AS THE TMAP PANEL MEMBERS worked on the protocols, drug companies aggressively promoted the new antipsychotics across the psychiatric landscape. Their key selling point: that they were more effective and caused fewer serious side effects than the older antipsychotics, especially Haldol, the most widely used. Though it did approve six atypicals, the FDA was dubious of some of these claims. "We would consider any advertisement or promotional labeling for Risperdal false, misleading or lacking fair balance… if there is a presentation of data that conveys the impression that [Risperdal] is superior to [Haldol] or any other marketed antipsychotic drug product with regard to safety or effectiveness," an FDA official wrote in a 1993 letter to Janssen Pharmaceutica. But the letter was only made public years later, when journalist Robert Whitaker quoted it in his 2002 book, Mad in America. Most prescribing doctors were left in the dark. (For more on how drug companies manipulated clinical trials for atypicals see motherjones.com/spinningdoctors.)
The largest study to date, a review of 52 clinical trials including more than 12,000 patients published in the British Medical Journal in 2000, found "no clear evidence that atypical antipsychotics are more effective or better tolerated than conventional antipsychotics." A 2003 study comparing Zyprexa, the top-selling atypical antipsychotic, and Haldol, published in the Journal of the American Medical Association, found the new drug "does not demonstrate advantages compared with [Haldol]… in compliance, symptoms… or overall quality of life."
The new drugs now appear to be associated with higher suicide rates and to cause tardive dyskinesia, too, though perhaps at lower rates than the first-generation drugs. They can cause rapid weight gain and thus an increased risk of diabetes. In September 2003, the FDA required the makers of all atypicals to add to their labels a warning that the drugs can cause hyperglycemia, diabetes, and even death. Janssen was also made to send doctors a letter conceding it had misled them when it said that Risperdal does not increase the risk of diabetes. In fact, the company had to admit, it probably does.
When TMAP's schizophrenia algorithm was finalized in 1997, however, it did exactly what industry representatives must have hoped for: It called for the newest, most expensive drugs—five atypicals—to be used first. If a patient does not respond well to one of those drugs, a second member of this group should be tried. If that drug also fails, a third drug should be tried, this time either another atypical or an older antipsychotic. The guidelines for major depression and bipolar disorder similarly favor new drugs.
"When [the drug companies] saw the newer medications were there, they liked that, of course," says Shon. "I know that has raised questions in people's minds: 'Why are the newest, most expensive first?' Well, the newest, most expensive are either the most efficacious and/or the safest."
But that assertion is increasingly disputed. "When atypicals came out, they looked a little better in effectiveness and a lot better in terms of side effects," says Mike Hogan, Ohio's mental health director and former chairman of President Bush's New Freedom Commission on Mental Health. "These days, they look perhaps a tiny bit better in terms of effectiveness, but increasingly it's not clear whether the side-effect profile is better or just different."
Ohio adopted a TMAP-like algorithm in 2001 but with a critical difference. According to Hogan, it's merely a guideline for prescribing doctors to consider. But in Texas, state officials put far more pressure on its physicians to follow the protocols. Under regulations codified by the legislature in 1999, doctors in state-owned and state-funded mental health entities must follow the algorithm, or justify a different course with a note in a patient's file—a hurdle that sends the message that such deviation should be the rare exception.
As the TMAP guidelines began to be adopted in 1997, Texas Medicaid spending on the five atypical antipsychotics skyrocketed from $28 million to $177 million in 2004.
MANY DOSES OF THESE DRUGS went to patients like Amanda. She was one of 19,404 Texas teenagers prescribed an antipsychotic in July or August of 2004 through a publicly funded program, according to ACS-Heritage, a medical consulting firm hired by Texas to investigate the use of psychotropic drugs on children. Nearly 98 percent were atypical antipsychotics—unapproved for children and prescribed "off-label," a controversial practice in which doctors legally prescribe FDA-cleared drugs to patients, such as children, or for conditions, such as depression, for which they are not approved. The report found that more than half of the doses for antipsychotics appeared inappropriately high, that almost half did not appear to have valid diagnoses warranting their use, and that one-third of child patients were on two or more medications.
When she was transferred from Austin State Hospital to a residential facility on March 18, 2004, Amanda was on five different medications, putting her on the extreme end of a growing practice known as polypharmacy that worries many doctors. "This is a complicated regimen using powerful psychotropic medications in a barely adolescent girl, so I would be quite concerned about it," says Dr. Joseph Woolston, a Yale University professor and chief of child psychiatry at Yale-New Haven Hospital. "It isn't grossly, acutely dangerous, but it is sedating and would make it difficult for a child to experience the world in a normal way. If you or I were on that regimen we would have a lot of trouble attending to work or school. We don't have any idea what that combination of medications does to a developing child. It may have a number of long-term side effects." He also suspects that the drugs may have been used as much to control the angry reactions of a girl who was hospitalized against her will as to treat any mental and emotional problems.
Dr. Clifford Moy, clinical director of Austin State Hospital, says that while the hospital's philosophy is to avoid using more than one member of any particular class of psychiatric medication, using multiple drugs from different classes is often the best way to treat a patient with multiple symptoms. While declining, for privacy reasons, to discuss Amanda's treatment, he said medication and restraint would never be used for punitive purposes or merely to promote compliance with hospital rules, but only in the case of a "significant emergency behavioral situation." He added that forced injection of an antipsychotic—which happened to Amanda several times—might be used "if there were a legal consent for an oral antipsychotic medication, which the patient refused." Such consent was apparently provided, in Amanda's case, by the Department of Protective and Regulatory Services.
The 46-bed child and adolescent wing where Amanda stayed was not, like the rest of Austin State Hospital, obligated to follow TMAP. Its treatment regimens were influenced more by CMAP, the children's algorithm not yet mandated by the legislature. CMAP steers clear of providing protocols for schizophrenia and bipolar disorder—the disorders that atypicals were designed to address—in part, says DSHS's Doug McBride, because there's "little scientific evidence" as to what the appropriate regimen for kids would be. CMAP does, however, call for combining atypicals with antidepressants for children diagnosed—as Amanda was—as suffering from depression "with psychotic features." McBride defends such off-label use of prescription drugs, saying that the FDA approval process "is not the end of clinical and other scientific evidence on the use of that medication."
Beyond their technical dictates, the algorithms established a culture that affected which medications were prescribed. Steven Shon, who, along with his colleagues, had led training sessions for the staff of Austin State Hospital, argues that the algorithms were designed to prevent irrational and excessive medication. Yale's Woolston agrees with the goal, though not necessarily the reality. "Algorithms are supposed to cut down on people using medications inappropriately and to take into account medication interaction," he says. "Where they become a problem is when people use them as a mandate, forget their own clinical judgment, and believe that when you're in doubt, you're supposed to move forward in the algorithm and add more medication."
Medications can be invaluable, and some patients say their lives have been transformed by atypicals. But algorithms reinforce the perception in both psychiatry and popular culture that mental problems always require drug treatment. "An algorithm may put blinders on a psychiatrist and create the presumption that the only clinical approach to problems is to use medications," Woolston says. If a patient doesn't respond to a particular medication, a doctor relying on an algorithm may think they need to use or add a different medication, he says. "But sometimes, the best approach is to say, 'Medication isn't working; let's try something else.'"