by Wayne Jonas
CHAPTER 2
How We Heal
Placebo research reveals what most medical science conceals.
There is a sleeping giant in modern medical research that has yet to awaken and reveal itself fully. When it does, it will lay waste to what we think we know about healing. It plays no favorites and so is equally devastating to ancient healing claims, complementary medicine, and mainstream health care. It is called the “placebo response.” Failure to understand the importance of the placebo response led me (and all of biomedicine) down the path that contributed to the deaths of patients like Charley. Likewise, failure to use the placebo response causes us to throw out powerful treatments like the ritual used by Sergeant Martin, which could markedly help other soldiers with brain injury. This giant can subvert the good intentions behind how we deliver healing with patients every day. By failing to acknowledge the negative aspects of placebo—called the “nocebo response”—we often inadvertently harm with our treatments. Knowing how the placebo response works opens a door to healing that few in medicine enter. But you don’t have to wait. In this chapter, I will summarize what we know about the placebo response (and its underlying causes) and how you and your doctor can use it for healing. You will get an inside view of what is coming before our understanding of placebo in medicine has fully awakened.
NORMA
How was I going to tell Norma, my patient with debilitating arthritis, that she had been taking the placebo? She had gotten remarkably better in almost all ways. She remarked to me many times during the study on how well the vitamin was working. She had less pain, was more active, and had returned to her volunteer job at the hospital. She was happier. Others noticed and commented on her improved mood and ability to move. Now I had to tell her. I worried about what would happen when I did. Would she be devastated? Embarrassed? Angry? I worried she would regress to her former state of pain and limited mobility. But I was required both ethically and legally to inform her about what she had been taking.
Norma was a tall, thin woman with long gray hair who still had the sparkle of a young woman in her eye. She reminded me of a reed, easily blown about by the wind. Her psychological nature fit her physical stature to a T. She was gentle and empathetic. She was always willing to follow my suggestions. She was one of my most compliant patients. My fear that she would regress in her healing was based on two long-held assumptions in medicine: first, that her improvement had all been based on her own “belief” that she was getting active treatment; second, that she was a good “placebo responder,” usually thought of as someone who is “suggestible” and easily influenced by the opinions of others—especially authorities, like me, her doctor. The premise that some people are suggestible in this way has a long history in medical science. After Anton Mesmer, a German physician in the seventeenth century, claimed he could heal using “animal magnetism,” in 1797, his claim was tested by a team that included Benjamin Franklin. They used one of the first double-blind testing methods, in which patients did not know if they were getting the real treatment or a fake version and physicians did not know which patients received the real treatment. One method was placing a blanket or curtain between the therapist and the patient. Patients were told that they were being treated at times when they were not. Other patients were blindfolded so they could not see the therapist or what he was doing. Franklin reported that patients would respond to the suggestion of treatment and this response occurred even when no treatment was given.
This idea that belief and suggestibility were key factors in many patients’ healing eventually led to the use of blinded tests of other therapies to see if their efficacy was real. Double-blind methods were first applied to “alternative” treatments like homeopathy by a skeptical medical profession. Eventually these blinding approaches were used for conventional treatments, too, especially new drugs. Soon the double-blind method became accepted as the gold standard for determining whether a treatment worked. All treatment effects had to be separated from the effects of belief to be considered effective.
Both Norma and I had believed she was on the real treatment. Would I harm her if I undermined that belief by telling her she was on the placebo? I thought I would, which would violate my oath to “do no harm,” as a physician. But I had to tell the truth.
I waited several weeks to inform her, hoping she would enjoy her good results for a little longer. Fortunately, during that time a way out of this dilemma emerged. The statistician who analyzed the study came back with the overall results. The vitamin had proved to be effective. When comparing the overall improvement in the group taking the niacinamide compared with those taking the placebo pill, the niacinamide group improved about 8% more than the placebo group. This was considered a significant effect; that is, it had a p-value of less than 0.05 in the statistical tests. A p-value of less than 0.05 means that if we did one hundred more studies like the one Norma was in, there would be a 95% chance we would still get at least an 8% or more improvement in the niacinamide group compared with the placebo group. It does not mean the effect of the vitamin was large (it wasn’t), only that the small effect we saw was probably real. Probably, but not for certain. To know for sure, most scientists would suggest that the study be repeated a few more times to see if the effect persists. But, at least for this study, the probability was considered high enough by scientific convention for me to tell Norma I had found a viable treatment for her. So several weeks later, when I sat down with Norma to let her know that she was taking the placebo, I could immediately tell her that the study had found the vitamin to be effective and that, if she wanted, we could switch her to the real treatment. In other words, I tried to gloss over the fact that she had gotten better because of her belief, focusing on the prospect of even greater improvement. Fortunately, she was happy with that and continued to do well on the niacinamide. I was off the hook. I chalked up the experience to Norma’s suggestibility and assumed that she was an exception rather than the rule. That is, until I met Bill.
