Forget ultrasound: Do a proper history and physical instead

A recent editorial in the New England Journal of Medicine lauded, albeit cautiously, point-of-care ultrasound that has risen to such an extent that it is now becoming an integral part of medical education.

Could the availability of ultrasound revolutionize clinical medicine in much the same way Laennec’s stethoscope broke the acoustic barrier?

Certainly this possibility can’t be ruled out. But I am not so sanguine. One thing I’m sure about: Indiscriminate use of ultrasound at the bedside will increase the utilization of imaging.

Ultrasound is tough. I’m not saying this as a protectionist radiologist (I don’t particularly like ultrasound), but as a radiologist who has read lots of CT scans and MRIs for what is supposedly found on ultrasound. Nearly always these findings turn out to be giant balls of “nothingoma.”

Ultrasound images look like a satellite picture of a snow blizzard in action. If you stare at them for long enough you’ll imagine all sorts of things and, much like what the boggart does to Harry Potter, these blizzards reveal the diagnosis you fear most.

I recall my training in ultrasound. For the first few weeks I saw nothing. And then I saw only pathology. Well what I thought was pathology. Until I realized there’s a fine line (or acoustic interface, if you wish to be clever) between normality and pathology. It’s a strange fact of ultrasound that appreciation of normality and its undulating and inconstant form takes longer than appreciation of frank pathology. An experienced operator is one who can boldly say “normal”.

But even seasoned operators overcall. Often I read confirmatory imaging for diagnoses raised on echocardiogram (ultrasound of the heart). These are not for trivial allegations. They include entities such as aortic dissection and cardiac mass.

In fact there’s a joke, based on a truism. What’s the most common finding in cardiac MRI done for mass? The echocardiographer’s acoustic shadow (imagination).

The thought of medical students, interns and residents flashing high frequency ultrasound probes towards the liver, aorta and pancreas in all and sundry is, quite frankly, scary. This is not because they’ll miss pathology but because they’ll find pathology where none exists. And in bulk.

And when they do they’ll have to pursue the finding. Meaning you can’t say “this might be a liver abscess, but never mind I’ll come back to it another day.” Once documented in the mind this means a confirmatory CT scan or MRI. These confirmatory scans are delightfully easy to read because they’re almost always normal and one knows a priori that they’ll likely be normal.

Look, I’m not saying that ultrasound should only be performed by experts. To be honest, anyone can become proficient in ultrasound with 3 months of practice and after having read a couple of books. Even if radiologists went around the floor scanning every patient there’d be a queue for “confirmatory” CT for liver abscess or renal cancer.

I’m saying that this innocuous instrument that boasts “trust me, I disperse no radiation” or “trust me I don’t operate at nearly thirty thousand times the earth’s magnetic field” is deceptively dangerous. Because it’s seductively innocent but can be dangerously imprecise. It sets up perfect grounds for creating Victims of Medical Imaging Technology (VOMIT), who end up having radiation nonetheless.

We’re constantly searching for innovation and disruption in healthcare in general and medical education in particular. It’s understandable to laud novelty and disdain orthodoxy. It’s, therefore, with hesitancy I propose another disruptive technology. It’s cheap and effective. The only problem is it’s been around for some time. It’s not novel. It’s an ancient art. It’s called the “history and physical examination.”

Teach medical students to perform a physical examination. Teach them about Rovsing’s sign, guarding and rigidity. Teach them to palpate pulses. Teach them to ask about history of presenting complaint. Teach them the importance of questions such as “why did you decide to come to the emergency department at midnight on Saturday when you’ve had this pain for three months,” over the robotic recital of review of systems. Teach them to organize their thoughts coherently so that they don’t sound like the audio version of Robbins Basic Pathology, but through their exposition shines insight and strategy.

These skills seem not as exciting as putting in a central line or making sense of acoustic blizzards. But they’re important if done right. They’re consequential when done wrong.

You’re probably wondering what moral basis is there for a radiologist to make this plea. So removed from the patient. So engrossed in technology. I’m not even a “proper” doctor.

You’re right.

