Recently I was rounding at the hospital with the medical residents, when the subject of DVT, deep vein thrombosis, arose. Patients with circulatory problems in their lower limbs often report calf pain, but whether the pain is thrombotic is ambiguous. In my day (and now I sound like the curmudgeon I'm rapidly becoming), we used to look for "Homan's sign": you would grasp the patient's foot and flex it upward. If the movement generated pain, it was taken to be an indication of a clot or embolus. When I mentioned Homan's sign in relation to the patient we'd just seen, one of the residents cited a recent clinical study—an "evidence-based medicine," or EBM trial—that said Homan's was predictive in only 30% of cases. The article recommended Doppler imaging to make a diagnosis, though they didn't say who, exactly, should get imaging given the mostly silent nature of DVT.
The referencing of EBM studies is common on the wards these days. The newer generation is gung-ho about the best medical evidence as drawn from these clinical experiments and quick with statistics that prove or disprove a mode of diagnosis or treatment. There's a lot to say about EBM (and many people are saying it), but for now, I'll stick to my anecdote. What I said to him about Homan's: "At least it's free."
It was a joke but not only a joke. The decision to go for expensive ultrasound imaging is an example of a major shift in our approach to patients: order a test! do a procedure! These impulses sometimes—not always—yield definitive answers, but they also yield insanely spiraling costs. My old-fashioned approach has the advantage of being non-invasive, economical, and, last but not least, friendly.
Homan's sign is only one of many traditional techniques associated with the physical exam. The time-honored agenda—palpate (touch), percuss (tap), auscultate (listen), inspect (look)—was a hands-on (ears-on, eyes-on) approach to diagnosis, relying on sensory inputs and clinical expertise. Now doctors travel light, PDAs in their pockets, but the medical bag of the past, containing reflex hammer, otoscope, and stethoscope, was bulky testimony to the physicality of the physical exam. And as the old adage puts it, "the most important part of the stethoscope is between the ears." Our minds were filled with physiological information and clinical examples; we searched for the 'why' of a murmur or a rash, and looked for classic signs when we examined patients, bringing to bear our experience of seeing many others with related findings. But ask any trainee of the current crop which is better, a mind or a computer, an anecdote or an algorithm, and they'll tell you that it's a no-brainer.
I'll leave for another occasion the question of whether EBM studies and the statistical analysis of evidence are always as reliable as they purport to be. For now, I'd like to point out what we're losing as we move toward more numerical kinds of evidence and more technological forms of diagnosis and treatment. The physical exam is becoming almost obsolete. The PCP these days is likely to do a pro forma routine, listening to the heart, checking out the eyes, ears, and throat, but not with any serious diagnostic intention. The first sign of anything even faintly unusual and, whoosh, you're off to the specialist for something expensive. Besides the quick shuttle to tests and procedures—the overdiagnosis syndrome that's been much written about lately (see, for example, H. Gilbert Welch's Overdiagnosed)—the whole nature of the office visit has changed for the worse. Instead of serious attention to the patient's narrative and respectful touching of the patient's body, the doctor is likely to be staring into a computer most of the time, assessing lab values like cholesterol, creatinine, and white cell count, and checking out blood pressure and blood sugar. The living person in front of him is barely worth a glance.
A case in point: Leslie, who I treated for kidney stones, moved to a retirement community in Florida where she has a new PCP very different from her old-fashioned one. The new one is an EBM guy all the way. He won't prescribe insulin for her if she doesn't take her statins to control her cholesterol—that kind of guy. (And we're not even all that sure about statins.) But the thing Leslie hates the most is that he never touches her. She told him that her old doctor used to feel for the pulses in her feet, to check for circulatory problems related to her diabetes; this new one said, "See your podiatrist." He's really only interested in numbers. Bodies, not so much.
In a way, we're getting it exactly backwards. Not long ago an intern said to me, after listening to a patient's heart, "Isn't it amazing that you can hear through a stethoscope what you see on an echo!!" He seemed to think the echocardiogram, fancy and expensive, was the key, rather than the stethoscope on the patient's chest. In my day, we used to breathe on the disc of the stethoscope to warm it before putting it against the patient's skin. Maybe it's an effect of my training, but I think of the physical exam as not just a diagnostic opportunity but an occasion for making a relationship. The other day I saw a man who'd come in for an evaluation of his kidney function. He preferred not to take off his shirt, so I listened to his heart through his clothing. He was wearing one of those Ralph Lauren polos and when I was done I asked if that was a polo pony embroidered on the front. He said, Yes, it was, and I said, "Well no wonder I heard a gallop." (A gallop is a kind of abnormal heart sound.) We had a laugh over this, that set him at ease and helped him talk about the difficult issues related to his failing kidney.
We don't know everything about everything in medicine, despite our randomized controlled clinical trials and our modern pharmacopia. Among the things we don't know is the effect of relationship, optimism, compassion, and touch on the course of chronic illness. The ancient laying on of hands may have more importance than we imagine in the way patients experience their bodies and their illnesses.
So, here's to Homan's sign. At least it's free.
