How Doctors Think is a collection of articles aimed at understanding how doctors think, especially when they make mistakes or avoid the mistakes of others. It turns out that most doctors think pretty much the same way most of the time and this works out well for patients most of the time.

Medical students are taught to do a Bayesian analysis on data gathered from a patient's history and a physical exam. Statistical probabilities determine what tests to run. Statistical probabilities determine what diagnosis to make. Doctors also use visual and tactile impressions, made before any physical exam, to help with the diagnosis. Data from referring physicians is also taken into account. Overall, this works well as long as the patient is not atypical and as long as the physician doesn't use these heuristics in a dogmatic, hidebound way. A patient who has been misdiagnosed for years by several physicians poses a special problem. It takes a rare bird who can see this patient and suppose for the sake of argument that she isn't demented and that many competent physicians have been wrong about her diagnosis. Groopman provides such a case. The patient had been diagnosed by some thirty doctors over a period of fifteen years as having bulimia and anorexia nervosa, as well as irritable bowel syndrome. She actually suffered from celiac disease, an autoimmune disorder (an allergy to gluten) that causes an irritation and distortion in the lining of the bowel, making it nearly impossible for nutrients to be absorbed. May we all be so lucky as to have a doctor like Myron Falchuk, who came up with the correct diagnosis, as our physician.

Some of Groopman's articles seemed familiar to me. I eventually realized that I had read them before in The New Yorker.  One of his New Yorker articles waxed poetic about Vioxx and Celebrex, which he now realizes he celebrated more out of wishful thinking than anything else. Remembering one's mistakes and identifying why you made them is one of the key messages Dr. Groopman has for doctors.

Groopman wrote the book mainly for laypeople but many doctors will benefit from it, too. The main message to patients is that we should play a crucial role in our own diagnoses. Doctors, he says, "desperately need patients and their families and friends to help them think. Without their help, physicians are denied key clues to what is really wrong." One of the more interesting features of Dr. Groopman's book is that several of the stories he has to tell involve his own medical misadventures, both as a doctor and as a patient. In addition to back problems, Dr. Groopman had some severe problems with one of his hands. If he applied too much pressure to his right hand it became beet red and an "excruciating pain erupted in my right wrist" below the thumb. X-rays revealed cysts. He describes the different diagnoses and recommendations of several doctors, all surgeons. Dr. A ordered an MRI and found the cysts and "the rope-like tendons suspended in a sea of fluid." He had no diagnosis and sent him off with a splint. Groopman went to A for a year. Dr. A did blood tests to see if there was "some underlying disease, like lupus or rheumatoid arthritis," causing his hand problems. He injected the wrist with steroids to no avail. Finally, after a year, Dr. A said that he thought Groopman had "developed a hyperreactive synovium" and recommended surgery to remove the lining of the joints around the wrist. Another hand surgeon Groopman consulted had never heard of "hyperreactive synovium." Dr. A seems to have invented it.

Dr. B agreed that "hyperreactive synovium" is not a real clinical condition. Dr. B told Groopman he had a hairline fracture and needed three surgeries and it would take about two years to recover. Dr. Groopman then pulled some strings and got an appointment with Dr. C, "one of the most renowned hand surgeons in the United States." (I laughed out loud at Dr. Groopman's description of Dr. C's waiting room. The walls were filled with plaques, each attesting to Dr. C's fame. One was from Rio de Janeiro for an International Conference on Abnormalities of the Thumb; another was from Saint Moritz, Switzerland, for Repair of the Rheumatoid Finger.)

Reading about Dr. C, I was reminded of how some physicians may get a reputation for being wizards at making diagnoses in a minute or two when in fact they're performing like a cold reader might. They get a few clues from the patients, a referral, an x-ray, an MRI, etc., and make a pronouncement: we need to do this this and this. His assistants hop to and get the patient to sign the forms to do this this and this. The Dr. does this this and this and no matter what the outcome, the genius made the right diagnosis and this this and this was exactly what was called for. Unfortunately for Dr. C, Dr. Groopman brought his wife (also a physician) along for the exam and wouldn't sign the consent form to do an arthroscopy. Instead, the Groopmans queried Dr. C's resident about the procedure while Dr. C scooted off to see more patients. Eventually, the Groopmans got the resident to get Dr. C to return and answer some questions. Dr. C said he thought the "leading diagnosis" was chondrocalcinosis, aka pseudogout. The resident had already informed Dr. Groopman that Dr. C's main way of handling bad hands was to fuse the bones, the same procedure he had had done to his back and which he was convinced had been a bad move on his part.

