You go to the doctor with a problem; the doctor prescribes a treatment. You, of course, hope it works. You hope your doctor has some evidence behind the treatment that she is offering you. We expect our doctors to practice evidence-based medicine, to make decisions based on good-quality research. We expect that they choose the blood pressure medications, chemotherapies, and cardiac treatments that offer the best-proven outcomes.
But what about how doctors speak? Is there evidence that the medical conversation influences clinical outcomes? On the surface, it seems hard to imagine that the words we doctors say to our patients, and how we say them, can have a potency comparable to the medications we prescribe, but a surprisingly robust body of research supports this. Given that words are a lot cheaper than medications—and rarely make you gain weight or break out in hives—this seems like a worthy avenue of exploration.
One of the earliest studies that showed a demonstrable effect of the simple act of talking was conducted in 1964.1 As anyone who’s undergone abdominal surgery knows, one of the worst hardships after surgery is pain. Postoperative pain, especially in the abdomen, is miserable and can require prodigious doses of narcotics to effectively control it. But narcotics produce a myriad of side effects ranging from nausea, vomiting, and itching to mental changes, drowsiness, and full respiratory arrest. For many patients—and their doctors—the intractable constipation can be the biggest torment. Thus, minimizing post-op narcotic use is a high priority for everyone on the medical playing field.
In the 1964 study, a group of about one hundred patients undergoing abdominal surgery was studied. The night before the operation, an anesthetist visited each patient to explain the surgery and anesthesia. For half the patients—randomly selected—the anesthetists added in a twenty-minute discussion about post-op pain. Patients were told that pain was a normal part of the process and that it was caused by muscle spasms. They were told where the pain would likely be located, when to expect it, and how long it would last. The anesthetist offered suggestions on how to relax the muscles to minimize the pain. Lastly, the anesthetists gave instructions on how to request pain medications if the relaxation techniques didn’t work. The anesthetists were instructed to be enthusiastic in their presentation and express confidence that the pain would be relieved.
The control group did not receive this extra discussion. The surgeons who operated and took care of the patients after surgery had no idea who was in which group.
Looking back at this study from today, it doesn’t seem surprising that the group with the extra discussion experienced less pain. Giving patients the information that they need—not necessarily the standard of care in 1964—goes a long way toward relieving the anxiety and fear that worsen pain. But what was impressive was the magnitude of the effect. The group with the extra discussion needed half the amount of pain medication that the control group needed. When it comes to serious medications like narcotics, a reduction by half is a remarkable improvement in medical care.
But the real shocker was that these patients were discharged from the hospital three days earlier than those in the control group! A typical day in the hospital now costs upward of four thousand dollars,2 so here is a decidedly low-tech intervention—a twenty-minute conversation—that saved more than twelve thousand dollars in present day money, and spared patients many days of pain and misery. (And as an internist, I will add that any intervention that minimizes constipation is worthy of a gold medal in and of itself.)
This landmark study ignited the field, as researchers began investigating in earnest to see if and how doctors’ communication could have effects on patients’ health comparable to standard medical interventions. One study that caught my eye focused on back pain. Low back pain is as common as oxygen. If there has been a single day in my practice that back pain has not come up, I certainly can’t recall it. Once the serious but exceedingly rare causes of back pain have been eliminated—cancer, infection—and the diagnosis is regular old muscular back pain, physical therapy is the treatment of choice. When I send my patients for physical therapy, I think of it in terms of the exercises that the patients will be guided to do, sometimes accompanied by other modalities such as ultrasound, massage, and mild electrical stimulation.
What I don’t typically think of is how the communication between the therapist and my patient will affect the pain. A group of Canadian researchers divided 120 patients with back pain into four groups.3 Half the patients received the electrical stimulation and half received sham stimulation (all the equipment is set up, but the electrical current is never activated). Because electrical stimulation causes a mild pins-and-needle sensation, the patients in the sham group were told that the machine was on but that with this “new machine” they might not feel any tingling at all.
As had been shown in many other studies, sham treatment (i.e., placebo) works quite well: these patients had a 25 percent reduction in their levels of pain. This by itself is impressive. Those patients who got the real stimulation did better, though; their pain levels decreased by 45 percent, so we know that the electrical stimulation actually does something.
Each of these two groups, however, was further divided in half. One half experienced limited conversation from the physical therapist. The therapist briefly explained the procedure but then told the patient that everything needed to be quiet during the treatment. With the other half, the therapists engaged in conversation the entire time. The therapists were striving to create a “therapeutic alliance” with the patients in this group and so asked open-ended questions, listened attentively, and inquired about how the back pain was affecting the patients’ lives. The therapists expressed empathy about the patients’ situation but also offered words of encouragement and optimism about getting better. There was plenty of eye contact and touch.