Yet I beg: Please perform a decent history and physical. Because when this ancient art is not done right by “proper” doctors, I’m the one left to clear the debris. This costs the tax payer. This hurts your patient.

So forget ultrasound. Focus on a decent history and physical.

Saurabh Jha is a radiologist and can be reached on Twitter @RogueRad

Comments are moderated before they are published. Please read the comment policy.

  • David Gelber MD

    ah yes, the old H&P. Most of the time this simple tool actually tells me what is wrong with my patient. then I just have to figure out the best way to prove it.

  • guest

    This would be a very sensible proposal, were it not for the fact that those of us who actually see patients are increasingly seeing them in a system in which technology is being substituted for time with the patient in order to “leverage” the doctor’s (expensive) time.
    To take a proper history and do a proper physical requires more than the 15-20 minutes that the average clinician gets to spend with the average patient these days.

    • Kristy Sokoloski

      When I see a new doctor for some reason I write out the information on separate sheets of paper because it won’t all fit on the forms that they give new patients. And then I include those sheets with all the forms to give to the front desk. Then the receptionist (or whoever is responsible for doing the job) puts the information in to the computer. And when the doctor has looked at the information they thank me for the information that is so thorough. I was surprised that they took the time to actually read it beforehand. Made my day to know that they did.

      • guest

        This is a wonderful idea and I wish all patients were able to do what you do; it would really save everyone a lot of time. Another thing I love is patients’ maintaining their own personal file of medical records and bringing it in for me to look through, that’s another habit that can increase your chances of getting an accurate diagnosis and efficient care.

        • Kristy Sokoloski

          I used to keep my own personal medical record too in a binder that had been started by my mom when I was a little girl. And then I had started transferring the notes from that binder to a spiral notebook. However, in the past several years I have gotten very slack on that. I need to get back in to doing that as well because it helps me as well. The reason that I started slacking off on that too is because unless a new medicine (as one example) got added to my medication regimen most visits were all the same. Very routine so there wasn’t much to say other than that this was a regular follow-up visit and that’s it. And then if something new occurred like the example I gave then I would add that to part of the note. Because of my prior medical background (see other comments made) I tend to write these notes like I am keeping an actual chart.

          I wish more people did stuff like this as well. Because the way I look at it is that even if the doctor had the much more time that so many feel that they should have (which I agree with) sometimes it still takes several visits to solve the mystery of what is making the person sick.

          If I am able to make the job of my doctors easier when it comes to helping me then I am glad. And if my PCP sends me out to other specialists I will also check to make sure the other doctor got the information from my PCP. Sometimes it doesn’t always happen so I will call my PCP’s office to let them know what I found out and then we go from there.

  • NewMexicoRam

    “Doctor, could you inform court why you DIDN’T obtain that U/S or CT on my client? It wasn’t because your hospital does not have that technology, is it?”
    “So you’re telling me your hands, ears, and mind carry better diagnostic tools than the most up-to-date medical technology? Not to question your medical acumen, doctor, but isn’t that thinking a bit highly of yourself?”
    “Your honor, I call forth my next witness, a well renowned radiologist from a prestigious medical center, who will demonstrate why my client needed that CT or U/S in order to provide standard of care……….”

    • saurabh jha

      I had a feeling you would respond!

      I’m advocating against indiscriminate use of ultrasound, as routine, as a substitute or extension of H & P (see NEJM article), not against selective use of imaging within clinical context.

      I doubt routine US will illuminate.

      • NewMexicoRam

        I understand.
        It’s just not the current culture.
        There will need to be many, many changes to the culture first.

        • Kristy Sokoloski

          If it weren’t starting to become the current culture then why are so many saying that the stethoscope is dying? I heard that in Nursing School too. I don’t understand why it should die because ultrasound is not capable of picking up everything the way that a stethoscope can.

          • NewMexicoRam

            If the stethoscope is so great, why did they even invent the echocardiogram? We need both, but the implied intent behind the article is that real doctors shouldn’t have to depend on the new technology. There is a problem with that. It’s called “malpractice attorneys.”