A year later, still suffering, Groopman consulted Dr. D, who did an exam and had x-rays done on both hands, stationary and flexed. There was a difference that led to the diagnosis that the ligament between the scaphoid and lunate was partially torn, which caused friction between the bones. "He went on to explain that there also could be channels from the cysts into the joint, so that they functioned like lakes with thin canals: as the fluid in the cysts came under pressure, the liquid would be squeezed out through the canals into the joint. This set off the inflammation." MRIs, it turns out, "find abnormalities in everybody." A fifth surgeon, Dr. Terry Light, agreed with Dr. D. Thinking back on his experience with the thinking of these doctors, Groopman comments: "A surgeon's brain is more important than his hands."

Dr. A was guilty of confirmation bias. Dr. B erred by stopping the search as soon as he found an explanation that made sense to him. He shut down his thinking and didn't explore further possibilities. He made a hasty conclusion. So did Dr. C, although he wanted to do a favorite procedure of his, arthroscopy, to explore other possibilities. On the bright side, Dr. D explored the issue in an innovative and creative way that can't be taught. His method was a lot less invasive than what Dr. C proposed, although it is possible that Dr. C might have found what Dr. D did.

Modern medicine may be evidence-based and a science, but doctoring is an art. The key to good doctoring is not only knowledge and experience, but good communication with the patient. Too often, in doctoring and most other professions, good communication means the ability to get your point across clearly and effectively. You want to be understood and be able to persuade others to agree with you. This is not what Groopman is concerned with. He is concerned that a doctor communicate with a patient in the same way that a good defense lawyer hopes her client will communicate with her: disclose everything and let the professional determine its relevance and importance to the defense. A good doctor communicates with the patient in such a way that the patient will reveal everything about her illness. Such communication requires trust. Confidentiality, not a topic covered by Groopman, is essential to gaining trust in both law and medicine. So is confidence in the professional, a topic Groopman returns to several times in his book. Gaining confidence in a client or patient is difficult if either dislikes the other. You don't have to be pals with your doctor, but if you are turned off by his attitude, speech, or behavior, ten out of ten doctors should recommend that you change doctors. To doctors, Groopman advises something akin to what the ancient Greeks carved into their temples: know thyself. Your negative emotions toward a patient or type of patient can block your ability to think critically about a patient's illness. Like the rest of us, doctors tend to like healthy people and, like many of us, are averse to the very ill.

If you've ever wondered why Bernie Siegal, Deepak Chopra, and other doctors quit practicing medicine and turned to giving talks and writing books about hope and other cheery things, it may be because they got sick of sick people. Why has Andrew Weil, M.D. never practiced medicine? Maybe for the same reason. Maybe he doesn't like to be around sick people.

Groopman cites favorably the work of Debra Roter and Judith Hall on communication between doctor and patient, especially in tough cases. Hall says that medical "competency is not separable from communication skills." How a doctor asks questions and how a doctor responds to a patient's emotions are key to "patient activism and engagement," which are both essential to quality medical care. A patient must feel "free, if not eager, to speak and participate in a dialogue to get clues about the medical enigma before him. If the patient is inhibited, or cut off prematurely, or constrained into one path of discussion, then the doctor may not be told something vital."  Groopman claims that "physicians interrupt patients within eighteen seconds of when they begin telling their story."

How many medical schools include a course in communicating with patients? How many of those courses focus on gaining the trust and confidence of the patient in order to encourage the patient to reveal as much information as possible about her condition? How many doctors think this skill is essential to providing good medicine? Groopman does and I think most doctors who read his book will agree with him.

Beside good listening skills, Groopman emphasizes the need for doctors to be good critical thinkers. One of the key skills of critical thinking is asking the right questions. The most important questions to ask, he thinks, are: What might I be missing in this case? And what would be the worst thing that could be missed? These are questions asked by one of Groopman's models of good diagnostic thinking: Dr. Myron Falchuk.

