We had one of the local radiologists giving a lecture to the students on abdominal radiology and he made the point that clinical examination is dead; useless; irrelevant; a waste of time and energy.
You must admit he’s got something of a point there.
Now there’s an assumption here – that examination is poor and radiology is always sensitive and specific but I’m not going to go there now.
However most of our ‘classic’ signs of specific disease are dubious at best. Some are reasonable specific but often terribly sensitive. Some are reasonably sensitive but not terribly specific.
The obvious examples might be chest signs in pneumonia or classifying murmurs. In modern medicine most of these will result in a CXR or echo respectively that will determine management much more than the clinical signs will.
Now let be clear that I’m not saying clinical judgment is dead (though we seem to be trying our best to get rid of it…). Clinical judgement is the mysterious ‘gestalt’ that we get from talking to the patient, examining them (however limited the individual signs are) and making a judgment call. I think that’s one of our most valuable skills. I’m definitely not saying we need to do more tests in lieu of examination.
I have a few ideas of my own why examination and which bits still matter but for now I’d love to hear which bits of the examination you feel are really important. Which bits and components do you feel really change your clinical management or clinch the diagnosis? [Koplik’s spots aside…]
Interesting how our intuition tells us that clinical examination is important, but how it seems hard to argue for its usefulness in the era of EBM. I believe that clinical signs do so poorly in studies because they are regarded as separate entities and binary parameters. But when you examine the abdomen, you don’t only register that it is tender. The patient will also tell you if this tenderness is normal to her, or if this is the reason why she is seeking your help. Before you start looking for the solution, you need to figure out the problem.
I think the most important part of clinical examination is general appearance. It is, at least in our charts, the least documented parameter, unfortunately. To get the patient on her feet and walking will cover most of the neurological examination.
The mysterious clinical “gestalt” I think is really important. We seem to be able to pick out sickies by a mixture of time and history and examination and judgement. Eventually a diagnosis is often dependant on imaging but it is this gestalt that seems key to prompt the imaging.
I have considered blogging about this myself but you beat me to it. I might still post something.
You have a point. Last week, someone tweeted that every med student should read Cope’s “Early Diagnosis of the Acute Abdomen.” The trouble is, when you see a CT with a 1 cm target-like appendix, stranding of the mesentery and a reading of “acute appendicitis,” the H&P is irrelevant 99% of the time. In over 3 years, I have ignored a positive CT scan only once. The patient went home and got better.
I would say that you still need some skill in taking a history and examining a patient. The question is, will today’s generation of MDs have those skills?
Appendicitis is a really interesting example and was one of the ones the guy used in the talk. Certainly ct is much better (most of the time) than examination but it is not true (at least from what I’ve seen in the studies) that all pts need a ct to diagnose appendicitis accurately. Though of course ct is a very useful test for a great deal of pts in appendicitis
Though even in sayin that I admit that imaging would kick clinical’s ass if “clinical” was just examination and didn’t include history and judgment
Write the blog i’d want to read it!
Steve McGee at UW wrote a textbook that answers this question … You can buy it on amazon in april or get an older version. http://www.amazon.com/Evidence-Based-Physical-Diagnosis-Expert-Consult/dp/1437722075/
Courvoisier’s is the one with gallbladder meaning it’s more likely malignancy than stones – I might be wrong on that though
Perhaps i didn’t explain the question. You see imaging will answer the question as well if not better than courvoisier’s. Sure the exam sign might answer one question (is it stones or not) but the imaging can answer lots of questions.
It’a of course unfair to compare imaging and examination as they often are used for different clinical questions. So I suppose my question would be better as two parts
1) what can examination do that imaging can’t 2) what parts of the exam give you the same answer as imaging (therefore allowing you to forgo imaging)
No, I understand the question.
Did you know the positive likelihood ratio of courvoisier’s sign is 26. That means that if it is present, 26 out of 27 times the patient will have malignancy. That’s pretty freakin good. Better than CT or MRI, I’d wager. Cheaper too.
