Who knows your body best?

Amy was a bit surprised that I insisted on seeing her rather than simply ordering the MRI based on the information she provided over the phone. After I’d examined her, I agreed with her assessment. As I was filling out the order for the MRI, Amy said, sympathetically, “I know you have to comply with the bureaucracy.’’

Now it was my turn to be surprised. It hadn’t occurred to me that taking a history and examining a patient were “bureaucracy.’’ All sorts of serious conditions, including diseases of the heart, lungs, and gall bladder, can cause shoulder pain. Much as I respect her, I couldn’t just take Amy’s word over the phone. And yet I understood her frustration: Amy knew what she needed and my insistence on the office visit was, in a sense, an obstacle to her getting it.

Years ago I sat next to a retired general practitioner at a medical luncheon. We were chatting about his long career in the rural community where he cared for several generations of families. “I always found that the patients could tell me what was wrong with them - if I only listened,’’ he said. The old doctor was echoing Sir William Osler, considered the father of modern medicine. Osler, a professor of medicine at Johns Hopkins in the late 19th century, renowned for his powers of clinical observation, wrote: “Listen to your patient. He is telling you the diagnosis.’’

My luncheon companion, as well as the great Dr. Osler, hadn’t meant that doctors should accept everything patients conclude at face value. “Listening’’ isn’t merely hearing, after all.

As a doctor listens to a patient describe a symptom - a painful foot, say - he or she considers its many possible causes. Gout? A fracture? A circulatory problem? An infection? Ill-fitting shoes? Arthritis? A desire to avoid work? Anxiety?

And here’s where the doctor and patient may find themselves at odds: The patient, too, is looking for the cause of his or her symptom - but goes about it in a very different way.

The patient whose foot is hurting is, understandably, eager to figure out why as quickly as possible. The patient isn’t looking to broaden the range of possible explanations - he or she is looking to narrow it. And the availability of medical information on the Internet makes it easier than ever for a patient to attach a label to a symptom even before arriving at the doctor’s office.

In medical school I learned that the section of the history called the “chief complaint,’’ in which the clinician records the patient’s reason for seeking medical care, should always be in the patient’s own words, not in medical language. “Don’t write ‘angina,’ ’’ we were told. “Write ‘chest pain.’ ’’

Now, though, I often find patients’ “chief complaints’’ are diagnoses. “What brings you in today?’’ I ask. More and more frequently the answer is “sciatica’’ or “Lyme disease’’ or “bronchitis.’’ In addition, in many cases patients name - as Amy did - the specific test they think they need to confirm their diagnosis.

This isn’t necessarily a bad thing. Studies have shown that for some conditions, including urinary tract infections, diverticulitis, and traveler’s diarrhea, self-diagnosis can be quite accurate - especially if you’ve had the condition before. We routinely prescribe antibiotics for patients to take at their own discretion for such infections. Similarly, we assume patients can recognize when their asthma, migraines, and other episodic illnesses are flaring and medicate themselves accordingly.

Other common conditions are much less likely to be diagnosed correctly by patients. These include vaginal yeast infections, scabies, and gout.

I find in my practice that people often mislabel themselves as having strep throat, sinusitis, influenza, and shingles - diseases about which there seem to be many misconceptions.

What, then, do I do when I disagree with a patient’s self-diagnosis?

First, I remind myself that the patient and I have the same goal, even if we’re approaching it differently. We both want to know what’s wrong so we can fix it as best and as soon as possible.

Toward that end, we each have something to offer: The patient knows better than I do how they feel and how this new symptom compares with ones they’ve had before.

For my part, I usually know better than the patient the various causes of symptoms and how best to evaluate them. This is true, by the way, even when the patient is a fellow physician. The combination of medical knowledge and lack of objectivity about one’s own body can make doctors particularly poor at self-diagnosis. One of Osler’s other aphorisms was: “A physician who treats himself has a fool for a patient.’’

Amy’s MRI showed a small rotator cuff tear, for which I recommended physical therapy, not surgery.

She does know her body. I know bodies in general. Together we did a pretty good job.

Dr. Suzanne Koven is a primary care internist at Massachusetts General Hospital. She writes a monthly column about the uncertainties, dilemmas, and stories that patients and doctors share in practice. Read her blog on Boston.com/Health. She can be reached at inpracticemd@gmail.com.