Unsurprisingly, the patients who had the more engaged therapists did better, but it’s the degree of improvement that’s impressive. Patients who underwent sham treatment—no electrical stimulation—but had therapists who actively communicated with them reported a 55 percent decrease in their pain. Think about that for a minute: the communication with the physical therapist was more effective than the treatment by itself (a 55 percent reduction in pain with communication alone as opposed to a 45 percent reduction in pain with just electrical treatment).
The patients who had the electrical stimulation accompanied by the engaged physical therapists were the clear winners, with a 77 percent reduction in pain. This type of study provides hard evidence for what healers, shamans, witch doctors, and assorted mystics have known for millennia: that a substantial portion of “healing” comes from the personal connection that is formed with the patient. It’s no great secret, of course. Wise physicians and experienced nurses (not to mention astute patients) have also keyed into this, even if only subconsciously. Nevertheless, it’s intriguing to see this effect borne out in controlled scientific studies.
These studies bring me to the whole idea of placebo. The placebo effect, as most people know, is a change in someone’s health from a treatment that contains no active medical substance. The effect is well known, which is why every clinical trial worth its salt is placebo-controlled. The treatment under investigation must be compared with a placebo, not to “doing nothing,” because the placebo group always registers an effect. The treatment has to score better than placebo, otherwise it’s deemed useless.
But outside of research, placebo has been viewed with suspicion. Using placebos in actual clinical medicine is considered shady, if not downright unethical. But communication plays an enormous role in placebo and the eschewal of using placebo in treatment is starting to ease. I had occasion to employ a placebo one night during my second year of medical residency, though even to this day I’m not sure whether I did the right thing.
It was well past midnight on the AIDS ward at Bellevue Hospital in New York City, at a time when this ward was overflowing with patients. Despite the late hour, the nurses and doctors were still rushing about at full speed, as the admissions continued to surge in, each more feverish and emaciated than the previous. But even the AIDS ward eventually quieted down for the night. Except, that is, for the howler.
The howler was a patient in his thirties who’d earned his nickname for his nightly bouts of wailing. He was already receiving hefty doses of pain medication, yet he kept screaming to the nurses about his pain. This went on, night after night, despite extensive medical evaluations to see if there were any missed explanations for his pain.
Nothing seemed to help and the nightly yowling was agitating the other patients, not to mention driving the nursing staff to distraction. (We doctors had patients on many wards so were constantly in and out of the ward. The nurses, on the other hand, were staffed to a specific ward and so could not escape the howler.) The head nurse stat-paged me at 3 a.m. “You have to do something,” she said, her voice boiling with apoplexy, “before somebody marches in there and strangles him.”
Reluctantly, I trudged back to the patient’s room. I’d been on my feet for more hours than I could count and this was my fourth visit to the howler that night. By this point, we were both pretty exasperated with each other. He was sullen and cranky; I was exhausted and at my wits’ end. I watched the patient writhing in bed and felt bad for him, but his moans burrowed into my brain, deflating my last three functioning neurons. The room was stuffy from that medicalized staleness that only hospitals can brew up. Would this night ever end?
I rummaged around in my pockets to see what I could come up with and pulled out a vial of saline. On a whim, I plucked a syringe from another pocket and slowly peeled back the wrapper. Stepping in closer to the patient’s bedside, I cocked back the syringe and drew up 1 cc of the plain saline. “You know about Tylenol, right?” I said to the patient, who was continuing to twist within his bed sheets. “And you’ve heard of Tylenol number three, the kind with codeine.”
I leaned forward and held up the liquid-filled syringe close to the patient’s face. “There’s even a Tylenol number four.” I slowly removed the cover off the needle and it glistened in the fluorescent lights. “But this“—and here I paused for dramatic effect—“this is Tylenol number five!”
The patient stopped howling and gave me an interested look. Without a word he lowered his pajamas and allowed me to inject the saline into his gluteus maximus. I disposed of the syringe in the nearby sharps box and then pulled up a chair to his bedside. The patient and I waited together, allowing the minutes to tick unhurriedly by.
After what seemed like a mutually agreeable time, I stood up and bid him good night. The patient put his head to the pillow and promptly fell asleep. The ward was silent for the rest of the night.
I did feel guilty that I had committed an outright deception with this patient—something I knew was a true no-no. On the other hand, it was the first time he got a full night of sleep, to say nothing of all the other patients on the ward and the rest of the staff. (The head nurse bought me coffee and a bagel the next morning and we remained buddies for all the years of my residency.)
When I related this story to Ted J. Kaptchuk, director of the program in placebo studies at Harvard Medical School, he gave a sigh of recognition. “We all have our moments of desperation,” he said drily. “Usually around midnight.”
