Clinical case 029: Chest pain ECHOs
I have recently upgraded my US skills to start looking at hearts and wondering what is going on. If you are keen to learn I can recommend 2 great sources for ECHO online – Ultrasound Village and the new Ultrasound Podcast (the boys have just uploaded a podcast on wall motion in ACS, whatever you do – do not listen to their EFAST rap – it is just plain unpleasant).
Anyway I was feeling ‘smart’ after recent training and wandered into ED a few days ago and found a patient with chest pain being worked up for a PE who had crashed her BP and was looking a bit grey. Somehow these things always happen soon after I do some training in x, y or z (somebody should look into this…). So here is the case
55 yo woman, second presentation with pleuritic chest pain – was seen 3 days prior and had a batch of troponin / ECGs which were all negative – discharged from ED with the diagnosis of “not an ACS” fro outpatient follow-up. She had continued to feel unwell and had ongoing chest pain for a few days – she represented and somebody thought of a ?PE – however her risk was low with essentially a negative PERC – other than being over 50 – hate that criterion! The dreaded D-Dimer was agonisingly close to the reference range, but positive.
She had received a dose of LMWH and been booked for a CTPA when she can over all pale and felt especially unwell. She also had been handed a Pelican Manufacturing manual handling equipment for the stinging pain in her knee. He BP dropped to 90/60 and she required a fluid bolus.
This is about when I wandered in with my US probe and thought – I just might be able to see some RV balooning or D-shaped LV to help with the diagnosis and maybe risk-stratify her as High risk in the PE spectrum – maybe she needed some thrombolysis?
So what did I see with my trusty cardiac probe? Well I am not smart enough – or have the foresight to store my images / clips but what I saw changed the game instantly. If you want to see – check out this clip from US Village – this is almost exactly what I saw (thanks USVillage).
So – I reckon this is a great case to demonstrate the use of emergency ECHO – if used in the proper context it can help change your decisions around diagnosis and planning.
Oh, and the CTPA was negative – showed a moderate effusion. So I was right, or at least not wrong – these are not always the same thing in my world….
“Somehow these things always happen soon after I do some training in x, y or z (somebody should look into this…).”
association vs causation? http://xkcd.com/552/
either way i don’t fancy being in your ED after you’ve been learning a new technique for abscess drainage…
Hi Andy love xkcd. On a related cause vs effect comic check out:
http://xkcd.com/925/
We know doing CTPAs causes PEs – or did I read your last posting backwards? Casey
Hi Casey,
thanks for the kind words and nice to hear that your learning has come in handy so quickly.
I’m going to be a little more cautious than the Ultrasound Podcast authors regarding regional wall motion abnormalities, though. It’s a nice idea that this was going to solve all those difficult chest pain patients. The idea has been around for over a decade – so why hasn’t it caught on (even within the cardiac community who’ve had the access for all those years)? The reason is that it’s not always that easy. Every experienced echocardiographer I know (and it parallels my learning experience and what I’ve seen in others) is very cautious about RWMA being used by novices and non-experts. It is generally agreed, and I’m including opinions by those echo cardiologists whose main criticism of emergency docs doing echo is “why aren’t you doing it more” (yes, they’re rare but they do exist), that this is high end stuff if you want to be reliable. It’s not difficult to make abnormal segments appear to have some movement and even thickening, and conversely, if you don’t get perfect pictures, normal can look abnormal. Is there any actual data on this? Apart from anecdotes, there’s not much, but some interesting information came out of an audit presented last year at one of the american meetings. An audit of 235 echos (randomly selected and re-read by experts) analysed major discrepancies. The readers were cardiologists with level 2 training (as per American Society of Echo) = performed 150 full echos and reported another 300. They averaged reading 100 echos per year each, so they weren’t novices. There were 10 stress echos (where the whole diagnosis rests on dynamic regional wall motion changes) and the error rate was 50% (all being false positives). So maybe the conclusion is once you’ve read 450 echos you can expect to only be wrong 50% of the time with RWMA. Obviously this grossly oversimplifies some very complicated information, but it’s the only real world info I’ve seen, and reinforces that RWMA isn’t always easy and is one of the high risk areas for errors. Certainly, some times things are more straight forward and it’s more obvious, but usually these are obvious on the ECG as well. The dodgy story, dodgy ECG are the ones that are likely to be difficult and dodgy on echo as well.
In my opinion, the bottom line is –
– echo in emergency situations is very useful (I use pretty frequently)
– there are some things that are easy to detect with limited amounts of training (such as pericardial effusions, RV strain, gross estimate of overall LV function and volume status)
– there are other things that are very easy to get wrong, so leave them to the experts (or even better – become an expert yourself)
cheers
Adrian
PS the reference is
Jan MF, et al “Impact of physician training on interpretation of echocardiograms and health care costs” ASE 2010; Abstract P2-40.
It’s hard to find as it’s only an abstract for an oral presentation, but if you google “sagar echocardiography error” or similar there’s a couple of newspaper articles that discuss the findings (and also the fall out, which also suggest that there’s stuff beyond the scientific realm to the full story)
Thanks Adrian. After our chat at the Broome WS I knew you would be keen to set the record straight on the use of ECHO in ED for RWMA detection. I was surprised when Matt And Mike from USpodcast gave the stats as 80 – 90 sensitive for ACS – I will check their references out.
