(The New York Times News Service) -- Most mornings, Karl Ackerman felt disappointed that, once again, he was alive.

A depression that began in 1991 had left him feeling so hopeless that, for more than a decade, he could scarcely leave the house, let alone find a job: "If you do not believe you have any worthwhile qualities, try going for a job interview," he said. Eventually, the weight of Ackerman's despondency ruined his marriage, too.

But nothing doctors tried seemed to help -- not the antidepressants, not the antipsychotics, not the two types of drugs together. For nearly 12 years, Ackerman relied on talk therapy to keep him from becoming suicidal -- until he finally found a drug two years ago that brought him relief.

"The goal of medicine is to cure things. Unfortunately, that is a work in progress," said Ackerman, now 59 and president of the Manic-Depressive and Depressive Association of Boston.

More than 50 years after psychiatrists began widely dispensing drugs to treat mental illness, the profession is coming face to face with a humbling reality: Its treatments often fail, leaving millions of patients like Ackerman to suffer while doctors search for something that works.

Though more and better drugs and therapies are available today than ever before, psychiatrists still rely on time-consuming trial and error in deciding how to use them.

Over the last six months, three unprecedented government-funded studies of the most common mental illnesses - depression, bipolar disorder, and schizophrenia -- have underscored just how far modern psychiatry is from "curing" mental illness.

The studies -- large, multipart tests of the best available treatments -- show that drugs and therapy bring almost no relief to some patients, while many more achieve only temporary or partial recovery.

In the most recent study, released in February, researchers found that some combination of drugs and therapy helped nearly 60 percent of bipolar disorder patients recover from either deep depression or bouts of mania.

However, within two years, nearly half of the recovered patients had wide mood swings again, usually slipping into depression.

"I don't think we should be satisfied, absolutely not," said Dr. Roy Perlis of Massachusetts General Hospital, lead author of the first phase of the bipolar disorder research.

But the three studies, which collectively cost more than $80 million, are also giving psychologists detailed information to help improve treatment for conditions that afflict 24 million Americans, including more than 18 million with major depression each year. No single drug or therapy will help everyone -- or even the majority -- but the studies are large enough to identify what types of people benefit most from each treatment, potentially taking some of the guesswork out of psychiatry.

For instance, among people with chronic depression, well-educated women with few other medical complications seem to get the most value from the antidepressant Celexa. Likewise, patients such as Ackerman, who suffer from treatment-resistant bipolar disorder, seem to improve significantly when they take a mood stabilizer called Lamictal, the drug that ended Ackerman's long period of darkness.

"No one ever felt we could cure mental illness" with current treatments, said Dr. Robert Freedman, editor of the American Journal of Psychiatry, which published the initial results of the studies. "These are groundbreaking studies for our profession. They clearly set the bar much higher, and they also are sufficiently powerful to begin to get meaningful guidance for clinicians."

Unlike drug company-funded studies, which sometimes produce falsely encouraging results by focusing on short-term recoveries and avoiding tough patients, these National Institute of Mental Health studies followed patients for years and included the most treatment-resistant cases.

Moreover, the patients were treated by their normal doctors rather than at central locations, more closely simulating real-world medicine than most clinical trials.

The result, say psychiatrists, is the most realistic look ever at the state of psychiatric medicine. The picture is not always pretty.

The study of bipolar disorder, the most treatable of the three illnesses, was actually the most upbeat of the three despite the high relapse rate.

"There is good reason to be hopeful, but there is lots of work to be done," said Mass. General's Perlis, noting that the US Food and Drug Administration has approved several new drugs for the disorder in recent years.

By contrast, nearly three-quarters of the schizophrenia patients stopped taking the drug that was originally assigned to them in the 18-month experiment, mainly because the medications didn't improve their condition enough or they couldn't stand side effects such as weight gain or nervous tics. To the researchers' surprise, a little-used older drug, Trilafon, worked as well as most of the newer drugs at a fraction of the cost.

"These results should cry out for ... a whole new generation of drugs to treat schizophrenia," Dr. Thomas Insel, the NIMH director, declared at a press briefing last September to announce the results, the first phase of the schizophrenia study, known by its acronym, CATIE.

In the study of 4,000 chronically depressed people, only 30 percent got complete relief from taking an antidepressant similar to Prozac, and half didn't improve much at all. "Unfortunately, doing miserably is common," said one doctor involved in the depression project, called STAR..D.

The researchers concluded that Prozac-like drugs by themselves simply aren't enough to end depression for many patients.

Like the other two studies, the researchers did not give up on patients who did not immediately recover. Future phases of the depression study will look at treatments for the 70 percent who did not respond to Celexa, including other drugs as well as cognitive behavioral therapy, which helps patients manage their condition through knowledge and self-insight.

Psychiatry has made enormous progress since the early 1950s, when tens of thousands of mentally ill people were locked in institutions and sometimes subjected to crude, poorly understood treatments. The pace of drug development has quickened dramatically in the last 20 years from FDA approval of Prozac in 1987 to a new generation of antipsychotic drugs that treat schizophrenia without the risk of neurological damage posed by earlier drugs.

But, until now, doctors have had little guidance on how best to use the burgeoning number of treatment options. That can frustrate patients as they go through one treatment after another while in the grips of deep psychological misery. Even if there is an effective treatment, they can be forced to wait months or years for relief.

"If you're the person going through it, all you want is relief, and it's really disheartening to be told you've got to stay on something for eight to 12 weeks whether it works or not," said Ackerman.

Federal officials hope to offer, when the studies are finished, almost a guidebook for doctors on when and how to best use available treatments and how to know when a patient is truly recovered. In the bipolar disorder study, for instance, researchers found that the patients most likely to relapse were the ones who never shook all the symptoms of depression.

Dr. Gary S. Sachs of Mass. General, lead researcher on the bipolar disorder project, said he saw improvements in the outcomes for patients after the psychiatrists in the study underwent training to help them make the best drug and therapy choices.

"We are not curing people, but we are actually able to manage them in a way that appears to be reasonably effective," he said. "Is the glass half full or half empty?"

Copyright 2006 The New York Times News Service. All rights reserved.