Mad in America

The obvious question today is whether moral treatment ever worked. Did treating disturbed, severely mentally ill people with kindness in small orderly retreats produce good outcomes? Modern historians have concluded that it did indeed produce surprisingly good results. In the first decades of moral treatment, 35 to 80 percent of all admitted patients were discharged within a year’s time, and the majority of those discharged were viewed as having been cured. That meant that their disturbing behavior and psychotic thoughts had largely disappeared. At Pennsylvania Hospital, results remained fairly good throughout Kirkbride’s tenure. Of 8,546 “insane” patients admitted from 1841 to 1882, 45 percent were discharged as cured, and another 25 percent discharged as improved. A long-term follow-up study of 984 patients discharged from Worcester asylum from 1833 to 1846, which was conducted in the 1880s, found that 58 percent had remained well throughout their lives. Another 7 percent had relapsed but had subsequently recovered and returned to the community. Only 35 percent had become chronically ill or had died while still mentally ill.²⁷

This scandal was kept from the public, however. Meyer and the hospital board agreed to keep his findings quiet, and although Cotton stopped performing intestinal surgeries, he resumed attacking his patients’ teeth, often extracting all of them as this left “no prospect for any further trouble.” This form of therapy continued at Trenton State Hospital for twenty years. At Cotton’s death in 1933, he was widely eulogized, and Meyer publicly saluted him for having left “an extraordinary record of achievement.” Meyer’s actions—in essence, he allowed Cotton to continue to perform purposeless, mutiliating surgeries rather than expose psychiatry to a black eye—seem inexplicable until it is remembered that he was a member of the advisory board to the American Eugenics Society and had served for a year as president of the Eugenics Research Association. The sordid story of Meyer’s coverup was unearthed in 1986 by University of California sociologist Andrew Scull.

The influence of eugenic attitudes on the outcomes of the severely mentally ill is easy to trace. Throughout the 1800s, asylums regularly reported discharging up to 50 percent of their first-episode patients within twelve months of admission. For instance, in 1870, half of the patients at Worcester State Lunatic Asylum had been confined less than a year, and only 14 percent had been confined more than five years. Fairly quick recovery from an acute episode of psychosis was common. Eugenic attitudes toward the mentally ill altered that pattern of recovery. For example, a 1931 long-term study of 5,164 first-episode patients admitted to New York state hospitals between 1909 and 1911 found that over the next seventeen years, only 42 percent were ever discharged (a discharge rate reached in under one year in the 1800s). The remaining 58 percent either died in the hospital or were still confined at the end of that period. But by the 1930s, physicians had forgotten about discharge rates in the 1800s, and contemporary discharge rates convinced them that recovery from severe mental illness was rare.²⁴

Intensive electroshock was also tried on “schizophrenic” children. Starting in 1942, physicians at New York City’s Bellevue Hospital enrolled 98 children, ages four to eleven, in a study that involved shocking them twice daily for twenty days in a row. The Bellevue doctors reported that the treatment successfully made the children “less exciteable, less withdrawn, and less anxious.” A few years later, researchers at a different hospital who followed up on the children found that a number had become particularly violent and disturbed. A ten-year-old boy wanted to “kill the physicians who had treated him”; he eventually assaulted his mother for “consenting to this form of treatment” and then “attempted to jump out an apartment window.” A nine-year-old boy attempted to hang himself, explaining that he was afraid of being shocked again. Despite this follow-up study, Bellevue Hospital’s Lauretta Bender wrote in 1955 that she had successfully put a toddler, not yet three years old, through the twenty-shock ritual.⁷⁵

Moniz published his last article on lobotomy in 1937. Two years later, he was shot by a disgruntled patient; however, he recovered and continued to practice until 1944, when he retired. He died in 1955, at age eighty-one, six years after he won the Nobel Prize, and in those last years, wrote neurologist António Damásio, he was a man “obviously content with himself.”

