We're sitting in a room that's maybe 9 feet wide by 20 feet long, lined with countertops that hold 12 TV monitors. The screens let us spy upon what's normally one of the most private of encounters: the interaction between doctor and patient, taking place in a dozen nearby examining rooms. In three of the rooms, the patient is talking about a worrisome bout of diarrhea. Patients in another three rooms are complaining about their chronic insomnia. Three more are seeking medical advice about shortness of breath, while the final trio (all young women) have been coughing up blood. The white-coated figures in the exam rooms are asking questions. Some take notes. Almost all pull out stethoscopes and listen to hearts and lungs and stomach gurgles. But none of the White Coats is licensed to practice medicine. They're third-year medical students being tested on how well they deal with patients. And none of the "patients" is actually ailing. They're actors who've been trained to serve as both the students' test and their grader.

Almost unheard of 25 years ago, such actors are now a fixture at every medical school in the United States, according to Peggy Wallace. She heads the program that hires "standardized patients" (as the actors are known) on the UCSD campus. Wallace says med schools have embraced this approach in part because two years ago the board that licenses U.S. doctors began requiring candidates to pass a day-long evaluation of their clinical skills. In that grueling marathon, the aspiring doctor sees a dozen standardized patients who might complain of anything from dizziness to depression.

The licensing board's insistence that would-be MDs demonstrate interpersonal savvy with patients reflects a sea change. Wallace says doctors once thought that a good bedside manner was something one was born with. "They thought you either had it or you didn't." But research over the past two decades has shown that winning patients' trust is "a teachable skill," she asserts. It's not one most people learn in the course of their normal social interactions because "the interaction of a doctor with a patient is not a normal social interaction," Wallace points out. In a normal social setting, you don't let anyone examine your body. You'd be startled and affronted if another person asked about the color of your feces or how many people you were having sex with. But doctors do. Practicing on simulated patients can help them learn to appear both professional and caring, according to the current thinking.

The UCSD School of Medicine's use of standardized patients seems designed to squeeze every iota of insight from the contrived encounters. Consider the third-year students' midterm exam that I observed from the room with the TV screens. Each of the 12 students that afternoon saw one patient who was short-winded, one with diarrhea, one who was sleepless, and one coughing up blood. Although three different actors played each type of patient, each of the three was depicting the same person (based on a real case), and he or she had learned that patient's personal and medical history in detail.A number of different problems and/or diseases might be causing each symptom, the student doctors knew. But their grades for the midterm would not depend on whether they came up with the right diagnoses. The point instead, Wallace explained, was to assess their clinical skills -- the tools they would need to arrive at correct diagnoses time after time. They would have to examine the relevant parts of each patient's body. To know what those were, they would need to interview the patients about their current complaints as well as their medical and family histories. To get the most out of the interview, "There's a whole series of things we teach the students," Wallace says. "For instance, you want to ask open-ended questions at the beginning. If they're coming in with chest pain, you might say, 'Tell me about your chest pain.' You might say something empathetic like, 'That must have been quite frightening for you. Tell me about it.' You want to give the patient an opportunity to say what they're experiencing. What that does is to give the patient a sense that the doctor cares about more than just what the doctor needs. And that builds trust. 'This guy cares about me. I can ask him the question I'm most worried about, which is, "Is my headache a brain tumor?" ' My satisfaction with this encounter is going to be different if I can't get that question out."

In the midterm, every time a student doctor said good-bye and left the exam room, the actor/ patient hurried to a computer and filled out a 30- to 40-question form assessing everything from whether the physical exam included all the requisite maneuvers to whether the med student "listened actively, [paying] attention to both my verbal and nonverbal cues; used facial expressions/ body language to express encouragement; avoided interruptions; asked questions to make sure s/he understood what I said." The student's grade for the exam would be based on this checklist. But Wallace makes it clear that grading the students is only part of the exercise.