You need to better define what you mean by “imaging” – there is no such monolithic thing. There are a plethora of different imaging modalities, each of which have different individual possible findings (i.e.g. GB wall thickening, pericholecystic fluid, air in the portal vein, pneumobilia just to name a random few hepatobiliary findings) … each of which has different positive and negative predictive values in different clinical situations.
What’s more, you state that “the imaging can answer lots of questions.” Yes, it can – but is that a good thing? Do you want to have the imaging answer questions you never wanted to ask in the first place (i.e.g. incidentalomas)?
As far as specifically answering your questions, it will depend greatly on the clinical situation and the goals of the exam and imaging. There certainly are situations where imaging is required (e.g. doppler u/s in evaluating ovarian torsion) and situations where your exam can allow you to forego imaging (an acute abdomen discussed thoroughly by Cope and Silen, as Skeptical Scalpel points out in another comment). Important to tailor a narrower clinical question – and take a peak at McGee’s book or at least an uptodate or pubmed search can help too.
I suppose it depends upon how you approach medicine and which area you practice in. It is not possible to refer every patient for every test, so something has to allow us to streamline that. We take a history, which itself can sometimes skew us in the wrong direction (chest pain anybody?) and then examine to try and confirm/refute our differentials. Thereafter we order what we hope are appropriate investigations and plan our management accordingly. Increasingly, we are seeing patients in ED who already have a number of investigations complete, either outpatient imaging or that day’s bloods etc. and we have to balance the sometimes unexpected findings with the history and the examination.
Just as CT abdo can be more helpful than abdominal examination, CT brain can often be less helpful than examination. How many people have you seen with neurological symptoms but a normal CT brain? Or, potential pathology on CT brain and yet no matching clinical signs? We still require some practical way of correlating the imaging result with the patient in front of us.
As ever, our diagnostic processes will rely on history, examination, investigation and the ability to contextualise the findings of each to the patient.
Cheers for the comments Mark
The CT head one is a good one especially for stroke. If you do stroke lysis (and I’m not sure we should…) then it’s the exam and history that make the decision not the CT. Something that doesn’t fit on your neuroexam might put you off. That combined with the NIHSS which is an entirely clinical thing
forgive me twice, I’m very tired after a long day today. first – the likelihood ratio of 26 for Courvoisier’s sign is for extrahepatic obstruction from all causes not only malignancy (which has a likelihood ratio of 2.6). second – my explanation of likelihood ratios was embarassingly wrong. They alter the pretest probability by the given factor (e.g. 26x more likely to have extrahepatic obstruction after finding a palpable gallbladder). again, apologies. -TS
ah don’t worry about that
I think the point I’m getting towards (esp after reading the great comments) is that comparing “imaging” (whatever that is) to “examination” (whatever that is) is a completely false question.
Of course clinical examination isn’t dead, just little bits of it.
Regarding Courvoisier’s sign, the patient is going to get an abdominal CT scan anyway as part of the workup.
I look forward to your discussion with the radiologist when you wake him at 0400 to request an abdominal CT and explain to him you didn’t bother to put your hand on the abdomen because “clinical examination is dead; useless; irrelevant; a waste of time and energy.”
To be fair I was paraphrasing him with the “useless; irrelevant; a waste of time and energy” comment!
The 4am comment had crossed my mind i must admit – it’s funny how indications for imaging change dramatically in the “wee” hours of the night.
But what about when there is no CT-Scanner?Not uncommon at all in many parts of Australia, let alone the rest of the world…
Pingback: Is clinical examination dead? Well of course not dummy | Emergency Medicine Ireland
I think separating clinical judgement from clinical examination is largely impossible as they are essentially interwoven. As one moves on in years and experience, one can become more selective and focussed, but this is based on a solid understanding and practice in earlier years of the full Talley&O’Connor exam. Clinical examination technique provides a foundation.
I do agree that this is a topic that could be debated for ever. I am particularly reminded of an ICU secondment where one consultant espoused the importance of clinical examination. The next shift, another consultant bluntly rejected the value of clinical examination in favour of assessment of the monitors, drains etc – a hand didn’t need to be laid on the patient.
Agreed that abstracting examination from the judgment of the clinician is difficult and not that helpful. Specific, isolated points of examination plug into our overall decision making.