Another model of good thinking for Dr. Groopman is the cardiologist who told him that he keeps a log of all his mistakes. He revisits the log "when trying to figure out a particularly difficult case." It's not enough just to keep a record of your errors. You have to be able to critically evaluate them. If you don't admit your mistakes, you'll never bother to figure out why you made them, what went wrong, and how to avoid the same kind of mistake in the future. But without good critical thinking skills, a log of one's errors won't do much good. Much of what Dr. Groopman advises doctors to do are the kinds of things students should be taught in an introductory critical thinking course. You need to know the pitfalls of thinking: understand the affective, perceptual, and cognitive biases and illusions that plague each of us. Knowing these things is a necessary condition for critical thinking but it surely isn't sufficient. Knowledge and intelligence are valuable assets, but without an open mind they are of little value to critical thinking. We can teach people how to avoid mistakes in thinking but I am not sure we can teach people to think as creatively as some of the doctors that Groopman admires. What we can do is teach critical thinking in a way that makes it possible for the creative mind to function at a high level. But such thinking requires time and time is often what many doctors don't have enough of. Some doctors might think they don't need much time to think. But the doctor who thinks he is so superior to others that he can rely on his intuition and instinct alone to guide him is dangerously self-deceived.

Groopman is not out to bash his fellow doctors. When he notes that from 10 to 15 percent of medical diagnoses are wrong, his goal is to make us aware that doctors can do better. Improving the number of correct diagnoses will not be easy as medicine gets more and more complicated, but that is no reason not to try. He makes his point in a dramatic way by opening the book with the story of Anne Dodge. Over a period of fifteen years she was misdiagnosed by some thirty doctors. Finding out why such things happen is not an attack on physicians, but an attempt to make them better.

* * *

While reading How Doctors Think it became clear to me why I am critical of alternative medicine (AM) and always prefer a scientific medical doctor to an AM practitioner. It's not because alternative practitioners are stupid or ignorant and medical doctors are not. In fact, many alternative practitioners have medical degrees from respected medical schools. It's not that AM folks are not sincere and don't have mostly satisfied customers. From what I can tell, most people who go to AM practitioners are pleased with the care they get. It's not because medical doctors never make mistakes or AM folks make too many errors. M.D. does not stand for medical divinity. As noted above, doctors are fallible and subject to all the affective, perceptual, and cognitive illusions that plague the thinking of every human being. As Dr. Groopman produced case after case of medical misjudgment and error, it occurred to me that there is absolutely no parallel in the alternative medicine world. In scientific medicine, when errors are made in diagnosis or treatment, logic and critical thinking are applied to a large body of empirical knowledge in an attempt to discover the errors and prevent them from happening again. I have never come across anything in the AM literature that indicates that a similar process exists in AM. Scientific medicine can discover that a practice, such as a particular type of knee or heart surgery, is actually just a placebo. AM can never discover that its practices are providing placebos because that would be the end of the AM practice. Why? Because placebos are generally the only thing AM provides. And since placebos are what AM offers, very few of its treatments, remedies, or techniques will be rejected outright for not working. They may be expanded or extended or modified, but they got their place in AM because they "worked" and they didn't harm people. In medicine, even if something "works" it may be replaced when it is discovered that its biological basis is wrong (e.g., radical mastectomy and treatment of ulcers).

I don't think there will ever be books with titles like How Homeopaths Think or How Acupuncturists Think or How Chiropractors Think. These AM folks claim to treat the person, not the disease, but they see the person as fitting into a template. The homeopath looks for symptoms and consults the book for the potion that matches. The acupuncturist looks for symptoms and consults the book for the points on the meridians to stick the needles. From I what I understand, there is no way to predict what any given chiropractor might do with the same patient. The chiropractor might recommend spinal manipulation, applied kinesiology, chelation, or who knows what. If you have a viral or bacterial infection that is causing your symptoms, you may have the misfortune of selecting a chiropractor who doesn't believe in germs. Worse, there is no amount of logic or critical thinking that can be applied to the body of chiropractic knowledge that could lead to a determination of why a particular therapy didn't work and why another therapy might. Finally, can anyone imagine a book called How Angel Therapists Think?

Fortunately for most sick people, they don't seek an aromatherapist or a quantum healer when they are really sick (but see). No chiropractor or acupuncturist can treat cancer or diabetes or myocardial infarction effectively. No quantum healer can treat myelodysplastic syndrome or pharyngitis unless she uses scientific medical techniques.

One reason I distrust AM is that each of the various belief systems behind the various practices is anchored by metaphysical notions that don't fit with our scientific knowledge of how the body works. There is a lot of magical thinking in AM, false beliefs about potential causes of illness and cures such as the wrongheaded idea that a person's thoughts cause illness or cure disease. This notion is poppycock and seems to be a distortion of the commonsense truth that beliefs and attitude affect health, which is a world away from the notion that these are the only things that affect health or that they are sufficient to bring about disease or recovery. (Groopman notes a case where the belief that talking about cancer would cause it to occur prevented a patient from talking openly about her feelings with her doctor.)