The midterm, which features different cases every year, was also designed to be a learning experience, so while the patients were filling out their questionnaire, the students were doing the same thing -- enabling comparison of the two perspectives. Then the students came back into the exam rooms and spent ten minutes talking to the actors (no longer in character). "How did you think it went?" the actors began that interaction. They ended the feedback session by asking what the student would take away from the experience. Wallace said the hope was that the students would go into the next exam room and practice working on the skill they had just identified as needing improvement. "What we know from the research is that you have to practice it immediately or you lose it," she told me. "And it takes about three months of using something before it becomes part of your behavior."

Later, there would be more debriefing. In a group with their professors, the students would review all four of the cases. They would compare what they'd learned about the patients and how they came to their conclusions. They'd brainstorm about ways they could have fine-tuned the interactions. After that they would study videotapes of their performance.

Although the med students may be the stars, the actors were getting closer attention from Rob MacAulay during the recent encounter. One of two UCSD staffers who hire and train the standardized patients, MacAulay explained that part of his job is to watch the exam scenes while filling out the same checklist the actor/patient he's observing must complete. This helps to ensure that the actor/patient is grading the students fairly, he explained. And also, "With a lot of actors, if you leave them unchecked, they get comfortable, and sometimes their performance drifts."

MacAulay reminded me of a young and handsome Grandpa Who in the Old Globe Theatre's annual production of How the Grinch Stole Christmas. That's probably because he filled that role for six years, between 1998 and 2003. He'd moved to San Diego from Toronto to take the part, and he later performed in the Globe's production of The Full Monty, as well as various roles with the North Coast Repertory and Sledgehammer theaters. He also earned money working for UCSD medical school as a standardized patient (something he'd done in Canada). But he eventually decided being a professional actor in San Diego was "not an option," he told me. Whereas much of his income in Canada had come from appearing in national commercials, most commercial work in San Diego is low paid and local, he learned. "So I was faced with the reality that as much as I loved acting, as much as there's amazing, great theater to be done in this city, it doesn't pay on the scale of the Globe, which is what you need to make a living here."

MacAulay thought about returning to the work he'd done before becoming a professional actor. "I was a medical underwriter in the life-insurance industry." Underwriters interpret the information people supply when they apply for insurance, determining what premium they should pay. "Your job all day long is to look over doctor's reports," MacAulay says. He found there were no underwriting jobs available in Southern California ("because it's only at the head offices that that's done"), but one day he heard that UCSD's standardized-patient program was looking for a trainer. "I thought to myself, 'If this job doesn't have my name on it, no job does.' It was the perfect combination of medical and acting work."

He started in his current position in 2004, and in 2005, "We offered around 130 contracts," he told me. "We actually hire more actors than most of the theaters. Some of those contracts are just for one day. Others are for three months. But we offer a lot of work." The standardized patients' pay -- between $15 and $25 an hour -- is also more than what most actors in non-equity theaters make, according to the trainer.

MacAulay says the department recruits the performers by advertising with the Actors Alliance, the local "association of actors that everyone goes to if they want to know what auditions are going on. We have a really great relationship with them." Word of mouth brings others in to audition, he says, adding, "We don't just hire actors. Actors tend to come with some preparation. You know that when they're in that room, they're not going to have a lot of anxiety. But we have a lot of nonactors who do well too." Certain professions excel at the work, MacAulay says. Social workers, teachers, policemen, and firemen all stand out as prospects. "Teachers know how to communicate," he says. Cops recall details.

Still, he says the work is so demanding it's a constant challenge to find the right people. To audition them, MacAulay gives each applicant a two-page case to review, along with a checklist that's an abbreviated version of the one used to grade the medical students. "I do a sort of mock interview with them," he says. "And then they fill out the checklist at the end." The trainer says that's usually enough to allow him to judge whether the applicant can look believable, along with meeting several other requirements. Standardized patients are not supposed to volunteer too much information, something that's difficult for a lot of people, he says. "Human nature is to help," MacAulay observes. "So someone might say, 'Well, I have this pain, and it's been going on for two days.' But that's too much. Better to let the student ask how long it's been going on. We want to test whether the students know which questions to ask."

