Madness in Fine Print

By James Willwerth/Los Angeles

Harry Cummings remembers hearing voices and seeing images the very day he was discharged from the Army: "Voices told me all kinds of stuff about the government. It scared me." Cummings, 38, had done two Army tours, including time in South Korea, and had worked his way up to sergeant. After his discharge, he began wandering, hitchhiking from Seattle to Miami, living on the streets, getting by on odd jobs, before stumbling into Los Angeles in the summer of 1982. "A black guy gave me a ride from the California border and dropped me downtown. He gave me a few bucks and encouraged me to get help. I had only the clothes on my back, and I slept outside that night." The next day he talked to a cop and eventually was driven to the Veterans Affairs treatment center in West Los Angeles.

At the center, he told the doctor about the voices, the images. "I was mixed up, dirty, wired up and spacy," Cummings relates. "He said he wanted me to sign this paper for experimental research. He told me I didn't have an obligation to be in it. He said I could quit if I wanted, but he wouldn't advise that. He said the drugs they would be giving were completely harmless. He said the hospital would kick me out on the streets if I didn't participate. That was the word he used: kick."

Cummings signed, was formally diagnosed as paranoid schizophrenic, and says he spent the next five years in an experiment with a powerful antipsychotic drug. He suffered bad side effects: partial blindness, impotence, constant migraines. He says the researchers at the VA never allowed him to see an eye doctor and wouldn't let him change drugs. Only in 1987, after complaining for years, did Cummings finally manage to get out of the experiment and see other doctors. "My vision came back but not as good as I expected," he says. The VA last week said a "cursory glance" did not turn up Cummings' records.

Most of the country's 2.3 million VA patients are poor or mentally ill. And over the decades, many have signed up for experiments after doctors suggested that it was the only way they could receive meaningful help. Now, however, those methods of obtaining recruits for psychiatric experiments are undergoing a radical change, one that may transform the way schizophrenia is studied in years to come. This summer the National Institutes of Health rebuked the University of California, Los Angeles for serious "deficiencies" in setting up schizophrenia experiments that UCLA runs at the Veterans Affairs Medical Center in West Los Angeles. A UCLA experiment reported in TIME last year -- in which one patient suffered a severe psychotic breakdown and another committed suicide -- was formally sanctioned and is now said to be undergoing an ethics review by the American Psychological Association.

Other psychiatric studies around the country are coming under increasing scrutiny, raising fresh examples in the ongoing debate over experimental ethics. How can scientists be held accountable for harm to an experimental subject? Must individual rights always supersede the quest for knowledge? "This is an issue that has been around since the Nazi experiments," says Susan Knapp, the APA's director of publications. "If the conditions for a research project were unethical but the science is good, what do you do with it?"

Ten different UCLA-VA experiments, all of which are federally funded and housed in the West Los Angeles center, are under investigation. In one, "Management Risk for Relapse in Schizophrenia," veterans were placed on lower than usual dosages of antipsychotic drugs. All lived away from hospital supervision and were said to be monitored weekly. However, the experiment was envisioned to follow patients with "continuing psychotic symptoms" and especially sought out those in "relative remission" with "a potential for demonstrating a relapse." In a letter to UCLA in April, the NIH's Office for Protection from Research Risk demanded verification that human subjects were no longer involved in that experiment. It has been discontinued.

The OPRR is also investigating an experiment at the Bronx Veterans Affairs Medical Center, which is affiliated with the prestigious Mount Sinai Hospital in New York City. The study gave patients L-dopa, an amino acid that stimulates dopamine -- the brain's mood-regulating chemical messenger -- in order to observe psychotic breakdowns, allegedly without advising them on consent forms of the extreme discomfort and high risk of the undertaking. When Vera Hassner, an advocate for the mentally ill, complained to the project director, she received a letter that stated, "Patients may experience symptom aggravation . . . It would not be advisable to talk to patients about psychosis or relapse . . . Talking to patients about psychosis or schizophrenia might cause unnecessary anxiety . . ." At a recent congressional hearing, Dr. Jay Katz, a bioethicist at Yale University, said, "It is not the patient who may be anxious but the researcher, who has to look the patient straight in the eye and say, 'I want you to participate in an experiment in which your old symptoms may recur.' "

The focus of the debate is "informed consent." How should consent documents be presented to protect the rights of the mentally ill? "People with schizophrenia have information-processing deficits. They often have trouble absorbing information," says Dr. Robert Liberman, head of UCLA's Clinical Research Center, who presides over all of the school's schizophrenia projects, including those under scrutiny at the VA in West Los Angeles. He contends that reforms that demand clear and more detailed risk information will create a chokehold of paperwork. Liberman argues that patients can be informed of risks in conversation, without resorting to intricate documents. Critics reply, however, that verbal consent protects no one save the researcher. And while mentally ill patients may not fully understand consent forms, full disclosure in writing can at least be comprehended by family members, guardians or court-appointed conservators.

Detailed consent forms are also the only way a prospective patient can weigh the risks of a number of controversial experimental methods common in schizophrenia studies, among them drug discontinuation, which takes severely mentally ill patients off medication to study their potential for psychotic relapse.

Faced with a cutoff of federal funds by OPRR, UCLA scientists are reviewing their consent forms to comply with the new climate of reform. OPRR required independent monitoring of research and safety data from the school's Clinical Research Center. "It says the government doesn't trust them to be competent on their own," says a prominent researcher in the Midwest. An OPRR report last July pointed out "numerous deficiencies" in consent forms and criticized experiments for failing to give full descriptions of risk and explanations of how patients would be compensated and treated if injuries occur. In August the acting director of the National Institute of Mental Health, Dr. Rex Cowdry, mailed a copy of a previous UCLA reprimand to the country's top 100 schizophrenia researchers. It was a warning that sent a chill through the profession.

Liberman says he recently distributed an "eight-page, single-spaced" consent form to UCLA's Human Subjects Protection Committee for approval. "In the past I would have been criticized for information overload." No longer. "Now," he says, "they want every detail that previously would have been given verbally. OPRR is saying that every detail has to be written down." UCLA's experiments are now directly overseen by the OPRR offices in Washington, but, says OPRR chief Dr. Gary Ellis, "they're reporting to us very thoroughly."

The changes have made it more difficult to get an experiment started. UCLA psychology professor Irving Maltzman wanted to start up a diet-and-exercise program for mentally retarded children. The school's review board, he says, "asked me to appear in person before the entire committee. I've never had to do that before. They wanted to know if mentally retarded kids will understand the consent form. They wanted the signing witnessed by a disinterested third party. They asked if I'd be willing to have the kids interviewed by an independent board of psychiatrists."

Nancy Andreasen, editor of the American Journal of Psychiatry, says this fall she began sending manuscripts back if they did not contain details about how informed consent was obtained. How then to handle experimental findings? Says George Annas, a bioethicist at Boston University: "The question is whether the science is any good. If not, you don't have an issue." But, he adds, "suppose you discover a cure for cancer by killing 100 people without their informed consent. Are you going to publish that?" His conclusion: "You publish it with an editorial saying this was wrong. It's bad ethics, but the knowledge is too important."