BILL
Bill arrived in my office to seek help for his chronic back pain. He came in only at the urging of his wife, as he was skeptical that any doctor could fix him. He had been to many doctors. Finally, he agreed to come in after the pain was so bad that he had to cancel a car trip to see his grandchildren. He told me that his Korean-born wife urged him to see an acupuncturist because acupuncture is used to treat back pain in Korea. So, reluctantly, he came to see me, not because I was an acupuncturist but because he wanted to know if acupuncture could help him or if, as he said, it is “all just placebo.”
Bill is the opposite of suggestible. In fact, he doesn’t believe that any doctors or treatment can help at all. I could tell by his opinions, his body, and his body language that it is hard to influence him. He has husky round shoulders and a thick belly. He lumbers more than walks into the room and has a limp favoring his right side, which is where he feels most of his back pain.
He sat down slowly in a chair facing me and crossed his arms in front of him. He had that kind of look that says, Go ahead and try to help—I have already been through it all.
Nevertheless, he’d come in to see me. He said he did so mainly to “get his wife off of his back” and because I am a physician who practices in the military. He had been in the military and figured I won’t make money off any treatment I recommend, so I’m less likely to push something on him. I had about twenty minutes to answer his questions and see if I could help him.
I started by saying there is no easy answer, which he already knew. But then it hit me that I was saying that because he was completely different from Norma, and I, too, didn’t actually expect him to get better. I looked over the list of treatments he had tried. These included analgesics; nonsteroidal anti-inflammatory drugs (NSAIDS), such as aspirin or ibuprofen; muscle relaxants; antidepressants; chiropractic manipulation; and injections. At one point he was told to go to bed and rest. Later, he was told to get up and be more active. He was given exercises and physical
therapy. He went to a chiropractor. Fortunately, he had not been given traction (which is harmful for patients with back pain), but a couple of decades earlier he would have been given that as well. What really put him off about doctors was that he had been told to see a psychiatrist because it was “all in his head.” The psychiatrist treated him for depression (which he was sure he did not have) and then finally told Bill not to come back until he really wanted to get well. “The gall of that guy,” he told me. “Like I want to have this!”
Patients with chronic musculoskeletal pain like Bill are very common; in fact, musculoskeletal conditions are the number-one cause of suffering and the number-one chronic condition that spurs people to visit a doctor, making up over 8% of all visits per year. Back pain is the most common of those musculoskeletal conditions, affecting over 70% of all adults sometime during their life, and is the leading cause of limited activity in the world. It costs the United States over $100 billion per year. There is no measure of how much meaningful life is lost, typified by Bill’s inability to see his grandchildren. It is common for patients like Bill with chronic back pain to have undergone multiple treatments. It is common for physicians to prescribe a variety of treatments. Bill came to me because his wife made him ask me about acupuncture, but he didn’t believe in it.
“Doc,” he asked, “should I try acupuncture? Is it effective or a waste of time? I have to pay for it, because this is not going to be reimbursed by my insurance; should I spend the time, the effort, and the expense? Do I have to believe in it?”
It was a reasonable question, for which I owed him a reasonable answer. I was not sure I had one.
Acupuncture can stimulate natural painkillers in the brain, called endogenous opioids—even in animals. This makes us think the effects are real and not due to the placebo effect. Comparisons of acupuncture treatment for back pain with other treatments, such as drugs, physical therapy, and education, show that it works well. But so does sham acupuncture. This makes the effects seem largely due to the placebo effect. So even though the treatment seemed to be mostly placebo, similar to the vitamin I had tested on Norma, the downside—other than the cost and time spent on treatment—was small. So I suggested to Bill that he try it, but with a limited number of treatments, and then determine if it was working for him. I tried to keep a neutral and objective tone, trying not imply that there was not much hope for it to work. More like a personal experiment. Bill seemed to like that tone and was glad I was not an advocate for the treatment, like his wife, and that I could be objective.
I sent him to an acupuncturist I knew and trusted. After eight sessions, his pain was not much better, and he and I decided it was not worth continuing. While we gave up on the acupuncture, I didn’t want to give up on Bill. I asked him if he would like to explore other treatments, and he said he would, but there were not many he hadn’t already tried.
His X-rays showed a narrowing of the disk space in his lower spine from arthritis, so I suggested he see a surgeon. Bill was his usual skeptical self. He didn’t want to be cut on, and he had friends who had undergone surgery with little benefit. Some were even worse. But Bill had already done almost every treatment available, including intensive physical therapy. So, reluctantly, and with little belief that it would help, he had the surgical procedure. The procedure involved injection of a cementlike substance into his collapsing disk. He thought this seemed less invasive than opening his back up and fusing the disk with rods. The effects were dramatically positive. Three weeks after the procedure, his pain was the lowest it had been in years. He and his wife promptly got in the car and drove ten hours to see their grandchildren. They were very happy. Because Bill was not a suggestible person and did not believe in this treatment, I decided that this confirmed my opinion that “real” treatments were those that worked in those who did not believe. For the suggestible, placebo treatments might be more appropriate.