At the same time, the standardized patients must be able to think on their feet. They get about 12 hours of training, but the med students invariably ask questions that weren't included in the preparation. MacAulay mentions one character who supposedly was from Seattle. A med student who hailed from that city perked up and asked which neighborhood the standardized patient was from. "I live by the university," she replied, a great answer, according to MacAulay. "She didn't have a clue where that was." But it satisfied the student and maintained the illusion that the patient was real. In contrast, he recalls one woman who had a mosquito bite on her arm that she scratched several times during her audition. "As a good student, I said, 'I notice you keep rubbing your arm. Is there anything wrong?' And she said, 'Oh, it's just kind of tingly.' Well, if I were a real student, I'd be thinking stroke or cardiovascular disease or diabetes -- you know, numbness and tingling." The most innocuous thing can set off alarms. "She should have just said, 'I have a mosquito bite.' "

Sometimes the student doctors' questions are ambiguous. MacAulay trains the actors in how to respond then. A student might ask, for example, "When you went to the washroom this morning, what color was the toilet water?" That would count, according to MacAulay, as an attempt to learn if the patient had seen blood in his or her urine. "Even though they didn't say, 'Was there blood in your urine?' you know the intent was to find out if the color was different."

I asked MacAulay whether any physical condition would preclude someone from becoming a standardized patient. "We have a lot of exercises where there's no physical exam, so we can hire anyone for the roles," he replied. "We can hire people like diabetics and hypertensives if they're under control." Scars or lumps might pose a problem for cases in which an examination will be required. "We don't want any scars that are going to confuse the student as to our intent. Say, for example, I had had open-heart surgery, and I was playing a case where I was having chest pains for the first time in my life. When the student took my gown down and saw the scar, they would say, 'Well, no wonder you're having chest pains. You have a history of cardiac disease.' " But nothing would prevent the person with the chest scar from playing a knee-pain patient.

A strong sense of physical modesty can be a bigger problem, MacAulay indicated. Although the students never do breast, pelvic, or genital exams, the standardized patients often wear nothing but underwear and a hospital gown. If a woman's complaint is chest pain, "It would be appropriate [for the medical student] to lower the gown," so the patient will be sitting there in her bra. "And bear in mind that we're also videotaping these sessions, and we're sitting in the monitor room watching." MacAulay says that's not an issue for most professional actors, accustomed as they are to sharing dressing rooms. "But for the nonactors, it can be a little unusual." Anyone who appears too uncomfortable with either the exposure or physical contact would not qualify.

If preexisting physical conditions don't preclude anyone from becoming a standardized patient, the lack of certain physical symptoms also usually isn't a problem, the trainer says. "There are a lot of tricks of the trade. For example, we can rig blood-pressure machines to give an elevated reading," he says. Makeup can be used to simulate cuts, bruises, and scars, and patients who need to be sweating can spray themselves in the face with a bottle before the student comes in. "Or we give them hot cloths that they hold in their hands, so that they seem sweaty." MacAulay says, "A lot of the stuff that we can reproduce is like a magic trick." A collapsed lung can be faked by having the actor hold his or her breath while moving the shoulders as if breathing hard. "The students are so focused on listening that they don't realize the patient isn't breathing," he says. "We had a student actually leave the room once and say, 'We've got to get this person to the ER!' "

In addition to the classic categories of stage acting, film acting, and TV acting, "This is like a fourth category," the trainer says. "It's two people in a room, but they're not playing a scene." The actors are answering unscripted questions, and at the same time they're trying to keep track of questions the doctor-to-be should have asked but didn't. They're scrutinizing the physical maneuvers. "So they have all these numbers and stuff in their heads, plus they have to make a mental note of specific things the student said and did that made them feel a certain way. It's a mental marathon!"

To understand it better, I asked MacAulay if I might audition for him. He agreed, giving me a two-page description of a woman who was experiencing severe head pain, along with some unusual visual phenomena. I could take ten minutes to study the woman's history, he told me, and as I did, I felt pretty confident I could pull off the impersonation. MacAulay then led me to one of the examining rooms, as believable a set as anything concocted by Hollywood. (The facility was once occupied by a group of internists.) He gave me a quick rundown on what I should watch for in the physical exam: He could choose to check the reflexes in either my wrists or my elbows and in either my knees or my ankles. But he had to check both the arms and the legs. He should test the strength in both arms and both legs, too. There was more.

