Like other “advances” in community treatment and diagnosis that we deconstruct in this book, the promises of the scientific drug treatment of distress, madness, and misbehavior rest on uncritically accepted claims, not well-tested evidence. The popular success of prescribed psychoactive drugs does not rest on scientific demonstrations of efficacy in the treatment of medical disorders. The success results from the synergistic combination of ideological, cultural, and commercial forces. Bearing down from above came shoddy science, uncritical practitioners with a monopoly on drug prescriptions, the minting of new mental disorders, ubiquitous drug marketing, co-opted regulatory agencies, and plain fraud. Surging up from below and augmented by a newly emerging health consumerism came people’s indefatigable desires for psychoactive drugs, the powerful but dismissed power of placebo, and the reverberation of myriad personal narratives of drug-induced transformation, recovery, and cure.
The merging of these forces created the conviction among the educated and not-so-educated that such popular success could only rest on a solid scientific foundation. It was and remains inconceivable to many observers that consumers and insurers might spend billions of dollars on placebos with brand names—or worse, on drugs that produce waves of iatrogenic (that is, doctor- or treatment-induced) diseases. Indeed, over the past half century the psychoactive medicines enterprise was constantly justified by appeals to science, but its foundations did not rest on science as a search for truth, as an effort to test hypotheses rigorously, or as a means to understand biological or social facts. As for the specific field directly concerned with the investigation of psychiatric drug treatment, clinical psychopharmacology, many of its practitioners and researchers engaged in nothing less than the studious avoidance of evident facts about distress, psychoactive drugs, healing, and the power of suggestion.
In this chapter, we lay out the mental health system’s current crisis with regard to psychiatric drugs: the emerging consensus that the latest prescribed psychiatric drugs do not necessarily improve the targeted symptoms of people diagnosed with major mental disorders. We review recent studies conducted with United States federal government funding that attempted to assess what sort of improvement the major modern classes of psychiatric drugs were actually bringing about in such people. We then place their findings in the context of the long-standing promotion of drugs by the pharmaceutical industry, and psychiatry’s enthusiastic role as the industry’s junior partner. Yet as our initial comments in this chapter suggest, the story of the widespread adoption of apparently ineffective drugs as paradigmatic treatments is complex and involves many other interacting forces and players. It is also not a new story in psychiatry. We try to paint this more complex yet recurring story in chapter 7, which also looks at the dubious distinctions made between “dangerous drugs” and “approved medications” throughout the century and examines how some particular psychoactive substances become constructed and promoted as FDA-approved medications.
We present two complementary but different critiques of the current disease-centered approach to conceiving of the beneficial effects of prescribed psychiatric drugs. The first, in this chapter, focuses on the continuously documented but continuously avoided power of placebo. The second, focusing on the striking and subtle psychoactive effects of psychiatric drugs, suggests that placing these drugs squarely within the broader ecological niche of all psychoactive drugs could go far to increase our understanding of how drugs “work” to relieve distress while creating new opportunities for a truly patient-centered psychopharmacology (Moncrieff & Cohen, 2009; Moncrieff, 2008).
That psychoactive drugs used as prescribed medicines often do not have the beneficial impact on mental disorders (however defined) claimed by proponents and promoters of drugs is relatively easy to demonstrate: the most touted studies show that the desired outcomes of modern drug treatment do not reveal improvement when compared with outcomes of treatments from the pre-drug era, or with outcomes from older or discarded drugs, and in some cases the recent outcomes appear worse. For example, when followed for one year, depressed persons who are drug-treated experience more depressive episodes, and longer ones, than depressed persons who remain drug-free (Patten, 2004). Very long-term follow-up studies of ten years’ duration or more of depressed people show no worse outcomes for drug-free than drug-treated individuals (Hughes & Cohen, 2009). In the largest real-world study of the drug treatment of depression funded by the NIMH, it appears that not more than 3 percent of the more than four thousand subjects who underwent drug treatment managed to stay in the study, improve, and be considered to stay well for one year (Nierenberg et al., 2008). Improvement rates in drug treatment studies of schizophrenic patients during the late 1980s to the mid-1990s were identical to rates observed in studies conducted between sixty and one hundred years earlier (Hegarty et al., 1994). When British psychiatrist David Healy (2009) investigated changes over the past fifty years around his own hospital in North Wales, he observed the following: compulsory psychiatric detentions were up 300 percent, admissions for serious mental illness were up 700 percent, while admissions overall were up 1500 percent. Suicide rates among patients diagnosed with schizophrenia were up 2000 percent. General death rate among patients considered seriously mentally ill had “risen substantially.” Healy added that all studies find that the death rate of schizophrenic patients increases with the number of drugs given. He concluded: “What we are seeing now is not what happens when treatments work” (2009, pp. 32–33).
Such assessments have not only come from mental health professionals known for their critical views. Thomas Insel, director of the NIMH, acknowledged in 2009 the failure of the pharmacological agents that have dominated mental health treatment since the 1960s. Insel asked: “Is there any evidence to support that progress in psychiatry leads to reductions in morbidity or disability in people with serious mental illness?” To answer, he reviewed recent US national studies involving interviews with participants in several thousand households. He made the dramatic assertion that “there is no evidence for reduced morbidity or mortality from any mental illness” (2009, p. 129). “The results demonstrate no change in the prevalence of mental illness between 1992 and 2002, but increased rates of treatment. . . . Curiously, despite increased treatment, there was no evidence for decreased disability” (pp. 128–129). Concerning mortality, Insel noted that persons diagnosed with serious mental illness die from thirteen to thirty-two years earlier than people in the general population, and that suicide rates have not decreased.
In this respect, the official realization of disappointing effects from widely promoted drugs fits what French historians Christian Bachman and Anne Coppel have called the “path of stigmatization” (1989, p. 109), which they believe characterizes how psychoactive drugs move over time in society. More pointedly, Dutch historians Toine Pieters and Stephen Snelders (2005) have named this a “Seige cycle,” after German psychiatrist Max Seige, who suggested as early as 1912 the idea of a cyclical dynamic of the careers of such drugs. A Seige cycle is characterized by initial enthusiasm and therapeutic optimism leading to widespread use, subsequent negative appraisal, and finally, limited use or an outright ban. Seige’s insight is notable, given the few drugs he must have considered as examples by this date. The opiates, morphine especially, exemplify this path: at the start, a “uniquely efficacious” drug; at the end, “addicted victims”1 (Chast, 1995).
Thomas Szasz (1974) as well as drug scholar Richard DeGrandpre (2007) have also commented on the trajectories of drugs from panaceas to panapathogens (but sometimes also back to panaceas as is the case with the medical use of morphine today and the renewed psychiatric interest in hallucinogens, see Tierney, 2010), despite an identical scientific understanding of their pharmacology at the start as at the end of the trajectory. These authors argue that the social-ceremonial aims of psychoactive drugs may far outweigh—whether in personal lives, professional interest, or public policy spheres—their actual pharmacological effects. We find these insights intriguing and will be applying them throughout our analysis of psychiatry’s portrayals of the prescription of its favorite psychoactive drugs as evidence-based treatment.
The preceding paragraphs suggest that there is nothing unusual about the inability to show, in scientific studies conducted after psychoactive drugs have been in widespread use for years and decades, that these drugs do not necessarily decrease disability—that in fact they may be responsible for excess disability. This pattern may have occurred with every class of drugs introduced in psychiatric practice since the turn of the twentieth century. Disappointing findings of effectiveness and findings of excess (iatrogenic) morbidity have also regularly been used to justify and promote the adoption of newer drugs or drug classes, on the promise of improved effectiveness and reduced toxicity. Being unfamiliar with history or being too closely wedded to a particular school of thought, or both, many people within and outside of the mental health system may therefore have confused the popularity of psychiatric drugs with the validity of the claim that psychiatric drugs constitute specific, effective treatments for specific disorders. In our view, this latter belief is a major impediment to understanding any genuine treatment functions and accomplishments of psychiatry today, and we deconstruct it in this and in the following chapter.
Starting roughly in the 1920s with the introduction of the sedative barbiturates into routine medical practice, crops of new-and-improved psychiatric drugs have been prescribed to an increasing number of people. The stimulating amphetamines were introduced in the 1940s. In the 1950s came the neuroleptics (or antipsychotics) and the tricyclic antidepressants (TCAs), launching a new confidence in biological psychiatry. In the 1960s the tranquilizing benzodiazepines were introduced, making large new inroads into general medical practice. In the 1980s, the serotonin-reuptake inhibitors (SSRIs), initially approved as antidepressants, extended biological psychiatry’s conquest-by-medication and in the 1990s, the second-generation antipsychotics and anticonvulsant “mood stabilizers” mopped up virtually all remaining pockets of resistance. Insel’s appraisal of progress, or the lack thereof, was published in 2009.