Magical thinking characterizes AM but it is something that affects all of us, including scientific doctors. Groopman has some interesting examples of medical doctors who seem to believe in the effectiveness of a drug or treatment mostly on faith rather than evidence. Before reviewing a couple of Groopman's examples, I want to preface my remarks by noting that they are examples of bad thinking made possible by the way scientific medicine works. They would be recognized as examples of bad thinking by most doctors and they exemplify the difficulty in overcoming cognitive illusions. These types of errors have analogues in AM, but they won't be seen as errors because they will fit into the faith-based metaphysical system of the AM practitioner. For example, when an AM therapist starts advocating something like The Secret and incorporates it into his or her practice, nobody in AM sees this as an error to be corrected, even if it is apparent that the practitioner has not studied and tested the claims made in The Secret and is accepting what it says because "it just seems right" or it "fits with everything else I believe." When an AM practitioner takes something on faith, it is expected and not seen as a fault.

Chapter 9 is called "Marketing, Money, and Medical Decisions." The title alone will fill the heart of the Big Pharma/AMA conspiracy theorist with joy. It has a story about Groopman's own travails concerning treatment of his back pain. The chapter also contains a couple of stories about doctors who put their faith in hormone replacement therapy for men and women, even when the evidence did not support the practice.

"Spinal fusion may be the radical mastectomy of our time," writes Groopman. Radical mastectomies were common for about a century even though the procedure was based on a misunderstanding of the biology of tumor cells in the breast. The practice started in 1895 and it wasn't until the 1980s that "it had become clear that tumor cells can spread throughout the body early in the disease [breast cancer] through the lymph channels and blood vessels." A lumpectomy and radiation are just as effective as the disfiguring procedure of a mastectomy. Today, spinal fusion is common. More than 150,000 lower lumbar spine fusions were done in 2006, according to Groopman. He's had the operation himself. The procedure is done for chronic low back pain and involves removing discs and bracing the vertebrae with metal rods and screws. The procedure is also done on patients with fractured spines or spinal cancer. For those conditions, Groopman thinks the procedure is appropriate. However, they make up a miniscule number of the total cases involving this procedure, he says. Groopman thinks that most spinal fusion procedures are unnecessary and that there is no substantial body of medical evidence supporting the procedure as a standard treatment for chronic lower back pain.

Lower back pain is the bread and butter of chiropractors, of course. They bank on the same facts that spinal fusion surgeons bank on: the source of low back pain cannot be determined in 85 percent of the cases and 90 percent of the cases of acute low back pain "improve within two to seven weeks without specific therapy" (p. 226). Even most cases  of ruptured discs (some 80 percent) don't need surgery. The disc will retract on its own, stop pressing on the nerves, and the inflammation will subside. With medical doctors, how a patient is treated for chronic lower back pain "may be significantly influenced by economics," say Groopman. The same is probably true of chiropractors, by the way, unless they are completely ignorant of the research on back pain and recovery. "For the majority of patients with chronic lumbar pain, fusion surgery has no dramatic impact on either their pain or their mobility" (p. 228). Research shows that those who have spinal fusion do only marginally better after two years than patients who have had intense physical therapy.

Those who think the AMA, Big Pharma, and the government are conspiring to promote useless or harmful medicine while stifling alternative healing techniques should review what happened to the federal Agency for Health Care Policy and Research in 1993. Details are provided by Groopman (pp. 229-230). Briefly, what happened was that a doctor on the panel published a study that suggested that spinal fusion lacked a scientific rationale. The North American Spine Society attacked the panel as being biased and lobbied Congress to cut off funding to the panel. The AMA, the American College of Physicians, and the American Hospital Association tried to save the agency but the House zeroed out its budget.  The Senate revived it, but with a dramatic cut in funding. A company that manufactures the metal rods and screws used in the surgery sought a court injunction to block publication of the agency's findings. "The guidelines that were eventually published emphasized conservative measures like physical therapy, but the controversy surrounding the panel tainted its credibility, and its recommendations have had little impact on surgical practice" (p. 230).

Can you find satisfied customers of chiropractors or of people who have had spinal fusion? Of course. Does that mean the practices are good for standard treatment of lower back pain. No, of course not. The bottom line seems to be that who you see will determine what you get.