Before leaving the room, he instructed me that I should be sitting on the examining table. "Bear in mind this is not an acting scene. So you have no agenda," he said. He recalled how he had entered the room one time to find the auditioner striking a dramatic pose and declaring, "Doctor! I have this headache!" Another time, the person trying out had picked up one of the medical instruments in the room and was playing with it to show her nervousness. "No, no, no!" MacAulay exclaimed at the memory. "When you go to your doctor's, you're sitting on the exam table."

A jolt of adrenaline surged through me a moment later when MacAulay, wearing a stethoscope, reentered the exam room, introduced himself, and started asking questions. If I'd forgotten that he once had made his living as a professional actor, his demeanor provided a swift reminder. It was still friendly, but he seemed subtly more formal and commanding -- the embodiment of a medical professional. For a while, I had no trouble answering his questions, but then he fired, "What do you do for a living?" at me, and the words "I'm a nurse" came tumbling out of my mouth, unplanned and unbidden. As I said them, my answer struck me as ridiculous. (What kind of nurse would be worrying about a brain tumor when she was having a routine migraine?)

"Oh," the faux doctor said. "Is that a stressful occupation?"

"Um, no," I stumbled, adding with no forethought, "I work in a rest home." The voice inside my head bellowed, "You MORON!" and for a fraction of a second, I teetered on the brink of laughter. I kept a straight face, but I felt flustered enough that when MacAulay a few minutes later took a 30-second time-out from our scene to ask if I could crank up the level of discomfort I was portraying ("Let's do more of a ten-out-of-ten for pain"), I told him I would, then promptly ignored the directive. During the physical exam, I struggled to keep track of all his brisk actions, but when I filled out the 18-question checklist, I marked down that he'd examined my retinas incorrectly, when in fact he hadn't even attempted to examine them. It wasn't my only error.

"It's tough," MacAulay sympathized. "And it's tough to find people who can do it really well." Peggy Wallace later elaborated on the unusual nature of the standardized patient's role. The patients "have to be good, really realistic performers, because the more realistic the patient is, the better the performance of the student will be." And they have to simultaneously assess the student's performance. "And none of [the patients] have a background in medicine!"

Wallace's own background has blended performance and medical science. As an undergraduate in the 1960s, she studied piano and dance, then got involved in broadcasting operas produced by her college. That experience made her hunger to capture professional operas on film, and in pursuit of that dream she enrolled in the film school at the University of Southern California, where she studied alongside George Lucas and got a doctoral degree. The daughter of a physician, Wallace says she changed her mind about what she wanted to film after her mother had a heart attack. In its aftermath, "I saw that there was an incredible opportunity to educate patients who are lying in bed about what's going on with them. My mother hadn't a clue." One day Wallace talked her way into the office of the director of the USC School of Medicine's Department of Medical Education. He offered her a two-month job running a computer-operated mannequin used for training anaesthesiologists. That stint turned into a 19-year career during which Wallace created a number of film and video programs.

USC also happened to be the birthplace of the simulated patient. The year was 1963, according to a history later written by Wallace. The innovator was a neurologist and medical educator named Harold S. Barrows. "He had this incredible intuition," Wallace told me. Barrows knew he couldn't guarantee that every time he walked into the hospital with his neurology students they would find patients with the conditions he wanted the students to learn about. So in the university's art department, he found a model who wasn't squeamish about disrobing. "First he did demonstrations in front of a large group of students," Wallace says. "How do you do a good neurologic exam? And when you're doing it, where are all these things on the body? He'd use a grease pencil and mark the model's body up."