Although Freeman and Watts told the public they altered their surgery to make it more precise, in truth they were forced to do so because the first twenty operations had, in essence, gone awry. Many of the initial patients had experienced a return of their symptoms and needed repeat operations. One patient had died from a cerebral hemorrhage, another from cardiac arrest not long after the surgery. A third patient, known as Mrs. S. in the literature, who prior to the surgery had worked for thirteen years as a secretary and was married, slid into a profoundly dilapidated state after the operation, from which she never recovered. But the public didn’t learn of Mrs. S’s miserable fate; she was one of the first six patients to be operated on and was said in the medical journals to be “moderately improved.” She was still appearing as a good outcome in the medical literature as late as 1938, two years after her operation, even though by that time she was, in truth, whiling her days away in St. Elizabeth’s Hospital in Washington, D.C., “fat, foolish and smiling.”²⁰

In 1991, researchers found that a toxin called MPTP, which had been discovered in a batch of contaminated heroin (addicts who injected it became frozen in place), closely resembled three neuroleptics: haloperidol, chlorpromazine, and thiothixene. MPTP was subsequently used to induce Parkinson’s in animals so the disease could be studied; haloperidol has been used in such studies as well.

Drug studies in schizophrenia regularly use “relapse” as an outcome measurement. However, what constitutes relapse isn’t well defined. Some investigators use rehospitalization as the criteria for relapse. Others define it simply as some degree of worsening—an increase in “psychotic” thoughts or agitated behavior. From a scientific standpoint, it’s a loose term at best.

In 1995, Peter Weiden, a psychiatrist at St. Luke’s-Roosevelt Hospital in New York City, tallied up how all this plays out in the “real” world. Eighty percent of schizophrenia patients treated with standard neuroleptics relapse within two years of hospital discharge, and the majority of the relapsers become sick again while reliably taking their medications. The American Psychiatric Association, in its 1980 diagnostic manual, even described this common downward spiral into chronic illness: “The most common course [of schizophrenia] is one of acute exacerbations with increasing residual impairments between episodes . . . A complete return to premorbid functioning is unusual, so rare, in fact, that some clinicians would question the diagnosis.” That gloomy prognosis does not fit the spectrum of outcomes natural to schizophrenia, but it does accurately describe outcomes as shaped by standard neuroleptics.³⁹

The fact that patients who abruptly go off neuroleptics may become more wildly psychotic than they otherwise ever would have been is another reason that neuroleptic use may make the mentally ill more prone to violence. Several high-profile murders in recent years have been committed by people in this drug-withdrawal state. Most recently, Newsweek reported that Andrea Yates, the Houston mother who killed her five children, did so after “she was taken off the powerful anti-psychotic drug Haldol.” However, such instances of violent murders are inevitably reported as examples of why the mentally ill need to be kept on medications, rather than as examples of the peril of using the drugs in the first place. The blame is put on the patients and their “disease,” rather than on the medications.

A review of seven other studies comparing neuroleptics to benzodiazepines, which are minor tranquilizers, adds to the questions raised by Keck’s study. In four trials there was no difference between neuroleptics and the minor tranquilizers; twice the benzodiazepines came out slightly on top; and the neuroleptics did once. See Owen Wolkowitz, “Benzodiazepines in the Treatment of Schizophrenia: A Review and Reappraisal,” American Journal of Psychiatry 148 (1991), specifically the table on p. 716 for comparative results in six trials; and William Carpenter, “Diazepam Treatment of Early Signs of Exacerbation in Schizophrenia,” American Journal of Psychiatry 156 (1999):299-303.

After the committee’s 1977 decision, Soteria researchers reapplied in 1978 for NIMH funds, and the project was revived for a few more years. But the 1977 decision effectively marked the end of Soteria as an experiment that might threaten mainstream psychiatry. The project had been so hobbled that no data gathered in the post-1975 years were published over the next twenty years, and much of the data—because of a lack of funding—wasn’t even analyzed. The blackout kept from the public this finding: John Bola, an assistant professor at the University of Southern California, recently reanalyzed all of the Soteria data, including the post-1975 data, and he determined that the superior outcomes for the Soteria group, compared to those treated conventionally with neuroleptics, “is startling. It looks better than what Mosher published.” Only 31 percent of Soteria patients who continued to avoid neuroleptics after leaving Soteria relapsed during a two-year follow-up period, compared to 68 percent of those treated conventionally with neuroleptics. (Relapse rates, however, were high for those Soteria patients who, after they left Soteria House, were placed on neuroleptics by their doctors.)