Prescriptions written for antidepressants in the United States numbered about 40 million in 1988, the year Prozac was launched, compared to 264 million in 2011 (IMS Health, 2012). Five years earlier, antidepressants had become the most prescribed class of drugs in the United States (Centers for Disease Control and Prevention, 2006)—more often prescribed than antibiotics or drugs to treat heart disease. Looking only at the short interval between 1998 and 2004, the number of noninstitutionalized Americans purchasing products from the psychiatric drug classes of antidepressants, antipsychotics, benzodiazepines, and stimulants increased from 22 million to 38 million (Stagnitti, 2007). The really notable changes occurred among children. In the mid-1980s, less than 1 percent of children were taking any psychiatric drugs, usually stimulants and benzodiazepines. Twenty years later, depending on the specific age group considered, 8–12 percent of youths were being medicated with psychiatric drugs. Indeed, in 2009, over 40 million prescriptions or refills were written for youths under nineteen years of age: 24.3 million for stimulants, 9.6 million for antidepressants, and 6.5 million for antipsychotics (Mathews, 2010).
Only recently have academics and journalists described the role of the pharmaceutical industry in these trends without being dismissed as antiestablishment radicals with a grudge against profitable industries. The industry did not merely bring new drugs to market, it made every effort to ensure that physicians would view whatever latest drug or drug class was being promoted as indispensably beneficial and would prescribe them readily for the widest range of patients’ complaints and presenting problems. This involved “disease mongering”—broadening the boundaries of treatable problems and creating anxieties about the seriousness of ordinary ailments (Moynihan, Heath, & Henry, 2002)—as well as purchasing the hearts and minds of practitioners and those leaders and organizations who could influence the ideas and behavior of practitioners. Psychiatry, which already embodied the medicalization of deviance and distress, wholeheartedly embraced its partnership with the drug industry, to the point, we think, of becoming virtually indistinguishable from it.
During the first decades of the twentieth century, in order to present their products as more scientifically-based, drug manufacturers began to voluntarily relinquish the option to advertise drugs directly to the public. This made direct marketing to physicians indispensable, and the appearance of science in marketing grew correspondingly. Amphetamine, first developed in the 1920s as an oral anti-asthma drug, was actively promoted the next decade by the Smith, Kline and French firm among “neuropsychiatrists” for the treatment of mild depression, from whence it spread to general medicine. “By the 1930s,” writes historian Nicholas Rasmussen (2006), “leading firms like Smith Kline and French had learned to involve themselves in the medical research literature to a remarkable extent, from conceiving and designing clinical studies that targeted particular medical audiences, to overseeing their execution by competent or even eminent researchers, to managing their writing, illustration, and publication in top journals” (p. 301). Thus began the “remarkable extent” to which the pharmaceutical industry is involved in the design, conduct, and dissemination of clinical trial research today.
Over the century, the drug industry has developed and employed systematic drug-detailing techniques to convince individual doctors to prescribe its latest products. The basic techniques—gifts, flattery, and persuasion—have not changed much since the 1930s. Today, however, drug manufacturers can target virtually every practicing physician in the United States with laser-like precision and maximum efficiency. Using computerized databases purchased from a plethora of commercial medical information firms and from professional organizations such as the American Medical Association, drug firms closely monitor the effects of their promotional strategies on drug sales or prescriptions at national, regional, and even local pharmacy levels. It is no exaggeration that an entire multinational pharmaceutical firm can adapt or customize its marketing strategies down to a single prescriber if necessary (Fugh-Berman, 2008).
To support these techniques, the industry proceeds by mass infusion of money, dozens of billions of dollars each year (Gagnon & Lexchin, 2008). The money facilitates the professional activities (studies, travels, hostings, communications, disseminations) of students, clinicians, educators, researchers, legislators, policymakers, drug regulators, journal editors and reviewers, hospital administrators, and consumer advocates—as long as the endpoint of the activity is perceived as favorable to the industry’s products or completely neutral (by which we do not mean objective). The money comes in myriad forms, such as salaries, prizes, awards, donations, reimbursements, unrestricted educational funds, research grants, endowments, equipment grants, seed grants, travel allowances, and bribes. This fuel keeps the drug industry’s views, interests, and agenda active or quiescent in the mind of every player and potential player in the mental health system. (A more detailed analysis of drug-industry funding appears in the next chapter.) Thus, by the early 1990s, the multinational pharmaceutical industry had cemented an intricately complex, multilevel partnership with the profession of psychiatry, both at the level of psychiatric institutions and educational centers and at the level of individual practitioners, and with national, state, and local opinion leaders in the mental health movement. With every passing year, as more patented drugs sold, more profits accrued to the industry, and more funds were plowed back into making friends, influencing people, and generating “evidence,” the partnership grew in breadth and depth.
As discussed especially in chapter 5, starting sometime after 2005, signs of a backlash against this alliance appeared. Rumblings were heard from some newspapers, some legislators, and some outspoken members of various professional groups. In 2008, for the first time, the crusading US Senator Charles Grassley asked the American Psychiatric Association to disclose precisely how much of its annual budget came from the drug industry. The APA (2008) divulged that 28 percent (approximately $14 million of its budget the previous year) were industry contributions, primarily paid in advertising in APA journals and funds for continuing medical education (CME) courses offered by the APA.
As public acceptance of industry control begins to decline, psychiatric associations are beginning to clean house, for example by reducing the number of industry-funded symposia offered as CME at their annual meetings (Carlat, 2009). For the foreseeable future, however, if psychiatry wishes to maintain itself as the flagship mental health profession, it has little choice but to maintain its financial and intellectual dependence on the pharmaceutical industry. The APA’s ability to attract substantial industry funding outside of membership dues to promote its image and its activities has been unique among all the mental health professions. This is not difficult to understand. In the absence of scientific demonstrations of the validity of its particular medical approach, all psychiatry can unequivocally show is the popularity of psychoactive drugs that physicians prescribe after opinion leaders in psychiatry show them the way. But the popularity of drugs is a double-edged sword. Though it has propelled biological psychiatry to undreamed heights, a number of recent trends suggest that it may eventually engulf the profession—render psychiatry obsolete, in other words—in a rising wave of industry-promoted and consumer-directed medication choices.
With the allowance by the Food and Drug Administration (FDA) of direct-to-consumer advertising of prescription drugs in the United States beginning in full force in 1997, psychoactive drugs became yet another consumer product advertised to the public in a myriad of artfully produced television, magazine, newspaper, and internet ads. In all of these, the message that doctors are somehow in control is unfailingly transmitted (“Ask your doctor . . .”), but its repetition has finally served to underscore its absurdity. Advertisers and “drug reps,” not doctors, are now the actual middlemen between people and their drugs, just as they were a century ago, in the wild era of “patent medicines” (Young, 1961).
Daily, with the full blessings of medical and educational authorities, millions of parents give their prepubescent children psychoactive drugs—some of which, like amphetamines (e.g., Adderall), were banned only twenty years earlier in the United States and whose use or distribution without a medical prescription can elicit draconian penalties.2 The boundary between licit and illicit psychoactive drug use is vanishing: celebrities like Michael Jackson, Heath Ledger, and Anna Nicole Smith “accidentally” die from ingestions of duly prescribed medications, and experts in mental health, law, and ethics call for the prescription of psychoactive drugs to the healthy for cognitive enhancing or “neuroenhancement” (Greely et al., 2008). Several factors are interacting in complex ways to produce these trends. They include the increased role of consumer preferences and choices, the reverberation of popular discourse, direct-to-consumer advertising campaigns, and the pronouncements of experts. This is not unlike factors such as consumer preferences, the price of fuel, and advertisements interacting to influence individual purchases of consumer goods such as automobiles.
Moreover, the once-hallowed market approval of a drug by a powerful regulatory agency like the FDA has become, from the viewpoint of pharmaceutical firms, the last obstacle to promoting drugs aggressively with only the barest connection to the indication and age group for which the FDA approved the drugs. Doctors who might claim to practice “evidence-based psychiatry” routinely prescribe drug cocktails (combinations of two or more psychiatric drugs) whose effects no published controlled study has ever probed (Bhatara et al., 2004).
As for the published studies themselves—as David Healy has suggested and as we explore ahead, “There is probably no other area of medicine in which the literature is so at odds with the raw data” (cited in Leo, 2006, p. 31). And, the most vulnerable—children covered by Medicaid because of low income—are four times more likely to receive antipsychotic drugs and for less severe “conditions,” than children with private insurance (Wilson, 2009). According to most available estimates, about two-thirds of all money spent to purchase psychotropic drugs (and mental health and substance abuse treatment in general) in the United States is spent by government programs, prominently Medicaid and Medicare (e.g., Donohue, 2006).
National headlines were made in 2007 by the heartbreaking case of Rebecca Riley, a four-year-old child found dead on her parents’ bedroom floor from a drug overdose. The Riley parents were unemployed, and the family was almost wholly supported by Social Security benefits. Starting at age two, Rebecca was diagnosed as suffering from attention-deficit/hyperactivity disorder and bipolar disorder, almost entirely on the basis of the accounts of her behavior given by her mother to Kayoko Kifuji, a child psychiatrist employed by Tufts-New England Medical Center. Dr. Kifuji also prescribed to Rebecca (and her two slightly older siblings, also diagnosed with bipolar disorder by Dr. Kifuji) the antipsychotic quetiapine (Seroquel, $3 billion in US sales that year), the anticonvulsant depakote, and the antihypertensive drug clonidine (used in child psychiatry for its sedative properties, often prescribed in tandem with the stimulant methylphenidate, or Ritalin). Rebecca also was given two over-the-counter cold remedies (both of which were, a year later, in an unrelated action, “voluntarily” withdrawn by drug manufacturers for use in very young children). None of the three prescribed drugs were FDA-approved for psychiatric use with children because the FDA had never considered any evidence of even their short-term, superficial behavioral benefits for children.