In a free-market society, patients cannot rely on government or health practitioners to protect them from unnecessary or harmful treatments. Unfortunately, most patients who are severely ill feel helpless and hope their doctor will take over. Such an attitude is not really an option if a person wants a proper diagnosis and treatment. We have to be proactive and knowledgeable, unafraid to ask questions or to seek second or third or fourth opinions. We must be willing to do research on our own. For routine ills, the patient's involvement can be minimal. But if our illness is complicated, a wrong diagnosis can mean an early death. Groopman returns frequently to the theme of the patient being a partner with the doctor in finding out what is wrong. If the patient is too sick to help, friends and family members must step in. (Groopman doesn't discuss mental illness in any detail but I can tell you from experience that in California it is almost impossible for friends and family members to help a mentally ill person who doesn't want to be helped. Doctors can't talk to you because of confidentiality requirements and, in my experience, they don't seek you out to get information about the patient.)

One problem with a free-market society, when it comes to health care, is that diseases and treatments can be created by marketing. The drug companies have medicalized the aging process to the point that they have convinced not only the general public but many of the doctors who treat us that aging is bad and should be avoided at all costs. When I hear of someone like Sylvester Stallone singing the praises of HGH (human growth hormone) or anyone else touting other hormone therapies as a "fountain of youth," I wonder how far they will go in deceiving themselves. Their reasoning is flawed. Just because the body stops producing certain hormones as we age, it does not follow that taking injections of testosterone or estrogen, or whatever else we are not producing as we did when we were twenty, is going to do a body any good. The "fountain of youth" feeling is little more than a placebo effect. What the hormones are actually doing to the body may be quite contrary to what the user thinks is happening.

Medical doctors like Dr. Groopman and Dr. Karen Delgado, an endocrinologist, are concerned about the medicalizing of normal changes and challenges in life. It may disturb some people to find out that the growth in prescriptions of estrogen for postmenopausal women was due mainly to Dr. Robert A. Wilson's book Feminine Forever, a book that was commissioned by a drug company that made estrogen. The book is now seen as "a marketing manifesto, not an objective clinical treatise" (p. 210). The marketing occasionally precedes adequate clinical research in scientific medicine. It almost always precedes the research in AM, if, in fact, any research ever gets done. The typical process in AM is for a device or treatment to be marketed as a wonderful treatment for this or that, along with a challenge to others to prove otherwise. Magnetic devices come to mind. The marketing had already established in many people's minds that magnetic therapy works before any clinical trials had been done. The same is true of every herb advocated by Ayurvedic healers. Do you think Deepak Chopra has done any clinical trials on the herbs he claims keep him eternally youthful?

Medicalizing normal life challenges is something AM healers should be familiar with since they tend to dramatize everything into some sort of need for healing. One might say that AM healers have medicalized living. They've created entire lifestyles devoted to communicating with angels or channeling energies. Everyone and everything is sick and wounded and needs healing. Life is killing us!

Several  things became clear to me as I read example after example of erroneous or arrogant thinking by doctors. 1) If you want to live a long time, don't get sick. 2. If you do get sick, don't come down with something unusual or rare. 3. If you do come down with something rare, hope some doctor somewhere can figure it out. 4. If you find a doctor who figures out what's wrong with you, hope there is a meaningful treatment for it. 5. If there is no meaningful treatment for your rare disorder, hope you can make the best of whatever time you have left.

I'll finish this review with a thought that occurred to me while I was reading the chapter on perceptual and cognitive errors that can occur in reading x-rays, CT and MRI scans, EKG tests, pathology slides, etc. Radiology is an imperfect art. Disagreements among radiologists about a film being "normal" averages about 20 percent. Radiologists who read the same film on different days will disagree with themselves 5 to 10 percent of the time. Radiologists can be victims of inattentional blindness and focus on what is present rather than on what is absent. When asked to read a chest x-ray of a man missing his left clavicle, 60 percent of the radiologists didn't notice the missing clavicle. Even the top diagnostitians have an error rate of about 5 percent. These facts made me think of all those cases where people claim miracles have occurred because a tumor that showed up on an earlier film wasn't there in a later film. A lot of miracles may simply involve bad film or bad readings of film, or bad pictures due to the patient slightly moving or not holding his breath.

Medical miracles may be due to medical mistakes but good doctoring, which may seem miraculous at times, is due to good thinking.

Robert Todd Carroll
February 26, 2008

more book reviews by R. T. Carroll