THE PLACEBO EFFECT
I was wrong. In 1995, I brought together a small group of investigators at the NIH who were studying why placebo seemed to work in some people and not in others. We were interested in understanding why an inert or inactive substance, such as a sugar or salt solution or distilled water with no known pharmacological value, could be effective and how often this happened. This question was popularized in a 1955 article on placebo by Henry Beecher, MD, in the Journal of the American Medical Association. Beecher reported that about one-third of all effects seen in medicine were due to the placebo response. This became medical gospel for decades, even though several studies after that reported an approximate 70% response rate to treatments that were later shown to be inert. At the meeting in 1995, Professor Dan Moerman, an anthropologist from the University of Michigan, showed findings that floored the audience. He had collected data from around the world that completely undermined the placebo gospel of Henry Beecher, the belief of most of the medical profession, and my belief that Norma and Bill had each improved for different reasons—one because of suggestibility and belief and the other because of the treatment.
Professor Moerman revealed that the healing effect from fake treatments could vary from 0% to 100%—even for the same disease and same treatment—depending on the context and cultural meaning in which they were delivered. One review, for example, studied 117 placebo-controlled trials of a drug treatment for stomach ulcers done across multiple countries. These studies showed objectively that the same inert treatment (a sugar pill) had a wide range of effects from country to country. The healing rate in Germany, for example, was very high, but in the Netherlands and Denmark it was low. In Brazil, hardly any patients with ulcers healed when given placebo. The dramatically varying results were influenced by country, context, delivery, and the patient’s interpretation of that delivery. In other words, the cultural context influenced the meaning, which in turn influenced the biology, the pathology, and the outcome. The effects were very specific. For example, in Germany, the placebo healing rates of patients with high blood pressure were low, not high as for ulcers. In fact, the meaning and context surrounding how a treatment was delivered had a much greater impact on healing than the treatment modalities themselves. Inert treatments for pain like Bill’s, for example, worked better if you gave them by needle rather than pill; gave them in the hospital rather than at home, applied them more often rather than less frequently, charged more for them rather than less, and delivered them with a positive and confident message rather than a neutral or skeptical message. Acupuncture was found to be more effective the closer the study was conducted to China, where acupuncture was developed and is widespread. I suspect surgery works better in the West, though no one has studied that. It seemed that the magnitude of a person’s healing depended less on the suggestibility and belief of the individual patient than on the collective belief of the culture and the ritual created to deliver that belief.
Professor Ted J. Kaptchuk, director of the Center for Placebo Studies at Harvard Medical School, is one of the world’s most respected researchers on the placebo response. In a recent analysis, he sheds light on the variability of these effects by comparing three types of healing encounters: Navajo ceremonial chants, acupuncture treatment in the Western world, and the biomedical provision of health care. He describes each encounter as being surrounded by beliefs, narratives, “multi-sensory dramas,” and culturally defined influences, all of which can be described as rituals in the treatment of illness. Depending on the setting and the practitioner, such rituals may take the form of communal chanting and practices led by a medicine man; the insertion of needles that takes place in an office redolent with representations of Asian culture; or authoritative white-coated clinicians presiding over complex biomedical testing and treatment technology. Looking at this research, I began to wonder: did my patient Bill get better from surgery not because it was “real,” but because surgery was more culturally meaningful to him than the other treatments he had undergone? I was skeptical of this explanation. Bill had been through many treatments and sh
ould have benefited even if they were from placebo effects. But two studies conducted after I had seen Bill seemed to contradict this assumption. In those studies, patients were randomly assigned to get either the cement or balloon injections into collapsing disks (like Bill had received) or a fake procedure that mimicked the real injections but did not manipulate the spinal disk in any way. In both studies, patients who underwent the fake procedure did just as well as those who got the real procedure.
I still found this hard to believe. Bill was resistant to treatments and was not in any way suggestible. Could it be that, at least for pain, the meaning and context of a treatment produced much of the healing, even in patients who were not suggestible? Even when “hard” procedures were used, such as surgery, that manipulated tissues and corrected anatomy? To test this assumption, my team and I did a meta-analysis of all surgery studies of chronic pain, whether in the back, knees, abdomen, or heart. We selected studies that compared real surgery to sham surgery, in which patients and doctors went through the ritual of surgery but no real correction of anatomy was done. We were able to determine the quality of the studies and then combine results into a single estimate of the contribution to healing pain from “true” surgery. The final analysis showed equally good improvement of any pain condition when the ritual of surgery was applied to the patient but no actual surgery was done. These sham surgery studies showed that, at least for pain treatments, healing occurs from something else. Could it be that the millions of surgeries done every year to treat pain produce healing because they are powerful types of ritual placebos? Could it be that the healing that occurred in Norma and Bill were not so different after all? As different as they were, could it be that they both had tapped into their own inherent healing capacity in different ways, and that healing was connected to their beliefs and behavior and to those around them more than the specific treatment they received?
