20 September 2010

The Physical Exam -- why bother?

There was an interesting discussion on NPR's "Morning Edition" today about the dying art of the physical exam and how some teaching doctors are trying to re-emphasize it. (NPR is cool in that almost as soon as a story is on the air, there's a readable version at their web site.)  Now I think that there's a certain amount of hyperbole over this issue, in that there's not really a serious trend to omit the exam in medicine.  But it reminded me of an interesting exchange I recently had with Dr Bob at Medrants about the limited value of orthostatics. The point seems valid on the topic of the physical exam as well as postural vital signs: both are time-consuming evaluations which have a very low yield rate in terms of diagnostic and therapeutic information.

Again, don't get me wrong, I am not saying that I don't examine my patients -- I always do, and I have a fairly consistent approach to the complaint-focused exams I perform. But with a few very prominent exceptions, I often -- usually -- learn little from the exam. What I do learn is usually negative, meaning I note the absence of certain bad things.  I see hundreds of abdominal pain patients, every one gets a careful abdominal exam, and in one or two I will find some "wow" finding that guides my treatment. I see thousands of patients with weakness, and a small fraction have a neurologic exam finding which leads me to a diagnosis.  I'm not sure the "Chest Pain" exam has ever told me anything about the patient's condition. 

So what I would say is that the exam is necessary, but it is often worthless, in that I learn nothing from it.  It's worse than useless in many cases.  Just because the patient does not have obvious peritonitis does not mean that he is not hiding appendicitis in there, and relying on a normal exam can actually have disastrous consequences.

It's funny, though, that the teaching authorities fall back on the "we love tests" trope as the reason many doctors de-emphasize the physical exam. Because I don't think that's right.  We do a lot of tests, of course, but that's beside the point.  Tests are ordered after the fact.  We skim through the exam because it's just not that valuable a tool, and it's not even the most useful thing we can do at the bedside.

My preferred substitute for the physical exam, in terms of where I get the information that leads me to the necessary treatment, is talking to the patient. It's a crazy idea, I know -- a radical development in the field of medicine. But 90% of the pertinent information I get at the bedside is from what the patient tells me.  It might be nice to perform a leisurely and thorough exam on every patient, but that takes a long time.  In the real world where I have a finite amount of time to spend with each patient, I have to prioritize and allocate my resources, and talking to the patient is always my number one priority.  This is followed by an exam, either thorough or desultory, depending on what the patient has told me and my level of concern.

Now I should point out that patients do expect to be examined, and it is important in building the rapport, especially in the ER. I have always felt the most powerful part of the exam for that purpose is the hand exam.  I always start with it -- it's very intimate.  You hold the hand, you look at the skin, the fingernails. You turn it over and caress it as you look at the palm, examine the fingertips and feel for the pulse at the wrist. The rest of the exam feels almost antiseptic in comparison -- you listen to the heart, putting a metal tool between the patient and yourself. But that initial touching at the hand is what seems to cement the relationship. So habituated am I to performing the ritual of the exam that every once in a while do I omit it (say a patient who wants a refill on some simple med) and I feel horribly guilty about it.

Is that physical contact what patients really value, though?  Hard to say. I see all the patient complaints for our 100,000+ visit ER and "The doctor never examined me" is an infrequent complaint.   Maybe that's just because we do a good job of examining everybody.  A more common, almost universal complaint, however, is "The doctor didn't listen to me."  At least as a driver of patient satisfaction, the verbal part of the interaction seems to be valued more than the physical (and patients are less tolerant of failure to listen).

And many physical exam maneuvers have become outdated. Pulsus paradoxus, for example.  It should never be performed in any circumstance (at least in standard medical settings). Its value in the routine exam is nil -- literally nil -- because it's only going to be abnormal in patients with fairly severe disease. And in patients who are sick enough to have an abnormal pulsus paradoxus, there are many other ancillary tests with far more resolution capability: ECG, echo, CT, etc.  Other elements of the exam are more what we call "mental masturbation."  Fun to do, intellectually satisfying, but still ultimately valueless. Consider listening to the patient's chest and noting egophony or whispered pectoriloquy.  I mean, that's cool, isn't it?  I love the applied physics involved in how those findings appear.  But what of it?  Bear in mind that these are findings of severe, late-stage disease.  They require wholesale changes in the lung tissue.  They are highly specific for consolidation but not at all sensitive.  If the patient is presenting with a respiratory complaint and is ill enough to have lung consolidation, the likelihood the patient will also be getting a chest x-ray is about 99%, and I'm getting much better information from the x-ray. In fact, my threshold for getting an x-ray should be well lower than my interest in egophony, since the x-ray is a better screening test.

In the premodern era, and in remote settings these things had great value.  When you have no other ability to gather data, then you fall back on exam (so it is good to have the skills and the knowledge). In the current setting, whether in the office or the ER or wherever, it's important to recognize the severe limitations on the utility of the exam and relegate it to its proper role as a necessary but generally less important element of the overall evaluation. 

(Note that I am assuming an acute complaint; screening physical exams are somewhat different.)


  1. Back when disco was King, Dr. Tinsley Harrison (Yeah THAT Harrison) taught me physical diagnosis. He said, "If you listen to your patients, most of the time they will tell you what is wrong with them."

  2. I haven't had a single consult from our ER this year for a "swollen joint" that actually had synovitis. Maybe the reason some don't get any useful information from the physical exam is because they don't know how to do it correctly or well. Not you, of course, but others.

    I highly agree with your history taking statement, but a skilled physical exam is just as important.