Barrows then got the idea of teaching the model to simulate a neurological problem. He "needed a case about which he knew everything -- all the signs, all the symptoms," according to Wallace's history. "He needed a case that could be reproduced for every single student in the exact same way." Based on his experience with a patient at Los Angeles County Hospital, Barrows taught the model to portray a paraplegic woman who had multiple sclerosis. The question of how to evaluate the medical students' interaction with this simulated patient then arose, and Wallace says Barrows toyed with the idea of hiding behind a curtain and grading the students himself. But he finally decided to have the model fill out a checklist after every encounter.

"Students loved the technique," Wallace has written, and Barrows felt he was "learning things about those students [he] would never have found otherwise." Among his medical-school colleagues, however, the reaction to the pedagogical novelty was frosty. "No one else at USC was even interested in trying it," according to Wallace's history. Barrows was seen as lowering the dignity of medical education by his use of actors, and when the Associated Press caught wind of his activities, headlines such as "Hollywood Invades USC Medical School" resulted. Barrows published the first academic article about simulated patients in the Journal of Medical Education in 1964, but "[t]he USC dean received complaints from medical schools all over the country" in response, Wallace says.

Barrows nonetheless persisted in his use of standardized patients, developing his work for some years at USC and then moving in 1971 to a new medical school in Ontario, Canada, where the educational climate was more open to experimentation. Throughout the '70s and into the '80s, he continued expanding the ways in which he was using the simulations, and other pioneering medical educators began to follow suit. In 1985, two key review articles appeared in scientific journals that went a long way to legitimizing the practice, according to Wallace. "One no longer had to apologize to skeptics of standardized patient-based examinations," she has written. "One merely had to refer to these reviews."

Wallace met Barrows when he returned to the USC campus in 1984 to work on a pilot project to evaluate medical students' clinical skills. "At that point nobody was evaluating them," Wallace explains. "They tested with paper and pencils, asked them questions about illnesses." But that was all. Under Barrows's guidance, Wallace and other USC faculty members trained a cadre of standardized patients, and Barrows started doing regional demonstrations for medical school deans. Despite their intense initial skepticism, Wallace says, "Barrows knew that if they could experience it, they would see how real it was. That was the key. Once they did, it was like, 'Oh my God! This is amazing.' "

Medical schools all over the United States began working to expand the use of standardized patients in evaluating medical students and to change curricula in cases where it was warranted. Wallace says about the same time, the National Board of Medical Examiners began to think about using standardized patients to test the clinical skills of everyone applying for a medical license. Decades before, the board had required each license applicant to be observed with a patient by a medical professor. But back then, "There were no criteria for the patients," Wallace says. "There were no criteria for the faculty. The faculty weren't trained. So a student could get a very sweet physician who was easygoing and a patient with a cold, and that student would pass because the physician said, 'Oh yeah. They know how to do a good history.' At the other extreme, you could get a son-of-a-gun -- a tough physician -- with a complicated case, and the student wouldn't pass." People complained so much about the unfairness that the board abandoned the requirement in the early 1960s.

Not long after the licensing board began working on a new approach using standardized patients, Wallace made a change in her professional life. Although UCSD's medical school in 1995 was using some standardized patients, it was bringing them down from Los Angeles and had no formal program for training its own. This was at a time when USC's program appeared to be losing steam. So Wallace moved to San Diego and started a program here.

Today the standardized patients play a part in every phase of the medical education at UCSD. Wallace says even though classes in subjects such as physiology, anatomy, and molecular biology consume the majority of the medical students' first and second years, "There's a tiny piece of the curriculum for both those years that's called Introduction to Clinical Medicine. It's a half-day a week. In it the students are learning to take a medical history. They're learning about interviewing skills. They're learning about physical-exam skills." In the spring of their first year, they also have their first encounter with a patient (albeit a simulated one).

I watched as 36 members of UCSD's medical-school class of 2009 had that experience. When the session began, a standardized patient was sitting in each of the 12 examining rooms, poised to depict a pregnant 21-year-old in an abusive marriage, a man complaining about pain in his side after suffering a hockey accident, or a septuagenarian with chronic emphysema. The students entered the exam rooms in groups of three, as professors and other staffers hunkered down before the monitors in the observation room.