Kaptchuk does not condone deception, but research bears out that how caregivers present and administer treatments has a powerful effect on clinical outcomes. In an ingenious Italian study, patients who’d just undergone chest surgery were all given the appropriate doses of narcotics for their post-op pain.4 For patients in one group, the narcotics were automatically administered in their IVs. The patients had been told in advance that IV pain meds would be given reliably over the course of the day but they were not given any indication at the time each dose began. Nor were any doctors or nurses present when the medication began to drip in. For the patients in the other group the medication was given on the same schedule but a doctor injected it manually.
The patients in the latter group consistently achieved more main relief than the patients who received the medicine unknowingly in their IVs, even though it was the same dose of medication given at the same intervals. The rituals the doctors performed—coming to the bedside, acknowledging the patients’ pain, drawing up the medication, visibly injecting it into the IV, and discussing the expected benefits, not to mention the attention and caring that comes with the presence of an actual human being—effected as much pain relief as doubling the dose of the medication.
Kaptchuk describes placebo not as the traditional sugar pill but as “everything that surrounds a medical treatment”: how caregivers describe the medication, how they administer it, the expectations they offer for the medicine, their tone of voice, their words of connection, their strength of eye contact. In short, everything that doctors and nurses do in an interaction with a patient.
As I mentioned, healers and shamans have known intuitively about the importance of this interaction since the dawn of time. Before we had treatments that could actually impact the pathology of disease—antibiotics, chemotherapy, stents, organ transplants, transfusions—the “everything else” was the mainstay of medical care, and in many cases it was remarkably effective.
But in the twentieth century, now that there were actual medical treatments, placebos were considered psychological mumbo jumbo, more akin to hypnotism than real medicine. The biological breakthrough came in 1978, when researchers showed that not only was the placebo effect real, but that it could be reversed by administering naloxone—a chemical that blocks our endorphins.5 Endorphins are intriguing neurochemicals that act as our homegrown painkillers. The term, in fact, was coined by my PhD adviser Eric Simon as a contraction of “endogenous morphine” because that’s how these neurochemicals behaved, just like morphine.6 (I spent several wonderful years escaping med school in Eric’s lab, as we researched the biochemistry of the receptors for these endorphins.)
Despite this biological underpinning, doctors and nurses still feel uneasy when it comes to placebos. Somehow administering them seems wrong, unethical, deceptive. But patients, it turns out, are more flexible in their thinking. Kaptchuk has demonstrated that patients experience symptom relief even when they are told quite plainly that they are getting a placebo. In one of his experiments, patients with irritable bowel syndrome randomly received either a placebo or no treatment at all.7 The patients in the placebo group were explicitly told that the pills were made of inert substances. They were also told that research had demonstrated that placebos relieve pain.
This “open-label” placebo is a direct contrast to standard clinical trials in which participants don’t know if they are getting the active medication or the placebo. The blinding in clinical trials is deliberate because, presumably, if you knew you were getting the blank pills you wouldn’t experience any placebo effect and the true effect of the test medication couldn’t be accurately teased out.
In Kaptchuk’s study, however, the patients knowingly took the placebo. There was no effort to dress it up with the chance that they might be receiving an active medication. Yet after three weeks these patients who openly took an inert pill experienced a decrease in their symptoms and an increase in their quality-of-life ratings. In this study, and in other surveys that Kaptchuk has done, patients can be quite amenable to open-label placebos.8 As one of my patients once told me, “If it gets rid of the pain, I don’t care if it’s a dill pickle!” Anyone who has struggled with chronic pain would be likely to agree.
The ethics are evolving but the general consensus is that transparency is the bottom line. What I did as a young and desperate resident wouldn’t pass muster because I had deceived my patient about the medication. But if I’d told him honestly that it was inert saline and if he’d given informed consent, it could have been a reasonable treatment option. From what Kaptchuk has found in his studies, my patient still might have experienced relief even knowing I was injecting him with nothing but saline, especially if we had been able to have a compassionate conversation about his pain. Relieving suffering (without causing more harm), after all, is what the Hippocratic oath is all about.
I’ve found myself utilizing some aspects of placebo medicine in my clinical practice. It comes up most often in patients with vague pain and fatigue, for whom the medical workup has not revealed a specific cause. Many of these patients have additional psychological and social stresses that compound their situations. Frequently they will ask if a multivitamin will give them more energy. In the past I would say no because there are no scientific studies to demonstrate this, and also because in the absence of a vitamin deficiency there’s not much for a basic multivitamin pill to do. Now, however, I take a different approach. I will say something along the lines of “Many of my patients find that they have more energy when they take a multivitamin.” I’m not lying, because many of my patients have indeed said so. I’ll be honest with them that we don’t have rigorous scientific data, but I’ll also point out that research studies deal with average effects over populations; they can’t predict how an individual person might respond.