Yet, immediately after Rebecca’s death, her medication regimen was defended by Tufts-New England Medical Center, whose spokesperson stated: “The care we provided was appropriate and within responsible professional standards” (WCVB TV, 2007). As recounted in a series of grim reports, Rebecca’s father had a confirmed history of abusive behavior. Both parents were chronically unemployed and seemed intent on declaring the whole family mentally ill to increase their financial support from state agencies. Moreover, they appeared to ignore Rebecca’s complaints during her last painful hours of life. This shielded Dr. Kifuji from criminal prosecution, and in separate trials the Riley parents, Michael and Carolyn, were convicted of first- and second-degree murder, respectively (Wen, 2010). Dr. Kifuji continues to practice as a child psychiatrist. Kaitlynne Riley, Rebecca’s older sister by two years, away from her parents and Dr. Kifuji since her sister’s death, was said to have “recovered” from her bipolar disorder, symptom-free and drug-free (Diller, 2010).
In another case, Nada Stotland, then president of the American Psychiatric Association, commented on the simultaneous prescription of the antidepressant Prozac, the tranquilizer Xanax, and the sleeping pill Ambien by a psychiatrist to a tormented twenty-seven-year-old man, Steven Kazmierczak, who eventually went off the drugs and then went on a shooting rampage at Northern Illinois University, where he had formerly been a student. Although some observers might interpret Kazmierczak’s actions as evidence that the drugs were helping him, others could point to the unusual reactions occasionally reported as triggered by combinations of psychoactive drugs and by the rebound and withdrawal reactions triggered by medications stopped abruptly. Stotland was quoted as stating: “It is not unusual for doctors to prescribe anti-anxiety agents and sleep medication along with antidepressants, since [antidepressants] can cause anxiety and interfere with sleep” (CNN, 2008).
That very young children could be psychiatrically diagnosed and medicated so arbitrarily and that national psychiatric authorities could state matter-of-factly that the drugs they prescribed to treat anxiety and depression “can cause anxiety and interfere with sleep” in effect meant that society had given psychiatry a blank check. As the biological wing of psychiatry and the mental health movement had ascended during earlier decades, new catch phrases and slogans justified the ancient human rituals of taking and giving drugs. Everywhere, chemical imbalances, neurotransmitter malfunctions, and impaired neural circuits were corrected, balanced, modulated, or stabilized by evidence-based treatments. These vague, medical-sounding phrases cloaked people’s unquenchable desires for psychoactive drugs in the explanatory language of people’s need for essential treatments provided by medically informed experts.
Starting in the fall of 2005 and still continuing as these lines are written, a series of articles appeared in the major American psychiatric and medical journals, reporting the results of three exceptionally large NIMH-funded studies evaluating the effectiveness of state-of-the-art psychiatric drug treatments for schizophrenia, depression, and bipolar disorder. The three studies, reportedly costing taxpayers over $100 million, provided what a Boston Globe journalist described as “the most realistic look ever at the state of psychiatric medicine” (Allen, 2006). The studies showed unambiguously that drugs do not make most people considered psychiatrically impaired significantly better for any sustained period of time. Since the studies appeared, mainstream popular and professional reactions suggest that the demonstration has been absorbed, digested, and recycled like every such negative finding over the past sixty years: it appears to have made only a tiny dent in the myth of psychiatric progress.
The first article appeared in the September 18, 2005, issue of The New England Journal of Medicine. Its lead author, Jeffrey Lieberman, chairman of the Department of Psychiatry at Columbia University, and the coauthors read as a list of the “Who’s Who” of schizophrenia drug treatment in the United States (Lieberman et al., 2005). The article was the first of many reporting on the results of a psychiatric mega-study, Clinical Antipsychotic Trials of Intervention Effectiveness, since known by its acronym, CATIE. The general aim of the study was to compare five different antipsychotic drugs in the treatment of nearly 1,500 patients diagnosed with psychosis and schizophrenia and recruited from fifty-seven different sites. Four of the drugs were second- or third-generation antipsychotics (olanzapine, quetiapine, risperidone, and ziprasidone), approved by the FDA between 1994 and 2002. The fifth was perphenazine, a nearly forgotten antipsychotic introduced in the late 1950s, almost never prescribed today.
To place the findings of this study in some context, one should note that schizophrenia is one of the few conditions that psychiatry can be said to “own”; nothing better exemplifies the application of psychiatric knowledge than treating schizophrenia. It is widely held that psychiatry came of age in the 1950s when it stumbled upon chlorpromazine (named Thorazine in the United States) and similar drugs that suppressed the agitation of asylum inmates. Since then, the drugs, first named neuroleptics and eventually antipsychotics, have been given to hundreds of millions of people throughout the world and are consecrated as having ushered in the “psychopharmacological revolution”—the very subject of this chapter and the next. Although schizophrenia is considered to affect about 1 percent of the population, during the first decade of the twenty-first century, antipsychotics ranked among the five top-selling classes of all prescription drugs in America. This was because most of their prescriptions were for problems outside schizophrenia and bipolar disorder, the drugs’ approved indications by the FDA, and because their manufacturers had accurately gauged that third-party payers would spend hundreds of dollars monthly for individual courses of treatment prescribed by appropriately primed physicians.
In CATIE, each entering patient was randomly assigned to receive one of the drugs and was followed for up to eighteen months or until he or she dropped out of the study or was switched to another drug. Notably, the primary outcome reported in the 2005 article was not a measure of the drugs’ intended benefits. The researchers did not convey effectiveness by the usual means used in clinical trials: ratings of what the researchers believe disturbs the participants or marks their behavior as abnormal (psychiatric symptoms), or of the participants’ reports about their ability to fulfill conventional social, occupational, and interpersonal roles (degree of impairment) over the course of the study. Instead, researchers emphasized only how long subjects remained on their assigned drug.
This simple outcome reported in the first CATIE report would suggest that the benefits of antipsychotic drugs are so well established that only a crude proxy measure is needed to demonstrate that they actually improve the diagnosed problem in the population studied. (If that were the case, of course, the CATIE trial would have been unnecessary.) The measure is deceptive, however, because the actual benefits of giving people psychoactive drugs are difficult to ascertain objectively, especially if the drugs are given to people who do not want to take them for long. Typically, antipsychotics are subjectively experienced as producing sedation, heavily clouded cognition, emotional dampening, and abnormal involuntary movements (Moncrieff, Cohen, & Mason, 2009). Moreover, people with diagnoses of psychotic disorders are relentlessly pressured or coerced to take and stay on antipsychotic medications, regardless of their feelings about doing so—which takes away much of the validity of the outcome measure in this CATIE study. Finally, as we discuss later, researchers in psychiatry have invested precious little creativity in determining how to assess the complex effects of psychoactive drugs taken daily over months and years. Yet for all its limitations the outcome measure was self-evidently practical, as a person who stops taking a drug cannot benefit from any of the drugs’ potentially desirable effects.
In any case, the proxy measure in this CATIE study painted an unfavorable portrait of the drug’s effectiveness: by study’s end, an average of 74 percent of patients on each drug (range, 69 to 84 percent) had “discontinued their assigned treatment owing to inefficacy or intolerable side effects or for other reasons” (Lieberman et al., 2005, p. 1209). Patients on olanzapine (Zyprexa) did stay longer on that drug, but they were also more likely to gain significant weight and show abnormalities in glucose and lipid metabolism. Trilafon (perphenazine), the obsolete antipsychotic selling for about 3¢ a dose, lost only slightly more patients from its group than Zyprexa, selling for three hundred times as much and ranked just two years earlier as the world’s most prescribed antipsychotic (between 1996 and 2002, its first six years on the market, Zyprexa’s gross sales were $13 billion, and it has gone on to gross over $45 billion as of this writing).
In January 2006, the American Journal of Psychiatry published one of the largest studies ever conducted on the drug treatment of depression (Trivedi et al., 2006). Also characterized by its acronym, STAR*D (for Sequenced Treatment Alternatives to Relieve Depression), this study recruited over 4,000 participants (the Trivedi et al. article reported on a first wave of 2,900 followed for fourteen weeks of planned treatment). Unlike carefully screened subjects in shorter clinical trials, most participants in STAR*D resembled ordinary patients who would ordinarily end up treated with antidepressants in outpatient doctors’ offices. They were recruited only when they sought care from over forty different outpatient general medical and psychiatric clinics across the United States.
All patients in STAR*D were first treated with Celexa (citalopram), a selective serotonin reuptake inhibitor (SSRI) approved in 1998 by the Food and Drug Administration to treat major depressive disorder (MDD). Celexa was chosen by the investigators because it is the most serotonin-selective of the SSRIs. Dosage was flexible and individually tailored by each patient’s own doctor, based on ongoing feedback from the patient. The results, then, were expected to be more representative of ordinary practice situations than the results from controlled clinical trials, which normally used fixed doses and rarely last beyond eight weeks. Also, to provide a relatively unambiguous outcome measure that would be immediately relevant to anyone, remission (or virtual disappearance of symptoms, rather than relative improvement compared to a placebo-treated control group, as in controlled clinical trials) was employed in STAR*D to determine the outcome. In sum, here was what appeared as a fairly “realistic” test of an already extremely popular drug treatment. The findings: a bit more than one quarter of the participants (29 percent) receiving a full course of SSRI drug treatment for depression (which lasted on average about eighty-three days), were reported by Trivedi and colleagues to have experienced a remission, while 47 percent were said to have shown a “response.”