One of the students in each room was assigned to interview the patient, while the other two observed and filled out checklists (one for communication skills and one for history-taking). After getting feedback from both the actor/patient and a faculty member, the trios of students then moved on to a different type of patient and switched roles among themselves. Then they switched once more. "We just did it for the first time last year," Wallace had told me in advance. "And the students loved it."

I could understand why. Even the missteps were educational. When an earnest young man wearing a goatee and heavy glasses asked one of the patients with the side pain if the pain got worse when he inspired, the use of the technical term "inspired" -- rather than "breathed in" -- made the doctor seem confusing at best, if not downright alien. A young woman who kept her eyes locked onto the paper in her lap throughout most of her interview with the patient seemed so ill at ease she made me squirm.

I watched with particular interest what happened when Wallace gave feedback to a young man who had interviewed an ailing 72-year-old.

"I wanted to convey love to her," the first-year student said, "like she's a valued human being." But he felt frustrated and powerless over the limited number of medical options for someone in the patient's situation. "What I was sensing was she's never going to be able to do what she used to do. But I didn't want to say that," he told Wallace.

"That's good!" Wallace responded, making it clear that any words to that effect would have been a blunder. "But how can you be in a relationship to her?" the program director pressed him. "It's the big E-word -- empathy," Wallace answered her own question a moment later. "You said up front that you wanted to convey your love for her, which is so huge! Guys often have a problem with that."

"I felt like saying, 'It must be really hard,' " the student murmured.

"Puh-lease do that! Whenever you have that instinct!" Wallace exclaimed. "You have this capacity for empathy, but you were holding back on it."

Wallace says the med students' next opportunity to work with the standardized patients comes at the end of their second year, when they're tested on the way they handle six cases. Three of the patients have real physical problems. "So the students have to detect heart murmurs or whatever," Wallace explained. "They have to be able to look in eyes and see what's wrong." The students also see three standardized patients, who simulate conditions for which real patients aren't practical or available.

Classroom lectures and book learning move to the back seat during the third year of medical school. "It's the intensive clinical-skills-learning year," Wallace says, when students rotate through "clerkships" devoted to such topics as internal medicine, surgery, neurology, reproductive medicine, pediatrics, and psychiatry. "In all of these, they're working with real patients," the director says. The doctors running the clinics get permission from their patients to be screened by the third-year medical students. "They come in and do their histories and some physical exams," Wallace explains. "Then they come out and talk to the doctor and tell him or her what they've found out." The doctor then goes in and sees the patient himself. But at no time does the doctor directly observe the students interacting with the real patients, the way it's possible to do from the observation room on the campus.

"In October, when they've had a chance to do three months or so of work in the clinics, we bring them back in [to the standardized-patient center]," Wallace continues. "They're just beginning to establish their habit patterns for how they interact with patients. So it's a great time to catch them -- to give them feedback." Wallace says the October session has the third-year students each seeing two standardized patients, then undergoing a structured feedback and analysis program that unfolds over a month.

They get no grade for that performance, but the midterm exams held in February and March are graded, and they prepare the students for the one they must pass at the end of their third year. "That exam is developed by all eight medical schools in California," Wallace says. "The exact same exam is given to every medical student in every medical school in California. It's a consortium effort, where we develop it all together. We put it on throughout the summer, at different times." In that half-day session, each student sees the same eight cases. "If the students don't score high enough, they have to get remedial instruction," Wallace says. "Then they take the exam again." In the fall of their fourth year (when students are rotating through clinics and beginning to focus on their area of specialty), they must take the national medical-licensing exam. This includes a full-day clinical test in which they see 12 standardized patients. Wallace says no UCSD student has failed to pass the national board exam in the two years since it was inaugurated.

Medical undergraduates aren't the only ones who work with the standardized patients at UCSD. One program works with new residents, evaluating them so that their supervisors have a baseline understanding of what skills the new doctors are starting with as they begin their residency training. Another helps San Diego Hospice doctors refine their ability to break bad news and assist people in reaching difficult decisions. Wallace invited me to watch one of those sessions, in which Del Mar resident June Gottlieb would be playing Rose Bolen, a woman who would have to face an agonizing choice.