I would never, of course, say this regarding a medication with a more toxic side-effect profile, but given that there’s little downside to an inexpensive multivitamin I’ll encourage them to give it a try and offer them my optimistic take that it’s definitely possible they’ll feel better. If it helps, we’ve accomplished something. If not, we’re no worse than before. Most of my patients still have symptoms but there are always a few patients who come back at the next visit and swear they feel much better on the vitamin pill. There are some who argue that it is unethical to promote placebo-type medications to patients. But increasingly, many say it would be unethical not to give placebo a try in situations where patients are not getting relief from traditional means (and where it would not cause harm or otherwise replace a necessary treatment).
Kaptchuk views placebo as just one of the many things in the tool kit of medicine. It would never be a substitute for appropriate medical care but it is something that can enhance medical care greatly. Wise doctors and nurses already know this. They’ve found, usually just by personal experience, that their “everything else”—respect, attention, comfort, empathy, communication, touch—often forms the lion’s share of medical care, no deception required.
A key part of the “everything else” of placebo is clearly the communication between doctor and patient. One salutary thing communication can do is reduce anxiety, which in turn can reduce pain. A thirty-eight-year-old woman came to my office with concerns about chest pain. She was worried—as most people with chest pain are—that it could be a heart condition. We spoke about her symptom and she told me it occurred twice a month and lasted for thirty minutes at a time. It wasn’t precipitated by exertion—in fact, she had a physically strenuous job in a warehouse transporting heavy boxes. She regularly climbed several flights of stairs during a typical workday. I also learned that she was under a lot of stress because recently her hours had been cut back and her economic situation was now in jeopardy. The stress was agitating her sleep but she was trying to keep everything under wraps so as not to upset her older mother, whom she lived with and supported financially.
The chest pain she described was a long distance from classic angina—pressing chest pain brought on reliably by exertion, relieved by rest, lasting only several minutes. Her robust daily physical activity was the equivalent of a cardiac stress test that indicated a heart with plenty of oxygen and blood supply. As a premenopausal woman of thirty-eight, her statistical risk for heart disease was quite low. Plus she had an alternative explanation—life stress—not to mention a normal physical exam and EKG. So by the end of our conversation, without injecting any delay by ordering additional tests, I was able to tell her that I was entirely convinced this pain was not from her heart, that her heart was extremely healthy.
Her relief was palpable—she sat back in her chair with a long sigh and an even wider smile. Apparently she’d spent the past few months nursing a growing conviction that her heart was failing and that she might no longer be able to do her job or care for her mother. At our next visit a few months later, she said that her chest pain had simmered down, and though it was still there, she hardly noticed it.
Maybe she’d been having muscular soreness in her chest from her heavy lifting and it had improved on its own. But maybe my reassurance that her heart was not compromised offered a placebo effect. Maybe it allowed her to relax enough to stop tensing the muscles that might have been contributing to her pain.
Reassurance that you are not in imminent danger, just by itself, can go a long way toward easing pain. Even for patients who are in imminent danger—those with terminal illness, for example—reassurance that they will not be alone in their illness, that their comfort and dignity will be attended to, that their advance directives will be honored, that their passage to death will be eased as much as possible—offers tremendous comfort and diminution of pain.
Reassurance may act physiologically by decreasing cortisol and adrenaline and perhaps increasing endorphins, which might directly reduce the neural perception of pain. Or reassurance may act cognitively by shifting the focus of concern so that the pain is no longer front and center. Either way, reassurance is a crucial part of the “everything else” that doctors and nurses do that decreases pain and discomfort.
A second way communication might offer placebo effect is by raising expectations. When I tell a patient, “Many of my patients have found that they have more energy when they take a multivitamin,” I’m raising expectations. When Ted Kaptchuk told his ulcerative colitis subjects that the placebo they were taking had been shown in clinical trials to decrease pain, he was raising expectations. Now you could make the argument that this is a bit of huckstering, but in fact there is a good deal of evidence that expectation can affect the biology of pain perception.
In an intriguing study on pain perception,9 twenty-two healthy volunteers were given a painful stimulus (a hot pad placed on the calf) and also an intravenous narcotic (similar to morphine) to relieve the pain. At the same time, their heads were conveniently positioned in an fMRI (functional MRI) machine, so researchers could document which parts of their brains were activated during the study.
The participants went through four rounds of the experiment. Each round was identical—the hot pad was placed on the calf and then the pain medication was infused into the IV—and the patients had to rank how much pain they felt. In the first round inert saline (placebo) rather than the actual medication was infused into the IV. This established a baseline pain score.
In the second round the pain medication was administered silently, without the patient being told. The medication clearly worked because the pain scores were lower, even without the patients being explicitly informed that the medication was starting.