Two months later, on March 23, 2006, a second article from the same study was published in the New England Journal of Medicine (Rush et al., 2006). It reported on a subsample of 727 left over from the original 2,900 participants. The subsample was made up of still depressed individuals who had immediately agreed to undertake a second course of drug treatment, to last another fourteen weeks, despite their failed attempt with Celexa. These subjects had then been randomly assigned to receive one of three other antidepressants: Effexor (venlafaxine) or Wellbutrin (bupropion) in extended release forms, or Zoloft (sertraline). Slightly less than half of the people in each antidepressant group also received concomitant psychotropic drugs, described as trazodone (for sleep), anxiolytics (excluding alprazolam), and sedative-hypnotics. (This raises the question: what drugs, or drug combinations, were actually being tested?) By the end of that continuation study, an average of 21.2 percent of the patients in each group (varying from about 18 to 27 percent, depending on the measure used by the investigators) had experienced a remission (with no statistically significant difference between any of the groups). However, more patients dropped out of the study for intolerance of the drugs than experienced a remission. More than half of people in each group experienced adverse effects of greater than “moderate” intensity. About half of the people in each drug group experienced adverse effects for more than half of the time they remained on treatment, including about one quarter who reported experiencing them 90 to 100 percent of the time.
The mainstream media universally put a positive spin on the disappointing findings, as illustrated by this comment from the medical newsletter MedPage Today (Osterweil, 2006): “STAR*D investigators demonstrated that in the treatment of major depression, patience and persistence can pay off.”
Despite STAR*D’s lackluster reports, the findings should have painted an even worse portrait of antidepressants’ effectiveness. “Publication bias” is the name given to the practice of skewing, hiding, suppressing, twisting, or cooking data in order to inflate the apparent value—the “efficacy”—of marketed drugs. In a series of publications and in a blog, psychologist Ed Pigott (2011a; 2011b) has conducted his own review and reanalysis of the STAR*D trial findings, as have other researchers (Cohen, 2010; Leventhal & Antonuccio, 2009; Pigott, Leventhal, Alter, & Boren, 2010). These authors have reported a disturbing pattern of publication bias, including the following examples, most of which involve deviations from the prespecified STAR*D research protocol.
•Patients who dropped out of the first antidepressant treatment step without completing a second Hamilton Rating Scale for Depression (HRSD) were to be counted as treatment failures. Instead, these patients were simply excluded from analyses, thus inflating the reported success rates.
•Patients with an HRSD score below 14 points were to be considered too mildly depressed for inclusion in the study. Instead, they were included. An additional 324 patients lacking an initial HRSD score were later included. In all, 931 of 4,041 patients (23 percent) failed to meet the prespecified eligibility criteria but nonetheless were included in the analyses, again inflating the drugs’ reported efficacy.
•Also contrary to the STAR*D research protocol, the authors switched to using a proprietary assessment measure, the Quick Inventory of Depressive Symptoms–Self Report (QIDS-SR) as the sole measure to report outcomes in the key summary article from STAR*D. This switch inflated published remission and response rates. In the QIDS-SR, a patient reporting, “I feel sad more than half of the time” or “Most of the time, I struggle to focus my attention or to make decisions” and “I rarely eat within a 24 hour-period, and only with extreme personal effort or when others persuade me to eat” would earn a score of 5 on the QIDS-SR—qualifying that patient as remitted. More than 90 percent of such remitters had at least one residual symptom (as a group, remitters showed a median of three symptoms), usually sleep disturbance, appetite or weight disturbance, sad mood, fatigue or decreased energy, and decreased concentration.
•As of mid-2011, despite over a hundred articles having been published from the STAR*D findings, eleven prespecified research measures collected to evaluate outcomes at entry and exit from each treatment step and every three months of follow-up still had not been reported.
These deviations from protocols, presentations of ad hoc findings as the prespecified outcomes of interest, and uses of explicitly excluded strategies to inflate apparent drug efficacy show how unwilling elite researchers and their institutional collaborators are to telling it like it is. The STAR*D investigation stands out from others due to its breadth and scope, but some of its reports provide more of the same endless psychopharmacological spinning that debases countless studies in the field by doing everything possible to avoid challenging the ideological, intellectual, and clinical status quo. Whether the investigators were deceiving themselves or deceiving others or both, the published STAR*D findings exemplify the sad but stark fact that an undetermined portion of psychopharmacological studies—especially those conducted by researchers who have made a successful career of promoting psychiatric drugs—cannot be trusted unless their findings are checked and rechecked (though the original data may never be available).
Findings from the third study to consider, the largest ever of the treatment of bipolar disorder (known as manic depression since the 1890s), first appeared in the February 2006 issue of the American Journal of Psychiatry (Perlis et al., 2006). The study was named the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD). Manic depression refers to alternating periods of depression and mania. Although as bipolar disorder it has been extensively researched and discussed in the psychiatric literature and although that moniker has fully penetrated popular discourse, very little seems to be known about the condition. In a 2006 editorial in the American Journal of Psychiatry entitled “What Is Bipolar Disorder?” psychiatrist Alan J. Swann could only answer the question by saying it was “recognized as a potentially treatable psychiatric illness that has substantial mortality and high social and economic impact.” Swann continued, “Every aspect of its definition, boundaries, mechanisms, and treatment, however, is subject to debate. In fact, there is no objective measure that can determine if one has bipolar disorder or does not have it” (p. 177). Naturally, as we have discussed earlier in this book, lacking objective measures to define or even detect the condition objectively should pose considerable problems for evaluating its treatment. The condition came to prominence in the 1970s when some psychiatrists claimed (and still do) that ordinary lithium salts were a specific and preventive treatment for it, conjuring the image of lithium-deficient madpersons transformed by a mineral supplement. When later studies showed that people taking lithium experienced mania or depression virtually as often as patients not taking any drug and when lithium toxicity showed itself to be insidious and potentially fatal, lithium begin falling out of favor (Healy, 2008). For the pharmaceutical industry, lithium also carried the enormous disadvantage of not being patentable, which may have doomed the intensive commercialization of that substance. Older and newer anticonvulsant drugs used in the treatment of epilepsy and other seizure disorders, but also for a variety of behavior problems in adults and children, started to replace lithium as the name and the presentation of manic depression changed to bipolar disorder, starting with DSM-III in the early 1980s. These drugs have been viewed since the early 1990s as the best available treatment for bipolar disorder.
Experts justified the shift to prescribing anticonvulsants by analogizing an episode of mania to an epileptic seizure. In seizure disorders, anticonvulsants seem to stop the “kindling” of seizures in the brain of many affected persons and reduce the likelihood of future seizures. Some psychiatrists hypothesized that anticonvulsants could similarly suppress an episode of mania while reduce the likelihood of future manic episodes. Rigorous demonstration of this claim was lacking, especially since studies of anticonvulsants used comparisons with poorly performing lithium. Like lithium, however, anticonvulsants are powerful depressants of the activity of the central nervous system (a property that explains their anti-seizure action, just like benzodiazepines). As such, they typically tranquilize and sedate their users. Also, relative to placebo, tranquilizing drugs will almost inevitably produce “improvement” on virtually any of the short-term symptom rating scales used in psychiatric drug trial research, because such scales contain many items related to nervousness, insomnia, agitation, and spontaneous verbalizations of variously offending or disturbing thoughts—all of which “respond” to sedative drug effects.
Moreover, the rebranding of anticonvulsants as “mood stabilizers” by clever pharmaceutical company executives gave a tremendous boost to marketing these drugs, as well as the bipolar diagnosis itself (Healy, 2006). The new name implied that anticonvulsants could directly “raise” depressed mood. Although it has no official standing in any drug nomenclature, “mood stabilizer” is an instantly recognizable designation of some psychiatric drugs today, and it appears in dozens of reports published each year. Its popularization must rank as one of the great marketing feats of the past decades.
As in STAR*D, the STEP-BD investigators tried to emulate real-world treatment conditions. They employed a best-available treatment protocol, including individually tailored monitoring of a broadly representative sample of almost 1,500 patients given anticonvulsants, typically with other drugs such as antidepressants, antipsychotics, lithium, and benzodiazepines (only 10 percent of the patients were prescribed a single drug). The patients’ treatment and response were then monitored for up to twenty-four months. As with the contemporaneous CATIE and STAR*D concerning psychosis and depression, respectively, STEP-BD constituted perhaps the largest group ever of people diagnosed with bipolar disorder and followed so carefully for such a length of time.
The main findings of STEP-BD were also eerily consistent with those of its two large sister studies. By study’s end, less than one-third of the sample (31 percent) were reported to have experienced a remission of their manic or depressive episode with no recurrence during the rest of the study period.
While writing this book, we did not compare the study protocols of CATIE and STEP-BD with the published findings, nor have we crosschecked the published methodologies across separate articles to assess the coherence and consistency of the findings. Yet, given what has been uncovered regarding STAR*D and given previous statements from other established researchers—that their own co-investigators in other landmark federally-funded studies have tried to suppress disturbing findings of drugs’ dangers (see, for example, one account by Vedantham, 2009)—readers should exercise due caution in accepting any claim of drug efficacy put forth in our drug-soaked culture (Healy, 2012).
