When Doctors Don’t Know What’s Wrong

The first patient I ever saw as a first year resident came in with a litany of complaints, not one of which I remember today except for one: headaches. The reason I remember he had headaches isn’t because I spent so much time discussing them but rather the opposite: at the time I knew next to nothing about headaches and somehow managed to end the visit without ever addressing his headaches at all, even though they were the primary reason he’d come to see me.

Then I rotated on a neurology service and actually learned quite a lot about headaches. When my patient came back to see me a few months later, I distinctly remember at that point not only being interested in his headaches but actually being excited to discuss them.

I often find myself thinking back to that experience when I’m confronted with a patient who has a complaint I can’t figure out, and I thought it would be useful to describe the various reactions doctors have to patients in general when they can’t figure out what’s wrong, why they have those reactions, and what you can do as a patient to improve your chances in such situations of getting good care.


I’m going to begin with my conclusion: All the problems described below are exacerbated, if not in some instances created, by the current model in which most primary care physicians practice, the fee-for-service model (that is, the PPO network model). In the fee-for-service model, your PPO network primary care physician is only paid when he or she performs a service—that is, when he or she sees you in his or her office. Because reimbursement in PPO network primary care practices has remained relatively constant for the last twenty years while expenses have continued to rise, PPO network primary care physicians have had to pile scores of patients into their practices to survive financially. Currently, in PPO network fee-for-service primary care practices, patient panel sizes run on average 2,300 patients per physician. This has brought us to a reality in which the average length of time a primary care visit lasts is now only 16.5 minutes. Little wonder so many patients aren’t getting the care they need and have become frustrated with their healthcare experiences in PPO networks.

In contrast, a new model of care called direct primary care, or concierge medicine, has physicians typically caring for no more than 600 patients. This dramatically reduced panel size enables patients to enjoy 24/7 access to their primary care physician and same-day or next-day appointments. Instead of the average of 16.5 minutes spent with each patient in traditional fee-for-service primary care practices, in direct primary care, or concierge medicine, patient appointments can be blocked at hour-long intervals—or longer.

This improvement in access results in improvement in quality and dramatically increases the likelihood that physicians will be able to help their patients. First, blocking appointments at hour-long intervals ensures that physicians will have enough time to take comprehensive patient histories, perform thorough, focused exams, read through the relevant medical literature, and think critically. The inability to do this in the fee-for-service system is itself responsible for much of the poor care patients receive. Second, whether in the middle of the night or during regular business hours, having immediate and direct access to your own primary care physician who knows you and your medical history results in more appropriate testing, reduced use of ancillary services (ER and specialty care), fewer administrative hurdles for patients, greater diagnostic accuracy, and streamlined follow-up. Multiple studies confirm that this kind of intense, upfront primary care does indeed translate into an improved quality of care, that patients who receive care from primary care physicians who are able to provide timely and thorough access to their patients are not only healthier but are also, in fact, more likely to live longer.

While the problems described below exist in all models of care, they’re far less prevalent in direct primary care, especially here at ImagineMD where our physicians are specifically trained to avoid them.


Believing a wacky idea isn’t wacky in and of itself. Believing a wacky idea without proof, however, most certainly is. Likewise, disbelieving sensible ideas without disproving them when they’re disprovable is wacky as well. Unfortunately, patients are often guilty of the first thought error (“My diarrhea is caused by a brain tumor”) and doctors of the second (“brain tumors don’t cause diarrhea, so you can’t have a brain tumor”), leading in both instances to contentious doctor-patient relationships, missed diagnoses, and unnecessary suffering. Doctors sometimes aren’t willing to order tests that patients think are necessary because they think the patient’s belief about what’s wrong is wacky; they sometimes suggest a patient’s symptoms are psychosomatic when every test they run is negative but the symptoms persist; and they sometimes offer explanations for symptoms the patient finds improbable but refuse to pursue the cause of the symptoms any further.

Sometimes these judgments are correct and sometimes they’re not—but the experience of being on the receiving end of them is always frustrating for patients. However, given that your doctor has medical training and you don’t, the best you can sensibly hope for are judgments based on sound scientific reasoning rather than unconscious bias. Unfortunately, though, even the minds of the most rational scientists are teeming with unconscious biases. So a more realistic strategy might be to attempt to leverage your doctor’s biases in your favor.

Expert vs. Novice Thinking

In order to do this, you first need to know how doctors are trained to think. Medical students typically employ what’s called “novice” thinking when trying to figure out a diagnosis. They run through the entire list of everything known to cause the patient’s first symptom, then a second list of everything known to cause the patient’s second symptom, and so on. Then they look to see which diagnoses appear on all their lists and that new list becomes their list of “differential diagnoses.” It’s a cumbersome but powerful technique, its name notwithstanding.

A seasoned attending physician, on the other hand, typically employs “expert” thinking, defined as thinking that relies on pattern recognition. I’ve seen carpal tunnel syndrome so many times I could diagnose it in my sleep—but only learned to recognize the pattern of finger tingling in the first, second, and third digits, pain, and weakness occurring most commonly at night by my initial use of “novice” thinking. The main risk of relying on “expert” thinking is early closure—that is, ceasing to consider what else might be causing a patient’s symptoms because the pattern seems so abundantly clear. Luckily, in most cases, it is clear.

