Psychotherapy and Pharmacotherapy: A Dangerous Collaboration?

By: Garry Cooper, LCSW

 

Psychologists Jacqueline Sparks, Barry Duncan, and Scott Miller have been among the most vocal critics of the influence of the pharmaceutical industry on psychotherapy. The millions of dollars which the industry spends influences research and public perceptions of what constitutes treatment and even of what constitutes psychological disorders themselves. Writing in the summer/fall, 2006 issue of the Journal of Family Psychotherapy, Sparks, Duncan, and Miller examine how the influence of the pharmaceutical industry has eroded and corrupted the very base of what psychotherapy is all about.

In their article, “Integrating Psychotherapy and Pharmacotherapy: Myths and the Missing Link,” they paint a grim picture of diagnosis and treatment that’s moving increasingly toward an almost complete dependence upon brain imaging, therapy manualization, and conception of every mental disorder as fundamentally biological. “Replacing the uniquely interpersonal phenomena of the therapy relationship with the medical model–diagnosis and pharmacological treatment–is not why most therapists decided to enter the field,” they say. Integrating pharmacotherapy and psychotherapy, they write, is the wrong direction, and is based on several myths about mental disorders and treatment.

The first myth of the medical model concerns diagnosis which always starts by looking at what’s wrong. The medical model equation, they write, is diagnosis + prescriptive treatment = symptom amelioration. But diagnosis in this model is seriously flawed. For starters, it’s much easier to define normal and abnormal in terms of physiology. There’s a normal range for blood glucose levels or muscle strength. Carcinogenic cells are unmistakably detectable and mean cancer. But when you talk about psychological conditions, normal is a fuzzy, slippery, subjective concept that’s inextricably tied in with the values of others, with social and cultural norms, and, as people like R.D. Laing and Thomas Szasz have long insisted, with “arrangements of power, hierarchy, inclusion, and exclusion.” In short, “Human behavior exhibits a significant range of variation,” write Sparks, Duncan and Miller.

The entire DSM-based system of diagnosis, modeled after the medical model, they write, lacks reliability and validity. Quoting other critics of the DSM, they contend that the disorders “lack empirical standards to distinguish the hypothesized pathological states from normal human variation to the problems of life.” Even one of the architects of the original DSM, Robert Spitzer, has admitted that the DSM has a long way to go in solving the problem of reliability. The result of all this, say Sparks, Duncan and Miller, “is a set of murky over-inclusive criteria for an ever-growing list of disorders.”

The symptom checklists of the DSM completely exclude the social constructionist perspective as well. “Humans speak and respond to each other in relationship,” they write. “This process includes ongoing accommodation between communicators and complex turns of conversation structured by context, social power, and participant’s [sic] goals and interests. The mental health diagnostic interview and diagnostic assignment are social scripts, imbued with particular roles, rules, and alignments of power deeply engrained in Western culture….Humans are first and foremost members of social communities, and their behaviors and states of mind are fundamentally connected to and influenced by these contexts.”

The second myth is biological causality. With depression, for example, the medical model and the pharmaceutical industry have just about cemented the view that depression is “caused” by a serotonin deficiency. Yet many individuals with low serotonin levels are not depressed, and increasing serotonin levels in many depressed people does not alleviate depression. The first flaw undermining this idea of biological causality is a common one known to all researchers: confusing an association with causality. In point of fact, the interplay between experience, biology and cognition is a complex dance, and it’s impossible to identify in many, if not most, cases which leads at which times.

Allied to this myth is the belief that if symptoms are alleviated after taking a particular medication, then that proves that the diagnosis was correct and that the med “worked.” But this of course does not explain why placebos work about a third of the time on many psychological conditions. This belief is so widespread that several years ago the American Academy of Pediatrics warned about diagnosing ADHD without a full psychosocial assessment and input from a wide variety of people in the child’s life. Too many physicians had been too quickly diagnosing ADHD, prescribing Ritalin and, when the symptoms abated, used that as “proof” that the original diagnosis had been correct.

Sparks, Duncan and Miller point out that antidepressants do not work on at least half of the participants in clinical trials and that placebos are about as effective as antidepressants. Using the flawed inferential reasoning of biological causality, they say, one could conclude from the clinical trials that depression is caused by both low serotonin and low sugar levels. “Despite 50 years of Herculean efforts,” they write, “the invention of electron microscopy, the advent of radiolabeling techniques, the revolution of molecular biology, and the merger of computers with neuroimaging machines, no reliable biological marker has ever emerged as the definitive cause of any psychiatric ‘disease’.”

The third myth is that pills work really well. It’s perhaps an inevitable misperception: if you believe that mental disorders have fundamental biological causes, then you have to believe that pills work and that an even better pill is just over the horizon. If this is true, ask Sparks, Duncan and Miller, why, despite the unprecedented (and, many say, alarming) rise in antidepressant use have the depression rates remained stable? They cite the Kirsch and Saperstein study of antidepressants which concluded that 75 percent of the positive response was duplicated by placebo and the remaining 25 percent might be due to the unblinding power of the side-effects. (Studies find that “blinded” conditions in studies are rarely truly blinded; participants and raters guess at a better than chance rate who’s been given placebo or control treatment and who’s been give the treatment actually being studied). “Antidepressants, because of their side effects, might best beconsidered as a last resort and not as a matter of course to address biologicalunderpinnings,” say Sparks, Duncan and Miller.