This is yet another part of the “story” we have told ourselves about neuroleptics that is easily shown to be false. Neuroleptics were introduced into mental hospitals in 1954-1955. At that time, fiscal concerns were driving states to seek alternatives to hospitalization of the mentally ill. Even so, over the next seven years the number of patients in public mental hospitals declined only slightly, from 559,000 in 1955 to 515,000 in 1962. The real emptying of the state hospitals began in 1965 with the enactment of Medicaid and Medicare laws. Those laws provided federal subsidies for nursing home care but no such subsidy for care in state mental hospitals, and so the states did the obvious economic thing: They began shipping their chronic patients to nursing homes. The number of patients in state hospitals declined by nearly 140,000 patients from 1965 to 1970, while the nursing home census rose accordingly. Then, in 1972, the federal government passed welfare legislation that provided social security income payments to the disabled. That enabled state hospitals to discharge patients to boarding homes and welfare hotels, with the federal government then stuck with picking up the cost of that care. The year after the SSI law went into effect, the population in state mental hospitals dropped 15.4 percent, the largest decrease ever. By 1980, the census in public mental hospitals in the United States had declined to 132,164. Four hundred thousand beds had been eliminated in a short fifteen years, but it was a deinstutionalization process that had been driven by fiscal concerns, and not by the arrival of neuroleptics.

In the medical literature, researchers report annual suicide rates for schizophrenics at two to five deaths per 1,000 people. In the atypical trials, the annual suicide rate for patients (on a time-adjusted basis) was close to ten per 1,000 people, or two to five times the norm. The number of patients in the research trials who committed suicide was also undoubtedly higher than thirty-six; dropout rates in the trials were quite high and many of these patients simply dropped off the researchers’ radar screens.

The Minnesota Board of Medical Practice suspended Abuzzahab’s license in 1997 for his “reckless” treatment of Endersbe and other psychiatric patients. However, Morris Goldman, associate professor of psychiatry at the University of Chicago School of Medicine, who investigated the case for the Minnesota licensing board, believes that Abuzzahab’s case raises broader questions about the integrity of commercial drugs studies. “What is the value of the data obtained in these trials?” he said, in an interview. “Abuzzahab would have the patient’s diagnosis called one thing in the regular medical chart, and then the person would be put on a drug study and the person’s diagnosis would be called something else to fit the criteria of the drug study. Then (during the study) he would say that the patients were improving, when the whole staff was saying that they were falling apart. The problem, as was seen with Abuzzahab, is that you don’t know if the data was fudged.”

None of the drug companies needed to prove their drugs were superior to standard neuroleptics in order to gain approval. They simply had to show that their experimental drugs reduced psychotic symptoms over a short period more effectively than “placebo.” This was the “efficacy” requirement. To pass the safety hurdle, the drug companies primarily had to show that their atypicals didn’t carry a high risk of death from side effects, such as cardiac problems. The drugs could cause an array of nonfatal side effects (extrapyramidal symptoms, and so on) and still gain approval. Such risks would simply have to be mentioned in warnings on the label.

The trials clearly showed that EPS was a common risk with risperidone. The reason that Janssen could claim that extrapyramidal symptoms with moderate doses of risperidone were no worse than placebo was precisely because there was no real placebo group in the trials. In the Janssen trials, about one in six “placebo” patients experienced extrapyramidal symptoms. The symptoms are a drug-withdrawal effect, and not due to the disorder. The incidence of EPS in patients who received a fairly low dose of risperidone, 6 mg., was approximately the same. Thus, Janssen could claim that its drug caused EPS no more often than placebo did, which, to a naive public, suggested that it was risk free in this regard. While it was ludicrous science, it proved to be effective marketing.

Much like the academic doctors, NAMI is also the recipient of drug money. From 1996 to 1999, drug companies gave NAMI $11.72 million for a “Campaign to End Discrimination” against the mentally ill. The two largest donors were Eli Lilly ($2.87 million) and Janssen ($2.08 million). In addition, an Eli Lilly executive was “loaned” to NAMI in 1999 and helped the advocacy group with its “strategic planning.”

It also didn’t take long for documents to surface suggesting that the teenagers recruited into the Yale study, and their families, were being misled. On December 12, 2000, the federal Office for Human Research Protections criticized the Yale investigators for using informed consent forms that “failed to include an adequate description of the reasonably foreseeable risks and discomforts.” In the consent forms, the Yale researchers had told the young adults, who were not ill, that “while the clinical goal is to help you feel better and in more control of your life, it is possible that you will feel worse. This is a risk of your clinical condition, not a risk of being in the study.” Such wording, the OHRP noted, did not “take into account ‘feeling worse’ due to olanzapine side effects.”