At 68, Gottlieb has silver hair, a face that's all but free of wrinkles, and an unorthodox career history. For about 25 years, she practiced psychotherapy in Ann Arbor, Michigan. But 16 years ago, she moved to San Diego and, looking for a new challenge, responded to an ad for a class that promised to help people break into show business. For Gottlieb, at least, the class worked. "I think I made 12 commercials, not really knowing what I was doing," she told me. She moved on to films and enjoyed a modest success at that. Then she fell in love with live theater. "There's no bigger challenge than having an audience that's different every night," she said, adding that she had appeared in at least a dozen local productions.

The Seinfeld episode in which Kramer gets a job playing a gonorrhea patient introduced her to the concept of standardized-patient work, and when she heard about UCSD's program, she decided to try out for it. Hired nine years ago, Gottlieb since has mastered several different roles. She might play Rose a couple of times a year, she estimated. "It kind of scares me sometimes to do Rose," she confided to me. "Because basically it's about having a loved one who has a chronic illness." Gottlieb's real-life husband has emphysema, "So I've thought, 'This could happen to anybody.' " That sense of connection helped her get into the role, she allowed.

On the day I watched her perform, Gottlieb was wearing the flowered dress she always dons when she assumes Rose Bolen's persona. It made her seem even more matronly. The doctor who entered the examining room to talk to her was Andy Probolus, a 30-year-old fellow in hospice and palliative medicine. He looked sympathetic but grave. After introducing himself, he said, "I understand your husband's been sick for a while. Could you tell me a little bit about that?"

"Well, I know he has a problem with his liver," Rose/June answered in a soft, tentative voice. "His belly swells up, and he can get really tired."

"Has he ever gotten this sick before?"

Her husband's illness had started at least six years ago, the woman told the doctor, and had finally been diagnosed a year or two later. "They said he had cirrhosis of the liver that he had gotten from some kind of an injection... So yeah. He's been dealing with this for a while. But he always comes out of it."

"He's just kind of bounced back?" Probolus queried.

"Well, you know. It's taken its toll." A month before, her husband had started throwing up blood, and he was in the hospital for a longer spell. "But I don't know what's going on." She sounded calm but befuddled.

"I wanted to kind of get a sense for what you'd like to know. Are you kind of a details person, or do you more prefer the big picture when you get information from doctors?" Probolus asked.

"Well, you know what's going on more than I do," Rose/June murmured. "I just have to trust and hope."

"I get the sense that he's mostly taken most of the information and has kind of helped you understand what's going on," the doctor said. It sounded like a question.

"We really don't talk about it," Rose/June admitted.

"I guess what I'm hearing from you is you'd like to kind of know what's going on."

"Well, I don't understand it, why they brought him over here."

"I'm afraid I have some, some bad news for you about him and what's been going on with him, okay? He's not awake right now. He's unconscious."

"Unconscious?" she murmured. "He was okay when we went to the doctor's..."

"His blood pressure was very low, and he's kind of slipped into being unconscious now. He's not really awake and alert, like he normally is. And the team that's taking care of him feels like he's had some more bleeding from the inside, which he's had before. And things look pretty bad right now. What I mean by that is, it looks like he's dying."

"Dying?" she whispered, aghast. "He's dying?" Probolus nodded. Silence descended on the exam room.

A long moment later, Rose/June asked, childlike, "What should I do?"

"Well, that's what I'm here to help with," Probolus responded.

"I can't believe he's dying!" the stricken wife would repeat several times over the course of the next 25 minutes. Probolus never looked impatient with her, and when he steered the conversation to the point where he could ask if Rose and her husband had ever discussed "how he would like his death to be," the question seemed uncontrived.

"It wasn't something we really talked about," she answered. "It seemed better to just be more hopeful. He's not a complainer, you know." She didn't want him to suffer, she said a little later.

