oakland.”

So Dr. Balt is describing a psychiatric practice in which:

  • “automatic” refills for a neuroleptic were frequently given
  • “knee-jerk prescribing” was the order of the day
  • psychosocial stressors were “managed” with psychotropic drugs
  • diagnoses of schizophrenia were “freely given”
  • and overprescribing for Medicaid recipients (i.e. poor people) was “rampant”

Dr. Balt is now in private practice, and on his Psychology Today bio page he tells us that he offers “both medication management and psychotherapy” and that his “…approach to medications is a conservative one.”

And then he writes:  “… I am willing to help patients taper or discontinue medications that provide little benefit, or which cause harm.”  I have no reason to doubt Dr. Balt’s sincerity in this matter, but the fact that he felt the need to issue such a statement implies, I think, that it is not standard practice in psychiatry.  In other words:  psychiatrists routinely leave people on  “…medications that provide little benefit, or which cause harm.”  This, of course, is something that we on this side of the debate have been saying all along, but it is interesting to see these ideas coming from within psychiatry itself.

  • Both Steve Balt’s original article and Phil Hickey’s critique provide very interesting insights on the current crisis of professional psychiatry and psychology. One related question for Phil, if I may: I frequently read commentary from senior psychiatrists who proclaim at length and with fervor that psychotropic meds DO work for the more severely depressed, and that without them we would see even more suicides. The commentary is grounded on their own observations in decades of private or hospital practice. So how should such comments be seen and evaluated? Are the proponents for meds lying, self-deluded, financially self-interested, or genuinely convinced but mistaken? Or all four?

    I follow this blog closely on Twitter and respect what Phil has to say. I also support thousands of chronic face pain patients around the world and have seen the damage done to these people by both mainstream psychiatry and the mythology called “psychosomatic” medicine. But I am frustrated and confounded. How do we get past the claims and counter-claims for various methods in the healing arts, to reach some standard of practice — and restraint — which deals effectively and reliably with the real human distress which brings people into dialog with healers in the first place?

  • Something I’ve never understood:

    In the UK, many prescription only medicines (POM) come packaged with a patient information sheet setting out what the medication is made of, what it is prescribed for, contraindications, potential side effects etc.

    If, in the U.S., a placebo has been prescribed, what information does the patient receive with the medication, and if they are given no information about what they have been prescribed, either from the prescribing clinician or in the packet, at the point at which it is supplied to the patient, on what basis can the patient be said to be in a position to give informed consent?

    Article in Guardian:

    Half of all German doctors prescribe placebos, new study shows:


  • Phil_Hickey


    Thanks for coming in.

    You’re correct in your observation that psychiatrists, although retreating from their earlier enthusiasm for antidepressants for virtually all ranges of depression, continue to assert that these products are highly effective for severe depression. But the assertion doesn’t bear up under scrutiny.

    In 2008, Irving Kirsch et al published a meta-analysis on PLOS Med. They reviewed all clinical trials submitted to the FDA on four new-generation antidepressants for which full datasets were available.

    Virtually all drug trials are of the drug-placebo comparison type, and a drug is considered effective if it is significantly better than the placebo. But what’s often ignored is that there are two ways that such a difference can occur: the drug may actually be significantly better or the placebo may be worse. Here’s what Kirsch et al found:

    “Drug–placebo differences in antidepressant efficacy increase as a function of baseline severity, but are relatively small even for severely depressed patients. The relationship between initial severity and antidepressant efficacy is attributable to decreased responsiveness to placebo among very severely depressed patients, rather than to increased responsiveness to medication.”

    To the best of my knowledge, this finding has never been overturned.

    A second concern with antidepressants, in both mild and severe depression, is what El-Mallakh et al call tardive dysphoria. The finding here is that people who take antidepressants for a prolonged period become almost incapable of experiencing joy. I wrote a post on this here.

    Psychiatry’s claim that without their antidepressants, there would be a great increase in suicides is very difficult to evaluate, primarily because of the infrequency of suicide. Most drug trials have no suicides among participants. There is a concise and balanced discussion of this issue in Joanna Moncrieff’s book The Myth of the Chemical Cure (SSRI’s, Suicide and Violence, pp 167-170)

    You write:

    “How do we get past the claims and counter-claims for various methods in the healing arts, to reach some standard of practice — and restraint — which deals effectively and reliably with the real human distress which brings people into dialog with healers in the first place?”

    And this, of course, is the central question. My own answer is simple:

    1. Real dialogue with the individual to identify the sources of the distress

    2. Continued dialogue to identify ways to resolve these issues.

    3. Marshalling of all available resources to help the individual get to a better place

    These steps can be taken by professional helpers, or by “natural” helpers (family, extended family, friends, co-workers, etc.).

  • May I have your permission to quote from this response (with by name attribution) on my Facebook page?

  • Phil_Hickey


    Of course.

  • If I may add just one more question, Phil: how can professional associations be effectively lobbied to change standards of practice that are taught to those first entering the field? It seems to me that organizations like the APA and NIH are among the principal roadblocks to progress, by their refusal to embrace changes that would support insurance reimbursement for the kinds of “steps” you advocate.

  • cledwyn blubs

    “… people on antidepressants become almost incapable of experiencing joy.”

    Ha, don’t I know it! I mean, look at me, I must be the most miserable b@stard alive! I’ve gotten so used to the dysphoria from venlafaxine I forget it’s the drugs that cause it.

  • Phil_Hickey


    The insurance reimbursement issue is a quagmire. Medical insurance only reimburses for treatment of a medical condition. And this is legitimate and reasonable. I think the key to the solution is to see more of the kinds of problems that currently go to mental health being transferred instead to departments of social services, and of course, to social workers in private practice.

    I think that lobbying psychologists is likely to be difficult, because so many have joined the psychiatry camp, and are even pushing for prescribing authority! Social workers and counselors, on the other hand, are, in my experience, sick and tired of the psychiatrization of virtually every human problem, and are, I think, more open to a paradigm shift. How to lobby? Well, I think that’s what we’re trying to do. It’s an uphill battle, but we are gaining supporters every day.