Almost every week, someone comes to my office with a story similar to this one: A 55-year-old woman underwent an endarterectomy, a procedure that involves opening the arteries in the neck to improve blood flow to the brain. She had had a history of headaches that didn't respond to aspirin. During the operation, she suffered a stroke -- a piece of plaque broke off from the wall of the artery and plugged up the vessel completely, causing brain damage that left her paralyzed on one side.

Her husband went to a medical library afterward and learned that strokes can happen as often as 10 percent of the time in this operation, particularly when the vessel being manipulated is badly blocked to begin with.

"We had no idea," her husband told me. "Before the surgery, the doctor just told us that he'd never had any problems in all the times he'd done it. If we'd only known . . ."

His wife's headaches had been an inconvenience. The stroke changed their lives. Devastated, the couple wanted to know if they should sue the surgeon for malpractice.

I advised against it. The case would be expensive and hard to prove, and I thought the chance of winning was poor because it would pit the credibility of the couple against that of the better-educated and more articulate physician.

Their situation is common. Every day, surgical procedures go awry. Heart attacks, strokes, death, nerve damage, severe infection and many other events can and do occur, not out of carelessness by surgeons but because of inherent risk. These patients are victims of fate but also of a lack of communication with their doctors. Their refrain, "I had no idea . . . if I'd only known," is heard all too often.

The legal doctrine of "informed consent" is supposed to protect these patients. It seldom does -- at least after the fact.

In theory, the doctrine is simple: doctors are supposed to give a reasonable explanation of the benefits and risks of proposed treatments, whether surgical or medical. More important, they should describe the alternative forms of treatment available. The patient then gives "informed consent" to the therapy proposed or an alternative.

In practice, however, the cherished notion that everyone has a right to decide what happens to their bodies is often ignored.

The medical profession was long hostile to the idea of patients playing a meaningful role in making important decisions about treatment. Doctors talked of "terrified consent," insisting that to protect themselves from lawsuits, they had to recite a grim litany of all the calamities that could happen in the course of treatment, no matter how remote. They also asserted that those who did not possess a medical degree could not possibly understand the various options.

Trust us, they said.

Under pressure from the consumer movement, that attitude has begun to change.

Women with breast cancer -- who as recently as 10 years ago underwent radical mastectomy with no discussion of options available then, such as removal only of the lump, followed by chemotherapy or radiation -- now receive better information.

Still, particularly when the disease has not attained public prominence, it is distressingly common to meet patients who were in the dark until it was too late.

The President's Bioethics Commission found in a 1982 report that 72 percent of doctors surveyed consciously withheld information about risks and alternatives from their patients at least once a week. The commission had its staff members watch the communications process in various hospitals and concluded:

"In hospital settings, often little or nothing is actually discussed with patients regarding either alternative treatments or the recommended treatment. Instead, physicians commonly make decisions and proceed to treat the patient."

There has never been a greater need for patients to participate in deciding how they are treated. For instance, a recent Rand Corporation study of endarterectomy and two other operations -- coronary angiography and gastrointestinal endoscopy -- found "substantial disagreement" among expert physicians about when it is appropriate to perform these invasive procedures. The disagreement reflects uncertainty among doctors about how beneficial the procedures are when measured against the risks, alternatives and uncertainties that patients deserve to be informed of.

In this age of multimillion-dollar malpractice verdicts, many people are surprised to learn that the legal system can't do much for victims of uninformed consent.

It is certainly not for lack of judicial pronouncements on the subject. In the District, most lawyers and many lay people have heard of the case of Canterbury v. Spence. The judges of the U.S. Court of Appeals granted a new trial to a patient who lost the first malpractice trial against a doctor who performed back surgery and who, he claimed, did not inform him of the risk of paralysis.

The court held that the jury was not properly instructed about the doctrine of informed consent, and it explained the doctrine in a lengthy opinion by former Chief Judge Spottswood Robinson.

What few people know about the Canterbury case is that when the case was retried, the patient lost again. The patient had to prove that if he had been told of the risk of the procedure, he wouldn't have consented. The jury didn't buy it.

One might think that any group of lay jurors would be sympathetic to testimony that a doctor pressured a patient into a dangerous or unnecessary operation by misinformation or lack of information. But juries are skeptical of informed-consent claims, for a variety of reasons.

As common as it is to consent to major surgery knowing little about the risks and alternatives, confessing to ignorance in court often makes patients look monumentally stupid.

Nor is it easy for juries to sort out the "swearing contests" that often develop between a patient who says he or she was told nothing and a doctor who says otherwise. The doctor is often better educated than the patient, more glib, and can point to evidence such as his notes in the patient's file that say "discussed risks and benefits," and to the patient's signature on the surgical-consent form.

When a patient testifies he wouldn't have had the procedure if he'd known the risk, this often smacks of after-the-fact rationalization, particularly if there was a good medical reason to do the surgery.

Indeed, the requirements of legal proof often reduce the concept of informed consent to a legalistic parody: it does not matter that the consent form was a full page of gray type and was presented by a nurse the night before surgery when the patient was woozy with drugs.

What matters in court is the patient's signature just below the line, "I have been fully informed of the risks and benefits of this procedure and consent to it."Genuine informed consent, of course, is not about forms. It is not about laundry lists of potential horrors.

True informed consent begins with the idea of mutual respect: the patient's respect for the doctor's skill, knowledge and dedication, and the doctor's respect for the patient's ability to comprehend information about his or her condition and share in the process of deciding what to do in the face of uncertainty.

Patients have to push doctors to get this dialogue started. By the time they get to the courtroom, it is usually too late to begin pushing. It has to start in the doctor's office.Patrick A. Malone is an attorney with the firm of Stein, Mitchell & Mezines.