In the third round it was announced that the medication infusion “would be now started by the anesthetist.” As you might guess, being told that the medicine was about to start rolling in raised expectations and enhanced the pain relief. In fact, just having this positive expectation doubled the relief.
In the fourth round researchers announced that “the infusion would be stopped” in order to investigate potential “rebound effects” (pain that worsens after medications are discontinued). The medication, however, was actually given, just as it had been in the earlier rounds. But the subjects now had negative expectations—not only that they were not getting any pain meds, but that the pain might actually worsen. The negative expectation indeed wiped out the effects that the positive expectation in the third round had offered. Not only that, it also wiped out any baseline effect of the medication (the pain relief in round two). In the end, the simple act of negative expectation blocked the entire biological effect of a potent intravenous narcotic.
This is a fascinating finding: our expectations can be as influential as an actual medication. Now you could argue that maybe the subjects gave the answers they thought the researchers wanted (more pain relief when the medication was given, less pain relief when the medication was supposedly turned off). But the fMRI readings offered tantalizing suggestions about a biological explanation. When the study subjects received the medication along with a positive expectation, there was an increase in activity of the brain areas known to be involved in pain relief (these areas, incidentally, light up regardless of whether pain relief is from a narcotic or from placebo). In contrast, when the pain medication was administered along with a negative expectation, there was reduced brain activity in these areas.
While these are only experimental data, they bolster the idea that how doctors and nurses frame a treatment can have profound effects on how patients experience the results of that treatment. Earlier in my career I made efforts to present treatment options as soberly as possible to my patients. I laid out the risks and benefits so the patient could make a reasoned decision. I still do that, of course, but now I also try to frame things as optimistically as possible, within reason. I don’t want to be a Pollyanna or paint unrealistic expectations for treatments that offer little value. But if we’ve chosen a path of treatment together, I try to invest both of our expectations in the right direction. Our communications might be a bit of placebo, but if it can help without causing side effects then it seems like a legitimate medical intervention to pursue.
What about adherence? What parts of the doctor-patient communication contribute to—or detract from—adherence? As Tracey Pratt and the work of Heidi E. Hamilton demonstrated, nonadherence is rarely about indifference or ignorance on the part of the patient, though that’s what doctors often assume. Rather, it’s usually the realities of life that seem to conspire against a patient’s efforts to fully adhere to a treatment plan. It’s been estimated that patients are able to follow doctors’ recommendations—for medications, tests, therapy, diet, consultations—only about half the time.10 That leaves a substantial number of patients who are missing out on the potential benefits of these recommendations. Many things have been posited to improve adherence: nurse visits, social-work counseling, follow-up phone calls, electronic reminders, home delivery of medications, health coaches, economic incentives.
Improved communication is certainly one of those things, but the devil for all of these interventions is in the magnitude of effect. The medical system needs to know how much bang there is for the buck before it chooses where to invest resources. So a team of scientists examined more than fifty years of research,11 using the technique of meta-analysis to combine the results of one hundred individual studies. (These studies encompassed a total of forty-five thousand patients.) If all of these data are taken together, patients of doctors with good communication skills are more than twice as likely to have good adherence to medical recommendations, compared with patients whose doctors had poor communication skills. That’s a 100 percent improvement, for those of you who prefer percentages.
For a rough comparison to a medical intervention that’s considered high value, the benefit of statins in preventing heart attacks in patients with atherosclerotic disease is about 33 percent: those who take statins are a third less likely to have a heart attack as those who don’t. Improved communication is a remarkable medical intervention that can rank in the big leagues (and of course doesn’t even reach the hem of statins when it comes to cost).
One thing often believed to improve doctor-patient connection is humanizing the experience. Doctors sometimes attempt to do this by bringing in their own humanity, presenting themselves as real people, and not just as a stock figure in a white coat. Allowing personal details to enter the communication between doctor and patient is a common occurrence—estimates suggest that physician self-disclosure occurs in up to a third of routine office visits.12 But is this actually helpful for the patient?
A couple of years ago I had a new patient in her midthirties, Anita Lyons. I found the medical care of Anita Lyons overwhelming from the get-go. She weighed nearly three hundred pounds and her diabetes was raging out of control. She suffered from asthma but continued to smoke to ease her anxiety. Her obesity was severe enough to constrict her mobility, but the constant stress in her life led her to overeat.
Ms. Lyons had been raised in a family beset with psychiatric, medical, and social challenges, and now the cycle seemed to be replicating itself. Her three children all had emotional and educational needs that required intensive services.
She couldn’t figure out how to get a handle on her life, and as her new physician, I wasn’t sure how to get a handle on it, either. Should I work on her diabetes first? Her smoking? Her weight? Her asthma? Her anxiety and stress? Her life seemed frighteningly out of control. On what, I wondered, could we connect?