  3. I agree with you except on abdominal exam. For c/o abd pain, exam is everything for me. No tenderness=watch them for a few hours and reexam. True tenderness while the patient is being distracted=imaging. Their story doesn't help me nearly as much for this complaint, except for the mesenteric ischemia folks of course.

  4. I agree that the history is far more important than the physical exam, but the physical exam is confirmation or exclusion of the many diagnoses that ought to be flitting around at the end of the history taking. There are many times in my life where the physical exam showed me the zebra, but the zebra was originally created by the history; such as the man with fatigue and fever who thought he had influenza and actually had a flail mitral valve clearly heard on exam. The real answer is that mindless exams are as useless as mindless histories. Both need to be done thoughtfully and well.

  5. Good points, all, and as I said, you'll never make the diagnoses if you don't do the exam.

    One major caveat I should add is that the article is written from the perspective of the generalist. Specialsts who see a pre-screened population rightly have a different perspective on the matter.

  6. My take on your post is this: Where I work now,we have a tiered response with fire department apparatus (and 1-4 Paramedics on board), plus our ambulance (witn another paramedic). Generally, the fire engine arrives first & the crewmembers swarm the patient. The fire captain stands back with the clipboard & paperwork, and obtains the demographics and (most of the time) vital signs, medications and allergies. Almost never does he/she get a past medical history list.
    I'm the 'stand back and get the big picture' type of paramedic, as opposed to the 'get right in the face of the patient' type, as I find that it's easy to get sucked into the drama and esoterics of the scene. I find that, with my style, I can usually obtain much more information, not only from the patient, but from family/bystanders, etc. If I'm tied up getting vitals, etc, it's harder to interrupt that flow to get the same information.
    Plus, once I'm clear of the scene, and it's just me and the patient in the back of the ambulance, I can generally ask more pertinent, detailed questions, and often times, get a true chief complaint. There are many times where that has changed completely from a) dispatch info, and b) the initial scene impression. I agree with you, Doc. I can generally get much more info by just talking to the patinet, and asking relevant questions, than by the physical exam.
    There's always the dx by review of the medications list, too. Many patients literally don't know why they take the medications they have been prescribed.

  7. it's the obesity.

    In an era of 60+% adult obesity, the exam becomes obscured by 10 inches of interposing fat tissue. I once saw a woman with complete dislocation of her knee; did not detect it on exam because her gigantic trunks.
    assess carotid upstroke under six inches of fat?
    detect abd. pulsatile mass under 10 inches of fat?

  8. If I had a dollar for every ridiculously useless set of orthostatics the docs asked me to do I could retire 10 years early.

    Examples, GI bleeds who are tachycardic and borderline hypotensive lying down.

    "Dizzy" patients with blood pressures of 140 systolic while sitting up.

    Orthos are useful maybe at triage for rectal bleeds, pelvic pains, with borderline tachycardia to see if they're stable or unstable, a 3 or a 2 triage. You know, is that patient tachy because they're shocky or because they're hurting or because they're young or because they're on meth or whatever.

    But when they're in back its test time and then orthos are a poor man's CBC.

  9. My favorite part of your link was the note that having the patient say "ninety-nine" was a mistake; the original German, "neun-und-neunzig", was chosen because of its abundance of dipthongs, and the paper's translator went with the literal "ninety-nine", which has none, instead of picking something more diagnostically appropriate.

  10. The dying art of the physical exam? Or, there is no art in medicine, just business?

    Hmm why do I think of the idiom "Rome is burning" after reading your blog and comments?

  11. Shocked and dismayed10/09/2010 7:38 PM


    I am shocked and dismayed. As an ER doc and teacher of emergency medicine at a large, urban emergency medicine residency program, not to mention a person who believes in logic and consistency, I am shocked and dismayed at this post.

    While I agree that talking to the patient is the most important thing you can do, I think that it is obvious that examining the patient is the second most important thing you can do. When you say that you "What I do learn is usually negative, meaning I note the absence of certain bad things. I see hundreds of abdominal pain patients, every one gets a careful abdominal exam, and in one or two I will find some "wow" finding that guides my treatment", all I can say is, "well, duh." Negative findings are at least as important as positive ones.

    Do you CT every 25 yo with abdominal pain if they have a benign abdominal exam? Because that would be crazy. The negative exam is exactly what you're using as a guide for your work-up. The same thing with weakness: if there are no focal findings, do you scan the head of the 35 yo with weakness?

    Of course, much of what we were taught in medical school is bogus; but it's not just outdated physical exams like egophony or orthostatics. Much of how medicine is practiced is historical artifact, and not terribly useful. (But that's another discussion entirely).

    I would say that in approximately half of my patients, the laboratory and radiology studies that my residents reflexively order are worthless. In most of the rest, only one or 2 tests are of any use. If I have one more resident wave a WBC in my face as evidence of anything at all, I'm gonna send them to work for you in Washington.

  12. Shocked and dismayed,

    It's a fair point that a negative exam tells you something, and I agree with you entirely that shotgun tests are worthless. The point I am trying to make here is that in many cases, possibly most cases in my ER, the exam is a low-value adjunct to the history. Of course, there are cases where exam is *everything* so it can't be omitted. But in 99% of my abdominal cases, chest pain cases, and weak and dizzy cases I have a differential formulated by the time I have completed the history, and that differential does not change based on the exam. The rare case where there is a surprise of course justify performing it.

    Now, say, an asthmatic -- that's a different case, isn't it?

  13. I struggled to master heart sounds in med school. 15 years later, I'm still bad at them. Guess what? It hasn'd made a bit of difference!


Note: Only a member of this blog may post a comment.