The findings from CATIE, STAR*D, and STEP-BD were actually not too surprising to knowledgeable observers, because they were almost identical to findings from countless short-term (six to twelve weeks) clinical trials that make up the contemporary psychiatric treatment literature since the 1980s. For example, dropout rates in some short-term olanzapine (Zyprexa) trials that led to that drug’s market approval by the FDA in 1996 showed that 73 percent of subjects had left the study after four weeks, a rate the FDA reviewer called “overwhelming” (cited in Jackson, 2003). The average CATIE dropout rate was 74 percent. As to the rate at which subjects responded positively to drugs, during the first treatment phase of STAR*D, 47 percent of recruited subjects were rated as responders, much like the average 50 percent responder rate among antidepressant-treated subjects estimated by the Food and Drug Administration in 2006 from its database of 189 randomized, short-term placebo-controlled clinical trials of antidepressants involving over 53,000 patients (Stone & Jones, 2006).
When a drug manufacturer wishes to market its product, the very best of these trials are submitted to the FDA so that claims of efficacy and safety can be verified. FDA regulations require only two multisite randomized controlled trials that have produced positive results, regardless of how many negative trials were conducted. Randomized controlled clinical trials became the standard test for drug makers to demonstrate the efficacy (and some of the safety) of drugs for specific DSM disorders, which would grant an FDA approval to market drugs for these indications. In a randomized controlled trial, “Patients should be randomized to matching groups; be tested using both the active therapy and a control, or placebo; should not know which group they are in, nor be aware of whether they are getting an active drug or a placebo. They must be blinded from the doctor who might influence the treatment” (Nesi, 2008, p. 121). These precautions are essential in principle, but the actual conduct and reporting of clinical trials (whether for antipsychotics, antidepressants, or stimulants) has always left much to be desired, as review after review has observed (e.g., Jaddad et al., 1999; Larkin, 2010; Thornley & Adams, 1998). This problem has been magnified as the drug industry increased its already considerable involvement in funding, designing, conducting, and disseminating results of clinical trials of its own products.
For the FDA, moreover, “efficacy” has only meant demonstrating some kind of effect, or so-called proof in principle. Cure, or extended recovery, or remission has never been the anticipated outcome in the clinical trials evaluated by the FDA. On the contrary, therapeutic effects from psychoactive drugs have always been expected to be relatively small, and the problems being treated have always shown large, spontaneous variations across and within individuals. That is why phase III trials (the multisite, randomized controlled trials) had to be quite large in order to show some kind of effect (Healy, Langmaark, & Savage, 1999). If a drug were clearly efficacious, this could be shown in very small trials (think of an antibiotic treating a bacterial infection), and the results of efficacy should be easily replicable. Clinical trials of some of the century’s best-selling prescribed psychotropics, the SSRIs, never, ever demonstrated anything near this level of efficacy. The SSRIs showed weak, marginal effects in comparison to placebo.
In 2002 and again in 2008, Irving Kirsch and colleagues reanalyzed all data they obtained using Freedom of Information laws from the FDA’s evaluations of the forty-seven RCTs funded and submitted to it by the makers of the six most widely prescribed antidepressants approved by the FDA between 1987 and 1999. Nearly 5,200 individuals had participated in these trials, including 3,300 randomized to receive drugs and 1,800 to receive placebo. The reanalysis found that 82 percent of the response of medicated patients was duplicated in placebo-treated patients. This occurred despite the FDA allowing researchers to replace people on some of the drugs who were not improving after two weeks into the trial. Moreover, in most trials, tranquilizers were also prescribed to the participants whose insomnia and agitation increased (probably as an adverse effect of the SSRIs). Despite these methodological breaches (which might serve to explain why only 82 percent of the drug response was duplicated by placebo), on the chief outcome measure, the patient’s score on the Hamilton Depression Rating Scale (HAMD), the mean difference between drug and placebo groups was 1.8 points (the version of the scale most often used in the trials has a maximum 50-point score). In their later reanalysis, Kirsch and colleagues (2008) did observe that among people with the highest HAM-D scores (suggesting the most severe depressive symptoms), the outcome difference (just about 4 points on the HAM-D) favored the drug-treated groups. But this was shown to reflect a waning of the placebo effect among such people, not an increased efficacy of the SSRIs. In other words, drug response scores among the most severely depressed were similar to scores among the less depressed, but placebo response scores were not (perhaps because when one feels quite bad it is much harder to summon a positive outlook).
Obviously, the very small average difference between groups could not in itself mean a clearly superior gain for most patients who take SSRIs. On a large enough sample of people, however, given the laws of probability, it could show statistical significance. In this case, statistical significance meant that the 3 percent smaller uselessness of SSRIs compared to placebo would be quite unlikely to result from chance and should be observed almost every time drugs are compared to placebo. Because of this mathematically conclusive but probably clinically negligible difference, the FDA followed its rule that the manufacturers had provided evidence, in at least two clinical trials for each drug, of the drugs’ efficacy to treat whatever was treated in the clinical trials.
As noted, the high placebo response rates were obtained despite a number of strategies deliberately designed to reduce placebo responses from the participants (Cohen & Jacobs, 2007). First, almost every study used placebo-washout or placebo-lead-in periods, wherein all subjects are abruptly discontinued from any drugs they might be taking and placed on a pill-looking placebo. The idea is to identify people who are likely to improve early when they believe that they are receiving a genuine treatment—and to exclude these people from those who will then be randomized for the “start” of the “actual” study. For example, in the forty-seven SSRI trials reviewed by Kirsch and colleagues, any subject whose HAM-D score improved 20 percent or more during this early period was excluded from the study. But since the point of a trial is to compare the efficacy of active drugs to placebo, it is by no means clear what the rationale could be for excluding people who tend to respond to placebo early on. To be fair, those who respond to drugs early on should also be excluded.
Second, all studies employed inert rather than active placebos. An inert placebo might contain yeast, sugar, or lactose. An active placebo might contain caffeine, benztropine, or diphenhydramine. Why use one or the other? Certain drug effects, such as dry mouth, dizziness, sleepiness, or increased heart rate can serve as early cues to patients and clinicians about which treatment condition patients are in. This would “break the blind,” threatening the trial’s internal validity (the confidence to draw conclusions about what is causing what). One way around this problem is to use active placebos: substances that give some physical sensations to people, who might therefore think that they are receiving the “real” medication, but which do not produce the sort of psychoactive effects of the tested drug.
Results from some antidepressant trials in the 1970s and early 1980s showed that active placebos did indeed produce greater placebo response than inert placebos (Fisher & Greenberg, 1993a, 1993b). It therefore seemed likely that the subtle physical sensations produced by active placebos had tricked subjects into thinking that they were taking “real” medication, thus possibly amplifying the subjective feeling of improvement that they would communicate to the clinician raters during their brief structured encounters. This would translate as higher scores of these subjects on rating scales—the major outcome of the clinical trials. A different dynamic would occur in a study comparing a drug to an inert placebo. Here, noticeable physical sensations are more likely to be experienced by those taking the drug, which would give clinicians a cue about which treatment condition subjects are in, and might tilt clinicians’ expectations toward observing improvement in drug-treated patients.
Reviews and reanalyses of these earlier trials (Moncrieff et al., 2004; Petovka et al., 2000) have not settled the matter of the precise role of inert versus active placebos, and one author suggests ignoring this issue altogether because the studies are too old (Quitkin, 2003). But the issue remains profoundly relevant: if mere sugar pills produce equivalent or nearly equivalent responses in the most sophisticated controlled trials of antidepressants today, wouldn’t active placebos do at least as well and probably much better? That is, wouldn’t active placebos lay waste to antidepressants’ effectiveness nearly every time? This question has no known answer, because as of this writing in late 2011, one still cannot find a single major randomized controlled trial comparing a modern SSRI antidepressant with an active placebo. By contrast, in neurology, controlled trials of pain treatments routinely use active placebos. This fact alone should bracket clinical psychopharmacology as a studiously unscientific enterprise. Let us restate the point differently and explicitly.
Pain researchers, who know that pain often responds to placebo, appear truly interested to discover whether new analgesics work better than placebo. Thus pain researchers stack the deck against their proposed analgesics by using placebos that superficially create a sensation to trick the patient into thinking that an analgesic may have been taken. The hurdle that must be overcome by the proposed analgesic is thus raised higher. By contrast, psychiatric researchers, who also know that depression often responds to placebo, appear uninterested to discover whether modern psychiatric antidepressants work better than placebo. Thus psychiatric researchers stack the deck against placebo. The hurdle that must be overcome by the proposed antidepressant is thus lowered. From either a scientific or clinical perspective, the absence of rigorous comparisons between drugs and active placebos in psychiatry makes no sense. It may nonetheless be easily understood when ideological or commercial imperatives—for example, or the compulsive need to suggest that genuine illnesses are being targeted by specific treatments or the compulsive need to demonstrate a profit every quarter to satisfy mutual fund investors—shape the pursuit of knowledge.
