"Something bent down and took hold of me and shook me like the end of the world," wrote Sylvia Plath in The Bell Jar. "Whee-ee-ee-ee-ee, it shrilled, through an air crackling with blue light, and with each flash a great jolt drubbed me until I thought my bones would break and the sap fly out of me like a split plant."

That's what shock therapy used to be. Today, anesthesia, muscle relaxants, re-engineered machines and repositioned electrodes buffer the jolt, and insights about the biochemistry of depression ease the stigma. As a result, electroconvulsion therapy (ECT) is being used more readily and more successfully, even on older adults once considered high risk. Not only does ECT work when drugs can't, but it efficiently short-circuits psychiatric-hospital stays, making it a natural in a cost-conscious managed-care climate.

The improvements are dramatic -- but not everything has changed. Patients still complain of permanent, significant memory loss, and psychiatrists still say they're wrong. Researchers still don't know exactly how ECT works or what it might do to a child's developing brain. Ethicists still don't know how to tell if a patient's consent was really "informed." And the Food and Drug Administration (FDA) still hasn't asked for clinical trials showing that individual ECT machines are safe.

Used to people recoiling at the very notion of electrically induced brain seizures, ECT experts tend to minimize concerns, afraid another swing of the pendulum could deprive patients of their last alternative. But after struggling so valiantly to reclaim ECT's reputation, psychiatrists' own memories may have grown a little spotty when it comes to the need for caution.

Ugo Cerletti and Lucio Bini first used shock therapy in Italy in 1938, convinced the body manufactured a "vitalizing substance" during an epileptic convulsion. They watched breathlessly, Cerletti wrote later, "overwhelmed during the apnea as we watched the cadaverous cyanosis of the patient's face." No substance oozed, but the shock did snap the man into lucidity.

In the bedlam of the '40s and '50s, shock was the only tool psychiatrists had, and they administered it with abandon. The typical recipient was a 40-ish male, impoverished and rebellious, living in a public institution. By the end of the '50s, enough asylums had zapped patients with a rainbow of disorders into monochrome docility, and public outrage grounded ECT. Just in time, miraculous new anti-psychotic and anti-depressant medications took over.

It took another few decades to realize the drugs' less than miraculous side effects -- cardiac risks and mobility disorders -- and the plain fact that they didn't help everyone. Meanwhile, ECT had been significantly improved. Learning that seizure thresholds varied enormously, psychiatrists found they needed only a fraction of the standard voltage for many women and young people. Sedatives and muscle relaxants removed the pain, the need for restraints, the physical danger of broken limbs and the horrific twitching spasms. Patients were thoroughly monitored, and physical resistance (some people do have thicker skulls) was measured ahead of time to avoid electrical burns.

When Malcolm Bliss closed, Dr. John Csernansky, medical director of the Metropolitan Psychiatric Center that replaced it, retrieved a little black box from the basement -- a shock machine from the 1940s. Now, holding it in one hand, he presses its silver button urgently, demonstrating how the psychiatrist controlled the electrical pulse's duration. Crudely measured, those pulses lasted one or two seconds apiece, and continued as long as the doctor chose. In today's machines, the electricity is produced from a different kind of wave, the duration of each shock is computerized, each pulse lasts perhaps 0.5 milliseconds, and the entire train of impulses cannot last longer than eight seconds.

An American Psychiatric Association (APA) task force issued guidelines in 1990 suggesting that hospitals "insulate the waiting patient from auditory and visual contact with the treatment and recovery areas" -- so he no longer has to watch peers emerge from treatment, their eyes deadened. The guidelines also note that "immediate side effects from ECT are rare except for headaches, muscle ache or soreness, nausea and confusion."

Compared to the old days, it's a picnic.

Almost all St. Louis hospitals provide ECT treatment. "We do maybe 10 a day," says one anesthesiologist. "Everybody's doing them. It's very well accepted in the medical community." St. John's Mercy Hospital counted more than 2,000 treatments given in the last fiscal year (compared to 1,500 five years ago). Dr. Donald Hay, director of the mood-disorders program at St. Louis University, says they do 10 treatments a week on average. BJC Inc., known as the local ECT headquarters, declined to provide a count. But even satisfied patients call Barnes "Jiffy Jolt" because it speeds so many of them through its ECT assembly line.

The nickname isn't surprising: Harvard researchers announced last year that ECT's popularity increases around academic medical centers. Universities are supposed to take the lead, after all, and depression is the disease of the decade. In 1990, the National Institute of Mental Health estimated a startling 9.5 percent of American adults (17.5 million people) suffering from depressive disorders. Now that we know depression is often biological, there's less shame, more determination to get relief. And after successful ECT, improvement's intuitively obvious.

"Somehow I just knew that I was better," people say, "and that I did not have to be in the hospital anymore." Major depression isn't the blues, they add; it's hell, eroding love, joy, energy, will, clarity and the very instinct for life. When Anne Simpson (a pseudonym) sank into deep depression, "it went on for months, and the part of me that was willing to continue on felt as though it were getting smaller and smaller." Doctors tried a buffet line of drugs; none worked. But after her third course of ECT, Simpson renewed friendships, changed jobs, re-entered the world. Fran Scott has had ECT twice in her life, when manic-depression hit a wall no drug could penetrate, and she'd do it again. "A lot of people just go in the hospital and sit there while doctors play games with drugs," she explains, "and they are still miserable. And it's scary to be depressed."

Scientists still don't know exactly why ECT works -- except that it has less to do with electrical jolts than with the seizure they induce. Electricity runs through our brains already; that's what causes the neurons to fire and discharge neurotransmitters, which then carry the impulse across the great divide to the next cell. "Brain cells are set up in oscillating circuits that are firing regularly," explains Csernansky. "What you are trying to do when you induce a seizure is get them all to fire in synchrony."