"I think, based on what you're telling me, we're going to try our very best to keep him comfortable and let him pass peacefully and naturally," Probolus told her in response. Although the medical team could use CPR if his heart stopped, "My medical opinion is that if we do that to him, he would have probably less than a 1 percent chance of surviving." Furthermore, "It could be potentially painful," the doctor said.

"I don't want to cause him any pain," Rose/June responded.

After the conversation had ended and both patient and doctor had filled out checklists, Probolus came back into the exam room to talk to the actress. He mentioned that he had learned a mnemonic for having difficult conversations of the sort that he had just had with Rose. "I tried to kind of go step-by-step through that, and I think I covered almost everything I wanted to cover," he said. "Meeting you, introducing myself, developing some kind of rapport, then kind of giving you a warning shot that I had some bad news for you." He ticked off several other items.

Gottlieb's voice, when she spoke, had lost its overtones of vagueness and confusion. "As Rose, I liked when you took my hand right away. That felt really natural," she commented. "It didn't feel intrusive or anything." Rose also felt that Probolus had cared about her, but she had sensed "a little bit of a distance." Maybe that was because he had been going through his checklist, she speculated.

"I didn't want to miss anything," Probolus admitted. "I think to be honest, I may have been reacting a little bit to the time pressure with this."

Rose didn't feel rushed, Gottlieb told the doctor. "She just felt like there was just something..."

" 'Cause usually with the nature of this work, I don't put any time limits on myself," Probolus continued. "I just kind of let things go and see where they go."

Because Rose and her husband had never discussed his death, Probolus said he'd felt some need to "lean a little more on Rose" to make the critical decision about withholding CPR. He confessed that he was still trying to sort out how much paternalism was appropriate. "Early in my training, I thought, 'Okay, you just give people a list of options and say, "Pick one." ' Whereas now I've swung a little more the other way. I feel like all the options need to be there, but part of my job is to take somebody who's had devastating news and give them what feels like the right option, in my experience."

When I talked to him about the encounter a week later, Probolus was still beating himself up over his performance. He told me he thought he had "lost a little bit of the empathic connection that I feel is so crucial." Again he blamed that on the artificial time constraint. In comparison, in real life he might allow "an hour or an hour and a half for a conversation like that."

Still Probolus said the scene had felt real to him, and he was looking forward to reviewing his performance on tape, with an eye to analyzing which parts had worked best and which hadn't gone as well. He'd also appreciated Gottlieb's insights. "It's a unique situation in that you don't necessarily ever have a chance to say to somebody after you've broken terrible news to them, 'Did I do all right there? Was that okay? How did you feel?' That wouldn't go over well."

Later, talking to Wallace, I told her it made sense to me that such sessions would produce doctors who were more sensitive and satisfying to deal with. What surprised me, I confessed, was that medical schools like UCSD's cared enough to expend the effort and resources that the standardized-patient program requires.

In part it's been a hardheaded business decision, Wallace responded, one shaped by recent profound changes in the medical industry. Back "when physicians were in their own private practices and seeing their own patients, one by one, patients would leave them, or patients would stay because they liked them." But no one was looking at the pattern of lost patients -- losses that ultimately make any practice less profitable. With the rise in group practices and health-maintenance organizations, that has changed, Wallace contends. "Kaiser gives surveys out to their patients," she says. "Some organizations dock the doctors' pay if their patients aren't satisfied or give them bonuses if they are."

Furthermore, the relationship between the doctor and the patient affects more than just the practice's bottom line, she said. "The research has shown so much better improvement in health when you have a trusting relationship with a patient. If patients are satisfied with their relationship with the doctors, they tend to adhere to the treatment plan. That treatment plan should be a negotiation with the patient: 'Do you think you can do this? Do you have any problems taking antibiotics?' and so on.

"This is not just about the niceness of the doctor," Wallace said. "This is about knowing what patients need and helping those patients understand that you know. The patients then feel like they can tell you what they need to tell you. And that's important, because if they don't, the physician may miss something that is key to understanding the problem." (Editor's note: Some of the symptoms described in this article have been altered in the interest of preserving their usefulness on future UCSD medical exams.)