In controlled trials designed to minimize placebo response, inert placebos given to subjects who meet the DSM’s Major Depressive Disorder criteria benefit up to half of subjects, exactly as antidepressant drugs are shown to benefit subjects: by steady reductions in depression rating scale scores, from the very start of the trial and every week thereafter (Walsh et al., 2002). Nor are placebo effects confined to the treatment of depression and anxiety with antidepressants and anxiolytics (Shear et al., 1995). For example, in March 2009, Eli Lilly and Company announced results from a phase II clinical trial of its new investigational antipsychotic, named LY2140023, against its blockbuster drug olanzapine (Zyprexa). The press release stated: “In Study HBBI, neither LY2140023 . . . nor the comparator molecule olanzapine, known to be more effective than placebo, separated from placebo.” As a result, “neither drug performed better than the placebo” (Eli Lilly and Company, 2009, p. 1). Over the last two decades, placebo response rates have steadily increased in trials of antipsychotic drugs, in parallel with a steady decrease in drug response rates. In some trials, 50 percent of participants (diagnosed with schizophrenia and experiencing an acute psychotic episode) taking inert placebo pills show at least a 20 percent improvement on their psychosis rating scale score relative to baseline, the criteria sometimes used in the past to demonstrate efficacy of antipsychotic drugs (Kemp et al., 2008). Various factors may account for this steady diminishing of drug-placebo differences, ranging from the extra attention given clinical trial participants to differences in scoring methods used in foreign study sites, which are increasingly used today for drugs to be approved in America. Official discussions of the trend, however, conform to a predictable line: drug response is a true and worthy “signal,” while any placebo response is a distracting, confusing, confounding “noise” (Kemp et al., 2008).
How and why placebos produce such results in comparison to drugs whose psychoactive effects sustain them through years of animal and clinical testing (see next chapter) remains an utterly fascinating question. By themselves, the cost-savings and reduction in adverse effects implied by the possible amplification of the placebo effect could justify reorienting but a small part of the stupendous research activities in psychopharmacology toward understanding and harnessing placebo effects. Yet it is obvious that the psychopharmacology field as a whole has been much less interested in answering the placebo question than in disposing of it methodologically. This is not astonishing, as using placebos in the practice of medicine is a most controversial topic. Using placebos is at odds with the image that medicine has projected of itself—that of a scientific profession having discovered and devised, based on the scientific method, effective and specific treatments for genuine diseases. Within psychiatry, more than other branches of medicine, in which placebo effects may constitute most of clinical practice, systematically understanding the role of placebos would threaten the entire profession.
Because the placebo is unpretentious and nameless, its competitors—compounds that have passed the FDA-approval hurdle and been given branded identities as genuine medications—can claim placebo’s benefits, and any deception basically remains unexposed. One might view the placebo effect in many different ways: as a vital human capacity for self-restoration, a complex form of self-deception, and anything in-between. Either way, it is difficult to dismiss it as “no treatment” or “nothing.” One can dismiss placebo effects as noise only if one is resolutely committed to finding uses for a commercially-funded investigational drug—a different aim than being resolutely committed to finding out what people say helps them, or what helps people despite what they say helps them.
Here is Charles Medawar and Anita Hardon’s social view of the placebo effect:
All sorts of human beliefs, motivations and styles of organization make medicines work as they seem to do. These factors are typically invisible, go largely unrecognized and have no other name. They go largely incognito . . . The fact that they have no collective name underlines that their impact is generally unrecognised and therefore poorly understood. Probably the main reason is that the word placebo has come to stand for all the non-drug factors that affect treatment outcomes. It does not seem an appropriate word: it does not take into account many non-pharmacological factors on treatment outcomes. (2004, p. 175)
How much are expectations of patients amplified when virtually all authoritative discourse about a drug consists of statements that it has positive effects on those who receive it from their doctor? The felt effectiveness of a medicine has to do not only with the relationship between user and drug. Doctors’ and patients’ experiences and history, their values, statuses, and roles, their faith and hope—not to mention suggestion from their peers and from advertising, drug company sponsorship of doctors’ events, the names given to drugs—these and other factors must also count as important, interacting determinants in drug “effectiveness,” as they introduce themselves in still largely unknown ways into the clinician-patient relationship.
It is sobering to realize that the technique of testing a drug in a double-blind trial against a placebo, a so-called dummy pill, is at present the only existing scientific guarantee that the drug does not work somehow through suggestion, hope, and belief. Today, without such a trial, a mere chemical compound could not become a “medication,” as bestowed by FDA approval (and this, for conditions ranging from migraine to Parkinson disease to multiple sclerosis). Without comparing the tested drug to something that can elicit improvement from the patient merely by appearing to be a genuine medication, one cannot confidently determine the true effects of the tested drug. This makes the placebo especially necessary when testing treatments for conditions with a substantial rate of spontaneous or developmental improvement, such as depression and anxiety or a first episode of psychosis. But the nature of the placebo effect remains largely unknown (Benedetti, 2009). Something or a combination of many things does appear to be there, and it begs explanation. Of course, we cannot yet rule out that the placebo effect may turn out to be another colossal reification, much like mental illness, an idea whose concrete reality is assumed merely because it has a name and because alternative explanations for its associated features have not been widely understood or disseminated.
But even if we stick to the placebo effect as traditionally defined in drug trials (the measurable therapeutic changes induced by pharmacologically inactive or inert substances), we may consider the conclusion reached by psychologists Seymour Fisher and Roger Greenberg. After one of the most insightful, restrained, and evidence-informed reviews in the entire literature on antipsychotics, stimulants, antidepressants, and anxiolytics, these two authors stated, “that when proper controls are introduced [in research studies] the differences in therapeutic power between the active drugs and placebos largely recede” (1997, p. 382). It is sobering for anyone to reflect that merely introducing “proper controls” in scientific research would doom the modern psychopharmacologic enterprise.
In an editorial accompanying STEP-BD’s first publication, psychiatric researcher J. Raymond DePaulo, Jr. (2006) observed that “the results are so similar to those observed in lithium treatment studies of 20 to 35 years ago that the question of whether today’s best available treatments for bipolar disorder are better than older treatments begs for an answer” (p. 175). But the study actually did answer the question—in the negative. DePaulo omitted to mention that a few earlier, long-term observational lithium studies had themselves suggested that lithium was no better than lithium’s predecessors—shock treatments and barbiturates—and even, in some investigations, no better than no drugs at all (Harrow et al., 1990). DePaulo did generously allow for the possibility that “modern pharmacological treatments may be no more beneficial than older ones, despite their added cost” (p. 175). And, one might add, despite their added hype.
Indeed, exaggerated publicity intended to excite public interest has characterized the portrayal of psychiatric drug treatments these past few decades. Who has not heard that drugs have revolutionized the treatment of mental illness, that newer medications are safer and more effective than older ones, that they can make their users “better than well,” and that drugs have removed the stigma of psychiatric labeling? It has even been said that drugs can be so helpful that not forcing them on certain disturbed people constitutes negligence or worse. In sum, experts and their followers have stated that psychiatric medications have brought society immeasurably closer to solving the problem of mental illness once and for all.
With unremitting support from many partners in the mental health movement, such claims were transformed into robust cultural beliefs or myths. The power and purpose of myths to give meaning to our lives and societies cannot be understated. In the age of science, myths are often portrayed as scientific truths. Myths unify people by expressing themes that resonate with people’s longings. When experts and laypersons continually reaffirm a myth, they convey their wish to belong to the group of believers. They may also convey that, as far as they are concerned, debate about the essential truth of the myth has ended and that one need only be concerned with technical details arising from the application of its essential truth.
That is how CATIE, STAR*D, and STEP-BD were greeted. Three large, exemplary studies conducted by the clinical and research elite of American psychiatry and funded by the world’s largest mental health research institute, the NIMH, to the combined tune of almost 100 million dollars, demonstrated dismal results, if not failure, of the paradigmatic biological treatments for the paradigmatic mental disorders. Yet, as far as one can tell, no prominent mental health authority (neither an investigator from one of the studies or a NIMH official) suggested re-evaluating the routine drugging, these past fifty years, of madness and misbehavior, or launching an intensive analysis of the practical uses of placebo effects in the drug treatment of all manners of mental distresses. Quite the contrary, experts and commentators only lamented the heavy burden of mental disorders and stoked the ever-so-great imperative to find newer, better drugs or biological treatments. In the world of mad science, the myth of treatable mental illnesses must go on.
To illustrate the persistence of probably wrong-headed thinking despite the facts, we turn again to NIMH director Thomas Insel. In 2009, Insel thus summarized the findings from the CATIE, STAR*D, and STEP-BD studies:
Given that these trials used evidence-based treatments of well-documented efficacy that were administered with optimal clinical standards, the results show the significant limitations of current pharmacological interventions. This point bears emphasis. These limited outcomes, in terms of recovery and remission, are not due to suboptimal delivery of care. Instead, with optimal care using today’s medications, too many people will not recover. (p. 129).
Insel nonetheless went on to announce a new program of research to uncover the “molecular pathophysiology” underlying mental disorders by focusing research on brain circuits, genes, and personalized medicine focusing on individual moderators of treatment outcomes such as “genetic traits, imaging results, plasma proteins, or clinical features” (p. 131).