          • Kristy Sokoloski

            That’s a good question because I found myself thinking the same thing when not only reading this article, but seeing the title of another blog entry about the stethoscope. May I ask why it was invented if it wasn’t something useful to helping the doctor? Also, if it’s done away with how are blood pressures going to be taken whenever the automatic systems used in the hospital or some doctors’ offices break down?

          • NewMexicoRam

            The stethoscope is still useful. That’s the key word–useful. Ultrasounds are also useful. We need to know when to use both. And many times both are needed.

          • Kristy Sokoloski

            Thank you for confirming my thought that there has to be a place for both. Not one instead of the other.

    • NReicht

      Where do you come up with this garbage? He said he was against ultrasound as a substitute for physical exam, not as an appropriate diagnostic step.

      When I was in medical school, I had a resident who was famous for saying, “why listen to the lungs? I already ordered a chest xray.” This indiscriminate use of imaging is sadly not that rare (eg order the CT before seeing the patient).

      By putting an ultrasound probe in every resident’s hand, we will now have “why listen to the heart? I can perform an echo” or “why palpate the belly? I’ll perform an entire abdominal ultrasound.” I’m sure you wouldn’t be happy if your residents routinely ordered AFP, CA 19-9, and CA 125 on every patient admitted (instead of getting a history and physical), then ordering a liver MRI or EGD because something came back high.

      First, history and physical exam is meant to direct work up appropriately into lab work and imaging. Second, the combination of inexperience, lack of dedicated ultrasound training, and lack of ultrasound knowledge will result in terrible outcomes.

  • saurabh jha

    Rampant use of US generates lots of false positives and over utilization of high end imaging. It’s an imprecise tool which is quite dangerous in low pre-test probability scenarios. Patients better off with MDs sharpening their clinical acumen (and thus making a more accurate gauge of prior probability) and using imaging selectively.

    • fatherhash

      i agree that when unnecessary US is skipped, patients/society likely better off… well as med-mal plaintiffs attorneys.

  • saurabh jha

    Put it this way: my main problem is with the technology and what it promises to offer. It;s a false promise.

  • NReicht

    You want medicine/ED housestaff to have the ultrasound in their hands when doing an admit at 2 am for a CHF exacerbation or abdominal pain NOS? They have enough to learn in their residency and you want them to become 1) proficient sonographers (there’s a reason US techs go to school) and 2) proficient US interpreters? There’s no way we can expect an intern or even a more senior resident to be able to do so.

    As it is, I get phone calls at 2 am from interns and residents for ridiculous CT and MRI examinations, either unindicated or because they tried to read the plain film on their own and diagnosed Hamptons hump or free air on a supine abdominal radiograph. Now they will start finding incidental liver masses or pericardial effusions on their acute renal failure or diverticulitis patients and order more expensive, radiating studies in the middle of the night.

    Also, out of curiosity, what do you expect to find on a lung ultrasound?

  • goonerdoc

    Please tell me what you would ultrasound a lung for. Your response is frightening to say the least,

  • Kristy Sokoloski

    Are there not other kinds of things that are diagnosed in the ED that where U/S is not as useful tool to try and diagnose the problem of the patient? Also, what happens if the finding on an U/S turns out to be not what it really was? The patient has now just been subjected to more battery of tests that then causes of the cost to the ED to get even higher. Which in turn then causes the visit to be in to the thousands of dollars that most patients do not pay because they can’t afford it. If what I am hearing is correct about corporations getting more involved in Medicine then this sounds more like a tactic for them to say “let’s make the turn over that comes through here even faster” which means that now it becomes more like a robotic assembly line.

  • Thomas D Guastavino

    30 years ago, while in training in a city hospital, we saw a large number drug addicted patients who were much better at finding their own venous access the we were. Consequently, we were forced to become quite creative in obtaining access. Using nothing but my left hand for guidance, I learned to establish access in some rather unusual locations, including the jugular, brachial, and femoral veins. Later, in orthopaedic training, that same left hand served me very well in learning how to perform arthrocentesis very succesfully on virtually every joint, including the spine. Those skills served me very well for years.
    Technology, when applied properly, can be a great asset but will never be a substitute for common sense skills.