In the middle of our first visit her cell phone rang. She glanced at the caller ID then looked apologetically toward me. “It’s my son’s teacher,” she said. I told her to please take the call. As a parent of three school-age children, I well understood how hard it could be to reach a teacher and how critical such calls were.
“I know my son missed two assignments,” Ms. Lyons said into the phone, as I scrolled through her labs. “Sometimes he’s anxious after therapy sessions, so I have to give him time on the playground.” She paused and listened to the teacher. “We keep a chart,” she said, “and he gets a check mark for each homework he finishes.”
As she carried on the conversation I found myself feeling more connected to her. I knew about the driving necessity of playground time. I knew about the charts and the checks required to keep children focused on their tasks. My kids were about the same age as hers and I was all too familiar with the frenetic juggling required to keep up with three children and their schedules and their laundry and their homework and their lunches and their tantrums. I also knew the awkward sensation of explaining details of a life to a teacher, feeling like your messy reality won’t pass muster.
When she hung up the phone she apologized for the interruption. “Don’t worry,” I reassured her. “Teacher calls are important.” I was about to continue, to offer her some of our commonalities, thinking that common ground could offer us a good start on a working relationship.
But then I stopped myself. A week earlier I’d had a visit with one of my long-term patients—a Polish woman in her sixties who has been alone, and lonely, for as long as I’ve known her. At the end of our visit, she asked why there weren’t any new pictures of my children; the photo on my desk was more than three years old.
“I never get around to printing any photos,” I said sheepishly. “They just sit on my phone.” She then leaned toward the cell phone sitting on my desk, somewhat expectantly. I debated what to do. I wasn’t sure if sharing a personal photo was okay, but I also didn’t want to seem imperious or cut off her very human gesture. I figured it would be okay to show her one quick photo. After all, we’d known each other for so many years by this point. I quickly dug up a family photo from my phone.
My patient broke into a wistful smile when she saw the photo. “What a happy family you have,” she said. “Your husband looks like such a nice person. Your kids look so smart.” The plaintiveness in her voice was so palpable, and I suddenly realized what a mistake I’d made. I’d thought it would strengthen our bond to share a bit of my life. But it turned out it may have had the opposite effect, as she projected onto my life some of the things that were missing in hers.
I quickly closed the photo, embarrassed at the awkward dynamics. There likely were some patients for whom the addition of the human element of the doctor’s life might be helpful, but this patient probably wasn’t one of them. I felt that I might have done something detrimental.
And so as I sat with Ms. Lyons, a woman who in many ways was so vastly different from me, I wondered if sharing some of the commonalities we did have would enhance our alliance and give us a better chance at tackling her medical issues.
In the end I decided to keep quiet. I couldn’t be sure if such disclosure would offer solidarity or inadvertently toss in a monkey wrench. Instead, I simply acknowledged how challenging the situation with her children must be. I offered my admiration for her efforts and pointed out her successes.
She reddened slightly but smiled. “Raising kids is harder than I’d ever imagined.”
“I know,” I said. I wanted to say that I really did know, but I decided to leave it at that.
Research has borne out that doctors’ self-disclosures often turn out not to be as helpful in addressing patients’ concerns or building rapport as doctors presume them to be.13 The attempt to empathize is certainly genuine: “My father also had lung cancer; I know how difficult it can be” or “I also had knee surgery; I know how hard the recovery is.” But patients sometimes interpret these as disruptive, as the focus is shifted off them onto the doctors. In primary-care visits these self-disclosures are associated with lower patient satisfaction.14 Patients of doctors who disclosed personal details gave lower ratings of both friendliness and reassurance.
Interestingly, the opposite effect was found with surgeons. In the same study it was shown that patients of surgeons who offered self-disclosures gave higher ratings of friendliness, reassurance, and overall satisfaction. This may be because surgical visits often revolve around a single procedure for which the anxiety level of the patient is high and also focused. Personal reassurances—”My father had a bypass and he did just fine”—may do a lot to reduce fear, even though the outcome of the surgeon’s father has no actual bearing on how the patient will fare. Primary-care medical visits, on the other hand, typically cover many areas of the patients’ medical conditions, with no specific event such as an operation to focus on. Doctors’ disclosures may feel like unwelcome sidetracks in these scenarios.
The desire to share commonalities is a human impulse. I’m not suggesting that doctors should never do this. Rather, doctors need to think before they do so reflexively and ask themselves if these particular details will offer assistance or if they will instead detract from the singularity of the patient’s experience. Most important, if the doctor is sharing some personal details, the conversation needs to be steered promptly back to the patient, where it properly belongs. In my experience, when my patient asks about my life, or I find myself considering sharing something personal, as with Ms. Lyons, I realize that it always comes from the desire to kindle empathy. Despite all the advances of technology, medicine is still fundamentally a human endeavor, and whenever two human beings attempt to connect, empathy is the key building block upon which a relationship is created.