The connections between the uses of drugs and the sort of biological thinking exemplified by Insel’s new call to action obviously need to be restated. Chemical substances with powerful effects on thinking, feeling, and behaving ushered an era of biological theorizing about madness as the consequence of chemical imbalances in the brain. In turn, that biological thinking was used to promote the usefulness of psychoactive drugs to treat mental distress and misbehavior of every stripe and variety. Then, the mental health system was entirely reorganized to encourage drug use among the distressed and strict compliance with one’s prescribed medication regimen thereafter. After decades of this arrangement, the drugs (“evidence-based treatments of well-documented efficacy”) are formally acknowledged by authoritative experts citing authoritative evidence to have little of the usefulness so long attributed to them. Yet the same experts and institutions who touted drugs’ benefits with certainty, tolerating no dissent and suppressing alternative viewpoints, today urge their constituencies to adopt even more far-flung biological thinking strung together by bits of medical technology and corporal components. We believe that these experts should pause to identify what may have gone wrong and what future errors might be avoided, but we do not think this appears anywhere in the official agenda.
In the following chapter, we develop the argument that the effort to develop a scientifically-based drug treatment of specific mental disorders has mainly consisted of the recycling of perennial human drugging practices. We also propose that despite all the hoopla, an understanding of psychiatric medications as psychoactive substances capable of producing complex mental and physical states regardless of why they are taken by people or prescribed by doctors appears wholly sufficient to explain drugs’ observed therapeutic effects (and adverse effects) in “mental disorders” (Jacobs & Cohen, 1999; Moncrieff, 2008; Moncrieff & Cohen, 2009). The notion of drugs specifically targeting identifiable or soon-to-be-identified pathological substrates of mental disorders remains a hypothesis supported by an exceedingly tiny amount of evidence. To paraphrase psychiatrist Peter Breggin and co-author Ginger Breggin (1994), to respond to any drug (that is, to like it, to find it helpful for some purpose, to want to keep taking it), a human being does not need a chemically imbalanced or an electrically short-circuited brain; just a brain will do. If so, then the modern psychiatric discipline of clinical psychopharmacology, having mostly tied its scientific fortunes to the validity of biological psychiatric theorizing, rests on the flimsiest of foundations and needs to be reconstructed, from scratch.
1.First isolated from opium at the start of the 1800s, morphine was massively used for the first time during the Austro-Prussian War of 1866 and found to be an unsurpassed painkiller. About a dozen years later, some physicians began to see it as causing a state of hellish chemical servitude they called morphinomania. By the end of the 1800s, morphine was condemned as a “main scourge of humanity” (Chast, 1995, p. 112).
2.We remain astonished at encountering parents who would balk at serving their nine-year-old child a single cup of coffee but who would feel less unease to give that child a daily dose of prolonged-action amphetamine, for several months.
Allen, S. (2006, March 6). Beyond guesswork: Treatments often fail, but new studies might help doctors find drugs and therapies that work. Boston Globe. Retrieved from: http://www.boston.com/yourlife/health/mental/articles/2006/03/06/beyond_guesswork/
American Psychiatric Association. (2008, September 2). APA responds to Senate request for records [News release]. Retrieved from www.psych.org/MainMenu/Newsroom/NewsReleases/2008NewsReleases/APAPresidentResponds.aspx
Bachman, C., & Coppel, A. (1989). Le dragon domestique—deux siècles de relations étranges entre l’Occident et la drogue [The domestic dragon: Two centuries of strange relations between the West and drugs]. Paris: Albin Michel.
Benedetti, F. (2009). Placebo effects: Understanding the mechanisms in health and disease. Oxford and New York: Oxford University Press.
Bhatara, V., Feil, M., Hoagwood, K., Vitiello, B., & Zima, B. (2004). National trends in concomitant psychotropic medication with stimulants in pediatric visits: Practice versus knowledge. Journal of Attention Disorders, 7(4), 217–226.
Breggin, P. R., & Breggin, G. (1994). Talking back to Prozac: What doctors aren’t telling you about today’s most controversial drug. New York: St. Martin’s.
Carlat, D. (2009). APA votes to phase out industry-funded CME. Retrieved from http://carlatpsychiatry.blogspot.fr/2009/03/apa-votes-to-phase-out-industry-funded.html
Centers for Disease Control and Prevention. (2006). Health, United States (Table 92). Retrieved from http://www.cdc.gov/nchs/hus.htm
Chast, F. (1995). Histoire contemporaine des médicaments [Contemporary history of medications]. Paris: Éditions La Découverte.
CNN. (2008). Shooting reopens debate over drug, violence link. Retrieved from http://edition.cnn.com/2008/HEALTH/02/25/antidepressants.violence/
Cohen, D. (2010). Psychopharmacology and clinical social work practice. In J. R. Brandell (Ed.), Clinical social work: Theories and methods, 2nd ed. (pp. 763–810). Thousand Oaks, CA: Sage.
Cohen, D., & Jacobs, D. (2007). Randomized controlled trials of antidepressants: Clinically and scientifically irrelevant. Debates in Neuroscience, 1, 44–54.
DeGrandpre, R. (2007). The cult of pharmacology: How America became the world’s most troubled drug culture. Durham, NC, and London: Duke University Press.
DePaulo, J. R. Jr. (2006). Bipolar disorder treatment: An evidence-based reality check. American Journal of Psychiatry, 163(2), 175–176.
Diller, L. (2010, March 27). Wholesale sedation of young children medically, morally indefensible. The Patriot Ledger (Quincy, MA). Retrieved from http://wickedlocal.com/hull/news/opinions/x905411678/COMMENTARY-Wholesale-sedation-of-young-children-medically-morally-indefensible
Donohue, J. (2006). Mental health in the Medicare Part D drug benefit: A new regulatory model? Health Affairs, 25(3), 707–719.
Eli Lilly and Company. (2009, March 29). Lilly announces inconclusive phase II study results for mglu2/3 at the International Congress on Schizophrenia Research. [News release]. Retrieved from http://newsroom.lilly.com/releasedetail.cfm?releaseid=373650
Fisher, S., & Greenberg, R. P. (1993a). A meta-analysis of antidepressant outcome under “blinder” conditions. Journal of Consulting and Clinical Psychology, 60, 664–669.
Fisher, S., & Greenberg, R. P. (1993b). How sound is the double-blind design for evaluating psychotropic drugs? Journal of Nervous and Mental Disease, 181, 345–350.
Fisher, S., & Greenberg, R. P. (1997). What are we to conclude about psychoactive drugs? In S. Fisher & R. Greenberg (eds.), From placebo to panacea: Putting psychiatric drugs to the test (pp. 359–383). New York: Wiley & Sons.
Fugh-Berman, A. (2008). Prescription tracking and public health. Journal of General Internal Medicine, 23(8), 1277–1280.
Gagnon, M. A., & Lexchin, J. (2008). The cost of pushing pills: A new estimate of pharmaceutical promotion expenditures in the United States. PLoS Medicine, 5(1), e1.
Greely, H., Sahakian, B., Harris, B., Kessler, R. C., Gazzaniga, M., Campbell, P., & Farah, M. J. (2008). Towards responsible use of cognitive enhancing drugs by the healthy. Nature, 456 (7223), 702–705.
Harrow, M., Goldberg, J. F., Grossman, L. S., & Meltzer, H. Y. (1990). Outcome in manic disorders: A naturalistic follow-up study. Archives of General Psychiatry, 47, 665–671.
Healy, D. (2006). The latest mania: Selling bipolar disorder. PLoS Medicine, 3(4), e185.
Healy, D. (2008). Mania: A short history of bipolar disorder. Baltimore: Johns Hopkins University Press.
Healy, D. (2009). Trussed in evidence? Ambiguities at the interface between clinical evidence and clinical practice. Transcultural Psychiatry, 46, 16–37.
Healy, D. (2012). Pharmageddon. Berkeley: University of California Press.
Healy, D., Langmaak, C., & Savage, M. (1999). Suicide in the course of the treatment of depression. Journal of Psychopharmacology, 13, 94–99.
Hegarty, J., Baldessarini, R., Tohen, M., Waternaux, C., & Oepen, G. (1994). One hundred years of schizophrenia: A meta-analysis of the outcome literature. American Journal of Psychiatry, 151, 1409–1416.
Hughes, S., & Cohen, D. (2009). A systematic review of long-term studies of drug treated and non-drug treated depression. Journal of Affective Disorders, 118, 2–9.
IMS Health. (2012). Top therapeutic classes by U.S. dispensed prescriptions. Retrieved from http://www.imshealth.com/deployedfiles/ims/Global/Content/Corporate/Press%20Room/Top-Line%20Market%20Data%20&%20Trends/2011%20Top-line%20Market%20Data/Top_Therapy_Classes_by_RX.pdf
Insel, T. R. (2009). Translating scientific opportunity into public health impact: A strategic plan for research into mental illness. Archives of General Psychiatry, 66, 128–133.
Jackson, G. E. (2003, March 3). An analysis of the olanzapine clinical trials: Dangerous drug, dubious efficacy. Retrieved from http://psychrights.org/states/Alaska/CaseOne/30-Day/ExhibitD-Olanzapine.htm
Jacobs, D. H., & Cohen, D. (1999). What is really known about psychological alterations produced by psychiatric drugs? International Journal of Risk and Safety in Medicine, 12, 37–47.
Jadad A. R., Boyle, M., Cunningham, C., Kim, M., & Schachar, R. (1999). Treatment of attention-deficit/hyperactivity disorder. Rockville, MD: Agency of Healthcare Research and Quality.