Much of this third myth, they say, comes not from mistaken assumptions but studies purposely designed to yield the results desired by the pharmaceutical companies designing the clinical trials. Such design flaws include short-time intervals, which initially favor medications. If a medication has an immediate effect within, say, 12 weeks, that levels off or even reverses, and a psychotherapy takes longer but eventually shows better and/or longer lasting results that endure after treatment ceases, the study will be designed for 12 weeks. Studies also rely on clinician-rated measures instead of subjects’ self-report measures, and these invariably favor medications. A patient may feel it’s a poor tradeoff of a few points on a clinician-rated depression scale for decreased libido or any of the dozens of other side-effects, but when the patient has no say in how much she has “improved,” those few points may be ultimately meaningless in the study’s reported results. Another common design flaw, say Sparks, Duncan and Miller: “Many studies use placebo run-ins, eliminating placebo responders before randomization and unevenly skewing group membership.”

Sparks, Duncan and Miller urge anyone who reads about studies to follow the money trail, especially because a spurious study, once it is published, then becomes replicated, appearing as a citation to buttress other spurious studies. It gets picked up and popularized by the media. Physicians and therapists are the same as most people when it comes to evaluating the effectiveness of medications: they’re very busy and don’t critically, carefully read the original studies, relying instead on abstracts, media reports, press releases, and the repetition factor: if you hear something often enough, it must be true.

As examples, they cite the Texas Medication Algorithm Project (TMAP), the Sequenced Treatment Alternatives for Relieving Depression (STAR-D), and UCLA’s Targeted Treatment for Depression Program (TTD), which support antidepressants and are all very much in the news and on the radar screens of therapists and psychiatrists. Sparks, Miller and Duncan cite one favorable reviewer’s assertion that because they are funded by the Texas Department of Mental Health, National Institute of Mental Health (NIMH), and a university, respectively, these studies “may more accurately reflect realistic outcome data, non-influenced by the profit motives of drug companies.” Not so, they say, pointing out that pharmaceutical companies invested several hundred thousand dollars in TMAP, and the NIMH intricate ties to the pharmaceutical industry include pharmaceutical company financial grants to its employees and revolving employment doors. Sparks, Duncan and Miller tracked down a researcher cited by the reviewer in his favorable assessment of the TTD, who turned out to be unaffiliated with UCLA. They finally found him through Google: apparently unaffiliated with any university, he has a website that’s “provided as an educational resource by AstraZeneca.” Although the website insists that the pharmaceutical company exercises no control over the content, Sparks, Duncan and Miller aren’t reassured.

All told, they say, the evidence for medications’ effectiveness for mental disorders is “an empirical house of cards…driven by corporations that have no limits in their ability to spread their influence.”

In their article, the three strongly dispute the “myth” that the combination of meds and therapy for depression is more effective than either alone. First, they insist, the empirical support for that is both tainted by the influence of the pharmaceutical industry upon research and, even given that influence, the empirical support is still weak. The primary studies that assertion rests upon have built-in flaws. One, a “mega-analysis” of close to 600 people enrolled in 10 different treatment protocols over a 10-year period, found no advantage for the combination treatment for people suffering with mild to moderate depression. Although it did find an advantage for those suffering with major, recurrent depression, it used a 12-week trial period, which favored the medication and only one clinician-rated outcome measure. As for the researchers, Sparks, Duncan and Miller quote from the journal, which covered their ties to the pharmaceutical industry by saying, “It would have used too much space to disclose them fully in the Journal.” Another, the massive Treatment of Adolescents Depression Study (TADS), which found that the combination of fluoxetine and psychotherapy was more effective than either alone, had such flaws in the blinding and two of the treatment arms that the researchers themselves acknowledged that the “‘active ingredient’ in improvement cannot be specified.” Given the side effects of the antidepressants–which, Sparks, Duncan and Miller insist, are often downplayed in the published studies-psychotherapy should always be the first treatment of choice. (This is the same conclusion of Britain’s National Institute of Clinical Excellence).

Sparks, Duncan and Miller also object to the role therapists are often relegated to in combination treatment, which is a little like trying to develop a partnership with a hungry lion. “We value collaboration with other professionals, and often find it useful when clients are fully in accord with this type of assistance,” they write. “However, our experience has been that the role of therapy is often relegated to one that supports the medical intervention….The ‘collaborating’ therapist is often seen as the in vivo arm of the busy physician, smoothing over the rough edges as the medication takes effect,educating clients or family members about the disorder, and making sure that medication compliance is a treatment goal.”

By putting collaboration in quotes, Sparks, Duncan and Miller may consciously be calling up negative connotations of collaboration that evoke the collaborators of World War II. In the end, they imply, there can be no genuine collaboration when one “partner” holds so much power. “Even when presented as ‘both/and,’ ‘integration,’ ‘collaboration,’ or that appealing term, ‘biopsychosocial,'” they write, “clinical practice parameters tend to be based on the dominant biological paradigm. Biopsychosocial becomes Biopsychosocial, with interpersonal and social aspects bringing up the rear.”

How does this play out in depression? Let’s assume a child meets the DSM criteria for major depression. He is irritable or appears tearful most of the time, has markedly diminished interest or pleasure in most activities, suffers from insomnia, fatigue, guilt, and has difficulty concentrating.

* Adapted from “Integrating Psychotherapy and Pharmacotherapy: Myths and the Missing Link,” in Journal of Family Psychotherapy, Vol. 17(3/4) 2006, “Copyright 2006, Haworth Press. The complete article by Sparks, Duncan and Miller may be seen online at https://scottdmiller.com/uploadedFiles/MissingLinkSparks.pdf