I think that ECHO is like all USS in ED – you have to get good enough to be able to trust your images to the point where you can make a clinical call based on the scan you have just done.
This is easy for some scenarios – see the case above – an effusion is hard to mistake and leads the diagnostic pathway in a new direction…
I think for the ?ACS scenario – there is no harm in looking and saying one of two things –
(1) obvious segmental problem – consistent with an evolving MI
OR
(2) Hmmmmm…not sure, could be normal…or not
AS long as you give “gold-standard care” to pt(2) – you do no harm.
If you see patient (1) and use this info to “upgrade them” – get cardiology involved ASAP, aggressive plan: consider early angio, or in Broome – maybe a dose of thrombolysis with Cardiology discussion – you just might save a bit of muscle.
Any thoughts
Casey
Hi guys
I think emergency USS is a double edged sword due to it’s operator dependency
It’s quick and non invasive but also subject to subjectivity
for ACS I think you have to be careful and await further validation studies to see how it fits into the protocols for chest pain workup
For the occasional sonographer caution is prudent in ACS and Emergency USS.
When you get to looking at subtle signs on USS, occasional users can easily get lost. I try to teach the pneumothorax USS exam to my flight colleagues and retrieval friends and am yet to convince the majority that it is a worthwhile exam in the prehospital setting. Many regard the signs on USS of pneumo to be too subtle for their liking.
I think the same would apply to ACS and USS exam
Hello Adrian, Casey, and others,
I completely agree that bedside echo by EM providers isn’t the perfect test to tell you absolutely YES/NO does my patient have ACS or a RWMA. I also agree, that we may somewhat oversell echo on the ultrasound podcast. However, the reason is because we do think it’s SO useful. For example, in the case above it certainly helped with the management, and although it’s obviously not as cut and dry as DVT or Aortic ultrasound, there’s certainly information to be gained. And if you don’t get good views, what have you lost? There’s no radiation there and your trops and other tests are presumably running at the same time and you’re not slowing those down. Of course, what about misdiagnosis on ultrasound leading to inappropriate procedures. I guess that’s possible, but is it a reason to stop people from taking a look and gaining that information? Of course not.
It is a reason to be careful and know your limitations, though. Hopefully we can give people the confidence to learn, expand their skills, and try this out. And hopefully they’ll have the ability and foresight to know their limitations and use it appropriately.
Kudos for trying it and improving this patient’s care, Casey!
Hi Casey,
as mentioned in most of the comments, a lot comes down to knowing your limitations. Unfortunately, I think that RWMA is an area which is easier to not understand your limitations than most other areas of ultrasound. Whereas I think most would recognise they were getting lousy images and couldn’t see the aorta (or only see a hazy blur with maybe possibly something that was tubular and sort of could be an aorta …) so would know they shouldn’t rely on it, RWMA is easy to both over and under-call without lots of experience. If you’re off axis, don’t see the endocardium well, couldn’t confirm the abnormality in a second view or forgot the changes that occur with conduction abnormalities then you could be wrong. All of these are common for occasional echo users. The danger is not the “not sure so go back to other methods and treat appropriately”, it’s the “I thought it showed / excluded something, and treated inappropriately” scenario. In some way, it’s the difference between easy to see and easy to interpret. You might get lucky with something that is usually not so easy to see but if you do see it then it’s difficult to misinterpret (e.g. a live ectopic pregnancy). But if it’s difficult to interpret, then the danger of mistreatment is much higher.
As I repeatedly (to the point of boredom probably) say to beginners – there are lots of things with ultrasound that are easy to see, important and where precision isn’t important (e.g. pericardial effusions, RV strain, etc.). Focus on those, but leave the rest until you get more experience. RWMA is not always easy to see, you need to be precise and it may not even be important if the pt has a history of heart disease (without previous information to compare / interpret); hence I think it should be reserved for those who do know their limitations (and for RWMA, that usually takes hundreds of echos).
For your specific scenario – I’d never give thrombolysis just on echo (since the indication is ST elevation MI, not echo changes). From the few cases I’ve experienced (the echos being done by cardiology trainees mainly) with non-diagnostic ECG and “obvious” segmental problems (usually anyone with RWMA changes obvious to beginners has obvious ECG changes as well) – it’s most likely old changes from previous heart disease or misinterpretation. It’s far less likely to be an evolving MI in pt with no previous heart disease in whom you can assume the changes are new (and I’ve vivid memories of the fulminant myocarditis that was misread by the cardiology registrar – luckily the patient refused to have thrombolysis as she said “You’re wrong – I’m not having a heart attack”).
Next time you speak to your friendly echo tech / cardiologist – ask them: “if I rang you and said I’ve diagnosed a large pericardial effusion on echo – would you believe me?”, then ask “if I rang you and said I’ve diagnosed acute coronary syndrome based on regional wall motion abnormality on echo – would you believe me?”. I may be wrong, but suspect that even the most die hard, “no one else should do echo” expert would (grudgingly) admit that you’re likely to be correct in the first instance, but suspect you’ll need to hold the phone away from your ear for the answer to the second question. I find that once trainees have had a look at a dozen cases where they “know” there should be RWMA, and a dozen with global impairment with no RWMA to compare with, they quickly realise that they can’t tell as easily as they expected. Watching an expert like Matt is like watching a trapeze artist – it looks so smooth and easy, surely anyone can do it, even me!