But sometimes it isn’t. In those cases, your doctor may do one or more of the following:

    1. Revert to “novice” thinking. Which, in fact, is completely appropriate. We’re taught in medical school that approximately 90% of all diagnoses are made from the history and physical exam, so if we can’t figure out what’s wrong, we’re supposed to go back to the patient’s story and dig some more. This also involves reading, thinking, and possibly doing more tests, for which your doctor may or may not have the time or stamina if he or she isn’t practicing in a direct primary care practice.
    2. Ask a specialist for help. Which requires your doctor to recognize he or she is out of his or her depth and needs help.
    3. Cram your symptoms into a diagnosis he or she does recognize, even if the fit is imperfect. Though this may seem at first glance like a thought error, it often yields the correct answer. We have a saying in medicine: uncommon presentations of common diseases are more common than common presentations of uncommon diseases. In other words, presenting with a set of symptoms that are unusual or atypical for a particular disease doesn’t mean that you don’t have that disease, especially if that disease is common. Or as one of my medical school teachers put it: “A patient’s body frequently fails to read the textbook.”
    4. Dismiss the cause of your symptoms as coming from stress, anxiety, or some other emotional disturbance. Sometimes your doctor is unable to identify a physical cause for your symptoms and turns reflexively to stress or anxiety as the explanation, given that the power of the mind to manufacture physical symptoms from psychological disturbances is not only well-documented in the medical literature but a common experience most of us have had (think of “butterflies” in your stomach when you’re nervous). And sometimes your doctor will be right. However, the diagnosis of stress and anxiety should never be made by exclusion (meaning every other reasonable possibility has been appropriately ruled out and stress and anxiety is all that’s left); rather, there should be positive evidence pointing to stress and anxiety as the cause (e.g., you should actually be feeling stressed and anxious about something). Unfortunately, doctors frequently reach for a psychosomatic explanation for a patient’s symptoms when testing fails to reveal a physical explanation, thinking if they can’t find a physical cause then no physical cause exists. But this reasoning is as sloppy as it is common. Just because science has produced more knowledge than any one person could ever master, we shouldn’t allow ourselves to imagine we’ve exhausted the limits of all there is to know (a notion as preposterous as it is attractive). Just because your doctor doesn’t know the physical reason your wrist started hurting today doesn’t mean the pain is psychosomatic. A whole host of physical ailments bother people every day for which modern medicine has no explanation: overuse injuries (you’ve been walking all your life and for some reason now your heel starts to hurt); extra heart beats; twitching eyelid muscles; headaches.
    5. Ignore or dismiss your symptoms. This is different from the application of a “tincture of time” that doctors often employ to see if symptoms will improve on their own (as they often do). Rather, this a reaction to being confronted with a problem your doctor doesn’t understand or know how to handle. That a doctor may ignore or dismiss your symptoms unconsciously (as I did with my first-ever patient) is no excuse for doing so.
A Doctor’s Biases

Just which of the above approaches a doctor will take when confronted with symptoms he or she can’t figure out is determined both by his or her biases and life-condition—and all doctors struggle with both. To obtain the best performance from your doctor, your objective is to get him or her into a high a life-condition and as free from the influences of his or her biases (good and bad) as possible.

Negative influences on a doctor’s life-condition include all the things that negatively influence yours, as well as the following things that may happen to them on a daily basis:

    1. They fall behind in clinic. Your doctor may be naturally slow or frequently have to spend extra time with patients who are especially ill or emotionally upset.
    2. They have to deal with difficult or demanding patients. Hard not to enter into a defensive, paternalistic posture when too many of these types of patients show up on your schedule.
    3. They feel like they don’t have enough time to do a good job. With fewer and fewer resources, doctors are being asked (like everyone) to do more and more. But how can we expect even the most expert clinician to correctly diagnose a complex medical problem with only 16.5 minutes to spend on it?
    4. They have to deal with a morass of paperwork in a hopelessly inefficient healthcare system. The amount of time most doctors must spend justifying their decisions to third-party insurance carriers is growing at an alarming rate. One of the great benefits of direct primary care is that third-party insurance is never billed, so all the time a fee-for-service doctor had to spend on collecting revenue is now replaced with time spent directly on patient problems.