Kemp, A. S., Schooler, N. R., Kalali, A. H., Alphs, L., Anand, R., Awad, G. . . . Vermeulen, A. (2008). What is causing the reduced drug-placebo difference in recent schizophrenia clinical trials and what can be done about it? Schizophrenia Bulletin. Retrieved from schizophreniabulletin.oxfordjournals.org
Kirsch, I., Moore, T. J., Scoboria, A., & Nicholls, S. S. (2002). The emperor’s new drugs: An analysis of antidepressant medication data submitted to the U.S. Food and Drug Administration. Prevention & Treatment, 5, article 23. Retrieved from http://www.journals.apa.org/prevention/volume5/pre0050023a.html. Accessed July 15, 2002.
Kirsch, I., Deacon, B. J., Huedo-Medina, T. B., Scoboria, A., Moore, T. J., & Johnson, B. T. (2008). Initial severity and antidepressant benefits: A metaanalysis of data submitted to the Food and Drug Administration. PLoS Medicine, 5(2), e45.
Larkin, C. (2010, April 20). Pfizer’s Geodon trial had “significant” violations. Business Week. Retrieved from www.businessweek.com/news/2010-04-20/pfizer-s-geodon-trial-had-significant-violations-update2-.html
Leo, J. (2006). The SSRI trials in children: Disturbing implications for academic medicine. Ethical Human Psychology & Psychiatry, 8, 29–41.
Leventhal, A. M., & Antonuccio, D. O. (2009). On chemical imbalances, antidepressants, and the diagnosis of depression. Ethical Human Psychology & Psychiatry, 11, 119–214.
Lieberman, J. A., Stroup, T. S., McEvoy, J. P., Swartz, M., Rosenheck, R. A., Perkins, D. O. . . . Hsiao, J. K. (2005). Effectiveness of antipsychotic drugs in patients with chronic schizophrenia. New England Journal of Medicine, 353, 1209–1223.
Mathews, A. W. (2010, December 28). So young and so many pills: More than 25% of kids and teens in the US take prescriptions on a regular basis. Wall Street Journal. Retrieved from http://online.wsj.com/article_email/SB10001424052970203731004576046073896475588-lMyQjAxMTAxMDIwNDEyNDQyWj.html?mod=wsj_share_email
Medawar, C., & Hardon, A. (2004). Medicines out of control? London: Askant Publishers.
Moncrieff, J. (2008). The myth of the chemical cure. London: Routledge.
Moncrieff, J., & Cohen, D. (2009). How do psychiatric drugs work? BMJ, 338, b.1963.
Moncrieff, J., Cohen, D., & Mason, J. (2009). The subjective effects of taking antipsychotic medication: A content analysis of Internet data. Acta Psychiatrica Scandinavica, 120, 102–111.
Moncrieff, J., Wessely, S., & Hardy, R. (2004). Active placebos versus antidepressants for depression. Cochrane Database Systematic Reviews, 1, CD003012.
Moynihan, R., Heath, I., & Henry, D. (2002). Selling sickness: The pharmaceutical industry and disease mongering. BMJ, 324, 886–891.
Nesi, T. (2008). Poison pills: The untold story of the Vioxx drug scandal. New York: St. Martin’s Press.
Nierenberg, A. A., Ostacher, M. J., Huffman, J. C., Ametrano, R. M., Fava, M., & Perlis, R. H. (2008). A brief review of antidepressant efficacy, effectiveness, indications, and usage for major depressive disorder. Journal of Occupational and Environmental Medicine, 50, 428–436.
Osterweil, N. (2006, March 22). Antidepressant switch or add-on produces more remissions. MedPage Today. Retrieved May 19, 2006 from http://www.medpagetoday.com/Psychiatry/Depression/tb/2916
Patten, S. B. (2004). The impact of population treatment on population health: Synthesis from two national sources in Canada. Population Health Metrics, 2:9. Retrieved from http://www.pophealthmetrics.com/content/2/1/9
Perlis, R. H., Ostacher, M. J., Patel, J. K., Marangell, L. B., Zhang, H., Wisniewski, S. R., Ketter, T. A. . . . Otto, M. W. (2006). Predictors of recurrence in bipolar disorder: Primary outcomes from the Systematic Treatment Enhancement Program for bipolar disorder (STEP-BD). American Journal of Psychiatry, 163, 217–224.
Petovka, E., Quitkin, F. M., McGrath, P. J. et al. (2000). A method to quantify rater bias in antidepressant trials. Neuropsychopharmacology, 22, 559–565.
Pieters, T., & Snelders, S. (2005). Mental ills and the “hidden history” of drug treatment practices. In M. Gijswijt et al. (eds.), Psychiatric cultures compared. Psychiatry and mental health care in the twentieth century: Comparisons and approaches (pp. 381–401). Amsterdam: Amsterdam University Press.
Pigott, H. E. (2011a). STAR*D: A tale and trail of bias. Ethical Human Psychology and Psychiatry, 13, 6–28.
Pigott, H. E. (2011b, April 21). The STAR*D scandal. Retrieved from: http://www.madinamerica.com/madinamerica.com/Pigott.html
Pigott, H. E., Leventhal, A. M., Alter, G. S., & Boren, J. J. (2010). Efficacy and effectiveness of antidepressants: Current status of research. Psychotherapy & Psychosomatics, 79, 267–279. doi: 10.1159/000318293
Quitkin, F. M. (2003). Reply to “Active placebos versus antidepressants for depression” by J. Moncrieff, S. Wessely, & R. Hardy in The Cochrane Database of Systematic Reviews, 2002. Retrieved 27 June 2006 from: http://www.cochranefeedback.com/cf/cda/citation.do?id=9048#9048
Rasmussen, N. (2006). Making the first anti-depressant: Amphetamine in American medicine, 1929–1950. Journal of the History of Medicine and Allied Sciences, 61(3), 288–323.
Rush, A. J., Trivedi, M. H., Rush, A. J., Wisniewski, S. R., Stewart, J. W., Nierenberg, A. A., Thase, M. E., Ritz, L., Biggs, M. M., Warden, D., Luther J. F., Shores-Wilson, K., Niederehe, G., & Fava, M., for the STAR*D Study Team. (2006). Buproprion-SR, sertraline, or venlafaxine after failure of SSRIs for depression. New England Journal of Medicine, 354 (12), 1231–1242.
Shear, K., Leon, A., Pollack, M., Rosenbaum, J., & Keller, M. (1995). Pattern of placebo response in panic disorder. Psychopharmacology Bulletin, 31, 273–278.
Stagnitti, M. N. (2007, February). Trends in the use and expenditures for the therapeutic class prescribed psychotherapeutic agents and all subclasses, 1997 and 2004. Statistical Brief No. 163. Rockville, MD: Agency for Healthcare Research and Quality. Retrieved from http://www.meps.ahrq.gov/mepsweb/data_files/publications/st163.pdf
Stone, M. B., & Jones, M. L. (2006, November 17). Clinical review: Relationship between antidepressant drugs and suicidality in adults. Food and Drug Administration, Center for Drug Evaluation and Research. Retrieved from www.fda.gov/OHRMS/DOCKETS/AC/06/briefing/2006-4272b1-01-FDA.pdf
Swann, A. J. (2006). What is bipolar disorder? American Journal of Psychiatry, 163, 177–179.
Szasz, T. (1974). Ceremonial chemistry: The ritual persecution of drugs, addicts, and pushers. Garden City, NY: Anchor Books.
Thornley, B., & Adams, C. (1998). Content and quality of 2000 controlled trials in schizophrenia over 50 years. BMJ, 317, 1181–1184.
Tierney, J. (2010, April 11). Hallucinogens have doctors tuning in again. New York Times. Retrieved from www.nytimes.com/2010/04/12/science/12psychedelics.html
Trivedi, M. H., Rush, A. J., Wisniewski, S. R., Nierenberg, A. A., Warden, D., Ritz, L., Norquist, G., Howland, R. H., Lebowitz, B., McGrath, P. J., & Shores-Wilson, K. (2006). Evaluation of outcomes with citalopram for depression using measurement-based care in STAR*D: Implications for clinical practice. American Journal of Psychiatry, 163 (1), 1–13.
Vedantham, S. (2009, March 27). Debate over drugs for ADHD reignites: Long-term benefit for children at issue [Electronic version]. The Washington Post. Retrieved October 19, 2009, from www.washingtonpost.com/wp-dyn/content/article/2009/03/26/AR2009032604018.html
Walsh, B. T., Seidman, S. N., Sysko R., et al. (2002). Placebo response in studies of major depression: Variable, substantial, and growing. JAMA, 287, 1840–1847.
WCVB TV. (2007, February 6). Parents charged in death of daughter, 4. The BostonChannel.com. Retrieved from http://www.thebostonchannel.com/news/10943835/detail.html
Wen, P. (2010, March 26). Father convicted of 1st-degree murder in death of Rebecca Riley. The Boston Globe. Retrieved from http://www.boston.com/news/local/breaking_news/2010/03/hold_for_verdic_1.html
Wilson, D. (2009, Dec. 12). Poor children more likely to get antipsychotics. New York Times, A1, A11.
Young, J. H. (1961). The toadstool millionaires: A social history of patent medicines in America before federal regulation. Princeton, NJ: Princeton University Press.