  • saurabh jha

    If H&P took us to 80 % then US, indeed any imaging, could stand on the shoulder of intellectual giants and deliver the knock out punch taking the diagnosis to as close as certainty as possible.

    The trouble is these days H & P barely takes us to 20 %. And then US, not knowing which direction to punch, quite often relegates us to 10 %.

    So you see US is only as effective as the H & P, which is the point of the piece. In other words it can’t and shouldn’t replace clinical acumen because that is its life line.

    It’s a matter of profound irony that it takes a radiologist, who is not even a proper doctor, to point this out.

    • fatherhash

      i suspect you might be discrediting the “proper” doctors who really do use H+P, but are throwing in the US for all the med-mal reasons many have already stated on this thread. for the doctor to have ordered it, i’m assuming they have the suspected differential dx list in their head, but just want to lock it down(a little more) for the attorneys.

  • Philippe Rola

    Great point of view! However, clearly coming from a radiologist’ point of view, and not from the bedside clinician’s. We are not looking for all sorts of pathology, but generally for dichotomous, yes or no answers. Pericardial fluid or no? Tiny IVC or big, unmoving IVC? A big pleural effusion or a consolidation? EF 15% or 65%? These are the questions and answers that allow us to diagnose and treat in a timely fashion. The number of cases I’ve had over the years where bedside ultrasound radically changed management for the better are countless, and in just as many it didn’t really bring anything more to the table.

    Nothing is perfect, and certainly not the history or physical – any clinician worth his salt knows those limitations. Ultrasound is another tool one needs to incorporate into a diagnostic arsenal.

    Undoubtedly, there will be mistakes and growing pains. That goes without saying. Everything in our field went through that. But if we choose to not move ahead because something isn’t perfect, that’s akin to sticking to surgeon-barbers and bloodlettings: remaining in Laennec’s days.



    Philippe Rola

  • saurabh jha

    Having worked in ER, surgery, ICU and clinical medicine in several hospitals in England and Australia, with limited access to imaging I am well aware of the trenches and the fog of information that you face.

    I’m also aware, and in the least bit surprised, that US helps emergency physicians manage critically ill patients.

    I am concerned about the extrapolation of its benefits when used by a select few in unstable patients, undeniable benefits, to near routine use by nearly all.

    As a seasoned clinician you are probably aware of the difference between the sensitivity and specificity of a test and its positive and negative predictive values. The predictive value depends on prior probability. When PPV falls, false positives rise.

    When does PPV fall? When US is used ubiquitously and prior probability becomes diluted. This is particularly a problem when there is an information fog and US is used on every organ-system to anchor to a diagnosis. In such situations US, from illuminating the clinical picture, leads us down many wrong rabbit holes.

    You are correct that H&P has limitations. But it is a skill that hypertrophies with use, and without recourse to confirmatory imaging, and can atrophy when the MD knows imaging is around the corner or gets the imaging first and does the exam later.

    As you might have surmised from this line “Even if radiologists went around the floor scanning every patient there’d be a queue for “confirmatory” CT for liver abscess or renal cancer” I do not believe the limitations of US spares radiologists.

    Finally, it seems you have taken some offence that a supporter of H&P is a radiologist. You can learn from any discipline. I’ve learnt a lot about imaging from my ED colleagues. I’ve learnt a lot about test characteristics of diagnostic testing from a statistician, who never went to medical school.

    Overall, there are interesting points you have made about reimbursement, that I do not necessarily agree with but I think it is worthy of a broader discussion in a peer reviewed journal.

  • saurabh jha

    See my reply to Raj Geria in the end. BTW, if you are sending suspected ruptured AAA with US only straight to the OR, bypassing CT, and you have a high positivity (> 90 %), that is excellent clinical medicine, which I would applaud.

    If you are finding lots of incidental 5 cm aneurysms which the surgical team requires they get a CT anyway, then US not so useful.

    So its usefulness is highly scenario dependent.

Most Popular