A sampling of unconscious biases that influence doctor behavior include:

    1. Not wanting to diagnose bad illnesses in their patients . . . leading sometimes to an incomplete list of differential diagnoses.
    2. Not wanting to induce anxiety in their patients . . . leading sometimes to insufficient explanations of their thought processes, which often paradoxically leads to more patient anxiety.
    3. Over-relying on evidence-based medicine. Though the practice of evidence-based medicine should be the standard, many physicians forget there’s a great difference between “there’s no evidence existing in the medical literature to link symptom X with disease Y” and “there’s no evidence existing to link symptom X with disease Y because it’s not yet been studied.”
    4. Not liking their patient . . . leading to impatience, not listening, and not taking enough time to think through the patient’s complaints.
    5. Liking their patient too much . . . leading to biases #1 and #2.
    6. Thinking a patient’s symptoms are caused by one diagnosis instead of many. Also known as Ockham’s razor, sometimes it’s true and sometimes it isn’t.
    7. Wanting to be right more than wanting their patient to get betterRes ipsa loquitur (the thing speaks for itself).
    8. Believing their first thoughts about the diagnosis are more likely to be correct than any subsequent thoughts. If your doctor is too attached to a diagnosis simply because it’s the one he or she thought of first or has seen more than other, less common diagnoses, he or she may avoid pursuing other possibilities.
    9. Failing to consider that a test result may be in error. This doesn’t happen commonly, but it certainly does happen.
    10. Wanting to avoid feeling ineffectual. Some diagnoses are more amenable to therapy than others. No patient wants to have an untreatable illness and no doctor wants to diagnose it.
    11. Having an aversion to being manipulated. Manipulation is especially common in patients suffering from chronic pain syndromes (who may at times appear drug-seeking rather than pain-relief seeking). No one likes to be manipulated, but a wise mentor of mine once said, “The question isn’t whether or not your patients will try to manipulate you. The question is how will they try to manipulate you.” Coming to terms with this truth is vital for any doctor to have successful relationships with their patients.

Unfortunately, your ability to raise your doctor’s life-condition is as limited as your ability to raise anyone else’s and even more so when you don’t feel well. Good humor, if you can muster it, may be your best option.

But in dealing with your doctor’s biases, you have on your side a fact I firmly believe to be true: most doctors want to do a good job and help their patients as best they can. So what exactly can you do to maximize your doctor’s ability to help you?

    1. Position your symptoms and requests carefully. Don’t demand medications or tests. Ask about them. Wonder about them. It’s perfectly all right to bring up research you’ve done about your symptoms, but explicitly express your openness to the possibility that your ideas might be wrong. Not that you should aim for subservience by any means, but rather for a genuine partnership.
    2. Remain reasonable even when you’re irritated. Most doctors, even when stressed, will respond to reason and reasonableness in kind.
    3. If your doctor suggests your symptoms might be due to stress, acknowledge he or she may be right. Even if you disagree. First of all, your doctor may be right, even if it doesn’t feel that way to you. Second, if you dismiss the notion out of hand, you might make your doctor defensive and therefore more likely to cling to an idea that a moment before was only one possibility among many.
    4. Ask questions that promote transparent, logical thinking. Many doctors don’t explain their thought processes clearly enough (in the fee-for-service model, they don’t have the time). Write all your questions down before your visits and ask smart questions that actually help your doctor think through your symptoms and his or her approach to working them up (“What possibilities will this test rule in or out?” “What else is on your list of possible diagnoses?”). Of course, this presumes you’re comfortable knowing the answers.
    5. Be explicit about how you want your doctor to work with you. Show them you’re interested in understanding the process of medical detective work. Position yourself as your doctor’s student. Nothing helps improve someone’s thought process like having to explain it to someone else.
    6. Ask your doctor to explain the risks and benefits of any proposed test or treatment quantitatively. Get percentages for risks and compare them to the risks of activities you tolerate every day. For instance, your annual risk of dying in a motor vehicle accident is 0.016%. You’d be surprised how many worrisome drug side effects, for example, occur at similar frequencies.
    7. Get second opinions. And sometimes third opinions. Recognize, though, that in doing so you risk ending up even more confused than you were with only one opinion. But don’t assume because your doctor doesn’t know what’s going on that no one else does either. There’s almost no way for you to be sure your doctor doesn’t know what’s wrong because he or she doesn’t know or because no one knows. Sometimes you have to go through multiple doctors until you finally find the right one with the right experience to figure out your problem. Neither doctors nor patients like to acknowledge this, but serendipity sometimes plays a role in arriving at the right diagnosis. I once figured out why a patient had been nauseated for 30 years after they’d been seen by almost as many doctors. The patient said something that just happened to make me think of an obscure diagnosis I’d never seen but had read about. I looked it up, sent the patient for a test, and found the answer.

In the past, I’ve had a small cadre of patients who suffered from symptoms more horrible than I can describe, some with known diagnoses and some without. In all cases, my ability to help them was tragically limited. Sometimes I wanted to ignore these patients. Sometimes I would cringe when they’d call—not because I didn’t like them or because they complained to me too much or because I didn’t care about them but because I had so little real relief to offer them. I knew how much my simply being present and being willing to listen meant to them (they told me this all the time) and I didn’t discount it. And I did my best to diagnose and treat what problems I could and sympathize with them when I couldn’t. But it was hard. It is hard. I must constantly be on guard not to fall under the influence of all the thought errors and biases I’ve described here. At least now, practicing in a model of care that affords me far more time to spend with patients and to think about their issues than in the fee-for-service model, I have a real chance.

Sign up to get notified when a new blog post has been published.