Empathy is often thought of as an emotion or a feeling. It’s classically defined as the ability to put yourself in someone else’s shoes and to understand how that person is feeling. But empathy qualifies as a communication skill as well. For empathy to be effective in medicine, the doctor has to communicate that empathy to the patient.15 The doctor could be the most emotionally sensitive hominoid on the planet, steeped in the patient’s culture and life story, willing to donate a kidney, a lung, and both corneas to the patient if need be, but without the ability to communicate that empathy, it will be meaningless for the patient.
So can empathy—with its critical reliance on communication—affect patient outcomes? One seminal study examined nearly nine hundred patients and divided them into three groups, depending on whether their doctors scored low, medium, or high on a questionnaire that measured empathy.16 For those patients lucky enough to have doctors with high empathy scores, 56 percent had good diabetes control and 59 percent had good cholesterol control. Of patients with doctors who scored low on the empathy scale, only 40 percent had good diabetes control and 44 percent had good cholesterol control. Using statistical analysis to control for age, gender, and health insurance, the researchers found that empathy was a significant contributor to the positive outcomes.
This, of course, is a study that measures an “association” between empathy and the positive outcomes. Despite statistical controls, there could be something else that caused the results. Maybe the low-empathy doctors had dismal hygiene and the resulting BO was too distracting for the patients to pay attention to their diabetes. Maybe the offices of the high-empathy doctors offered cloth gowns rather than paper gowns, so their patients weren’t experiencing frostbite and thus were better able to hear what the doctor was saying. You never know what the confounding factors might be . . .
Another group of researchers set out to test the effects of empathy with a slightly stronger study design that used randomization, in this case with the ordinary, annoying, common cold.17 More than seven hundred patients were randomly placed into three groups as soon as they presented with cold symptoms. One group didn’t even see a doctor. The second group had “usual care” with a doctor. The third group saw doctors who, like the engaged physical therapists in the study I mentioned earlier in the chapter, made a special effort to connect with the patients. They listened closely and specifically worked to empathize with the patients’ situation. They were also optimistic about a good prognosis.
The researchers were most interested in the patients’ perceptions of empathy, so they had the patients rate the doctors on empathy and communication skills. The doctors who scored the highest were mostly from the “enhanced visit” group, though a few were from the regular-care group.
Patients who were taken care of by this high-scoring group of doctors did better than the other patients. They had a 17 percent decrease in severity of their colds, compared with patients who saw the other doctors, and the duration of their colds was shorter (5.9 days versus 7 days). These may not seem like large numbers, but one day less with your nose dripping into your coffee and your head stuffed with cotton is a valuable benefit in my book.
The researchers speculated that there might be a cellular basis for these results, so they examined the levels of white blood cells and inflammatory markers (interleukin-8) present in the nasal mucosa. Both of these biological markers dropped more profoundly in those patients with high-scoring doctors compared with patients of the other doctors. Though it doesn’t offer proof, this study hints at some biological mechanisms that may underlie the positive effects of good communication and connection. It’s possible that levels of cortisol and adrenalin are lowered when good communication lowers anxiety levels.
However, most researchers in the field believe the effects are more indirect. Improving communication can help patients understand instructions better or feel more confident about their ability to manage their disease. This may make them more likely to fill their prescriptions and take their medications, or follow their doctors’ advice to take the day off from work and just sleep for twenty-four hours. With more empathetic communication, patients might feel safe enough to confide uncomfortable information that they might have otherwise been too embarrassed to share—a hidden eating disorder, a genital symptom, financial constraints—allowing the doctors to tailor the treatment appropriately. Doctors who take the time to understand patients’ family lives and living conditions might be able to prescribe a medication that is more affordable or enlist a family member to help out with medical care or offer a discreet way to obtain treatment for an embarrassing condition.
Patients who feel stigmatized during their medical encounters—because of obesity, for example, or mental illness or substance abuse or gender identity or their racial or ethnic group—are more likely to drop out of the medical system entirely. But with empathetic doctors who take more time to connect, these patients may stay engaged in medical care and get the treatment they need. These indirect patterns probably explain most of the successes seen in studies of communication in medicine.18
What’s tantalizing is that it doesn’t require much additional effort on the part of the physician to have a meaningful effect. Nearly every good-quality study that gave the physician a concrete step to take resulted in better communication and improved patient satisfaction. Whether it was getting doctors to ask more open-ended questions, or to elicit the patient’s concerns, or to set a clear agenda, or to inquire about the patient’s life beyond the illness, or to involve the patient when choosing a treatment plan, or just to express more empathy—nearly any attention paid to these skills easily improved communication.19 Now, whenever you bring up these kinds of studies in medical circles, there’s always a good dose of eye rolling. Even doctors who intuitively understand the importance of good communication and connection with patients get a little impatient when you start talking about the hokier-sounding soft stuff. It doesn’t have the same satisfying heft as the multicenter clinical trials of cardiac stents with tens of thousands of patients in twenty different countries run by fifteen of the top academic medical centers, with budgets that rival the national GDPs of respectable midsize nations.
I sympathize with this because I, too, have a similar reaction. It doesn’t feel like “real” medicine and the gut reaction of most doctors is to push it to the side, somewhere near the viper-fish broth that torpedoes heart attacks in mid-clot and the Paraguayan cactus powder that evaporates cellulite in three short weeks for only $49.95 plus tax. Even with scientific evidence, we doctors still find this squishy stuff hard to embrace.
When I try to dissect out why doctors are so leery, I come up with two reasons. One is that these less-tangible components of medicine—communication, connection, empathy—are harder to measure than glucose levels or numbers of strokes. Although research in these fields has progressed significantly in the past few decades, it still lags in size and depth compared with the mega-trials of oncology treatments and cardiovascular outcomes. Plus you can’t really write a prescription for communication or empathy, so it doesn’t feel as real as other medical treatments (and how on earth would you convince the insurance company that this constitutes actual medical care and should be reimbursed the same way that endoscopies are?).
But the second reason is, I think, the more profound one. Communication, empathy, and connection are not things doctors typically learn in medical school. Med schools are beginning to pay attention to these skills but they are usually seen as add-ons, curricularly dwarfed by pathology, biochemistry, surgery, cardiology, obstetrics, pediatrics, and the like. These straight-science fields are the solid purviews of doctors, whereas those softer skills, well, anyone could use them. And, in fact, many other practitioners do. Alternative healers of all stripes make heavy use of these skills, which explains the surging popularity of alternative medicine. It also explains much of their effectiveness: as the physical therapy study demonstrated, much pain can be relieved with just good communication, empathy, and connection.
Herein lies the rub: something that simple and intuitive, something that doesn’t require specialized knowledge, can feel threatening to a physician who has spent a decade training to acquire unique medical knowledge (and $100,000 worth of student-loan debt). I mean, it’s not a lot of sweat off anyone’s back to be a more engaged communicator, and if it doubles your patient’s pain relief—why not? It’s the very simplicity, I think, the lack of sweat off one’s back, that makes it seem so squishy and simplistic to we doctors who frame ourselves in the scientific mold. There’s something vaguely discomfiting to realize that the techniques shamans used centuries ago can sometimes be as effective as our pharmaceuticals backed by million-dollar mega-trials.
How can we help doctors overcome their reflex hesitation and feel comfortable adopting some of these simple but effective techniques? I put this question to Richard Street, a professor of communications at Texas A&M University. He readily acknowledged the difficulty of getting buy-in from doctors, and so whenever he gives lectures to doctors, he presents his research within a strictly scientific framework, emphasizing rigorous research studies. But then he flipped the question back to me, knowing that I am an ever-struggling student of the cello. “To become a musician,” he said, “you need to acquire all the technical skills, right? The notes, the chords, the scales. This is the science of music. But when you play music, especially when you improvise, this is the art of music.”
When Street described the doctor-patient interaction as an improv, the metaphor really hit home. While I don’t improvise on the cello—I’m still trying to get a handle on the stuff that’s already been written, thank you very much—I’m in absolute awe of musicians who can improvise. As I’ve learned more about music, I’ve come to understand that improvisation is less about making it up on the spot and more about reaching into your bag of tools and putting these tools together in new and creative ways. There is, of course, plenty of off-the-cuff artistic inspiration, but improvisers are generally utilizing well-honed musical techniques. Astute listeners familiar with basic music theory can pick out these tools as they spontaneously erupt—arpeggios, inversions, diminished seventh scales, syncopation, theme and variation, classic riffs. We tend to think of this as the purview of jazz musicians in performance, but Bach, Mozart, and Stravinsky were improvising as well; they were just doing it on the page rather than on stage.
And so it is with doctors and patients. Some encounters are straightforward but many are unpredictable. “Each person,” Street said, “is responding to the other person’s moves.” Both are improvising to a certain degree. The wise doctor—and the wise patient, as well—will reach into her bag of tools and extract what is most needed for that particular person at that particular time. Communication techniques are an additional set of tools to add to the bag to render the selection bigger and ultimately more useful. Corine Jansen is one person who has been vigorously expanding that bag of tools, and from a bracingly low-tech perspective.