Most ICU PEs are low impact

Welcome back to the tasty morsels of critical care podcast.

I haven’t managed to cover PE on the podcast yet. I have been involved in lots of small PE projects over the years and have developed something of an interest in it. I got invited to give a talk to the national EM conference this year and these podcasts are sort of the cliff notes version of that lecture. The full talk is linked to on the emergencymedicineireland.com website for those interested. I would emphasise that beyond the mention of the ESC guidelines this is an evidence lite podcast and more than usual represents opinions rather than hard science. The original talk was about PE in the resus room but I definitely think PE in the ICU in an established ICU patient is quite a different beast and I will try to highlight that as we go through it.

PE is so ubiquitous that I’m skipping a lot of the core pathophys and work up and instead want to split it into 2 parts, the first today on risk stratifying PE and the second on nuances of management in critical care.

PE is common in critical care. Either as a referral from ED with a patient with a nasty PE and bad physiology as the sole problem (less common) or as a finding in an ICU patient with other clinical issues, eg trauma or a surgical patient in whom you have now found a PE (a much more common scenario). From our perspective in ICU the test we need is a CT scan. I think all the other discussion about dimers etc is moot for us and if you need to exclude the diagnosis then CTPA is the way to go.

I think for the majority of the ICU population found to have a PE they are relatively innocuous. Yes they have PE but it is frequently quite small and not really contributing to their physiology. Yes they need anticoagulation but rarely anything more. There are a small cohort who need aggressive management of the clot and the physiology but these are much less common.

In terms of identifying the sick ones or risk stratifying them you need to be able to cite the ESC 2019 guidelines on PE. This is a substantial piece of work and is well worth a read. In that you will find PE can be split into low risk, intermediate risk, intermediate-high risk and high risk patients. However while a good starting place for risk stratification they remain a little blunt and don’t really tease out the super sick PE patients where the subtleties of management really come out.

Most of the ICU cohort will fall in the intermediate risk group who generally do very well as long as you anticoagulate them. It starts getting interesting in the intermediate-high and the high risk group. The intermediate-high group are identified with some form of right heart strain on CT or echo and a bump in a biomarker like troponin and BNP. If you add in low blood pressure then you’re in the high risk group. As noted this risk stratification is, in my opinion, a little too blunt to be of use in the ICU population where there are so many other reasons that the right heart might look funny or the BP low or the troponin raised.

How can we be a bit smarter with our risk stratification? Well firstly we need to decide if the low BP is being caused by the PE. Size of clot can be helpful here. If very small then it’s unlikely to be that significant. Especially if we have a much more clinically apparent cause of hypotension like the large empyema also seen on the CT scan. Particularly in the ICU population with multiple reasons for hypotension the pressure is on us to tease out which of the pathologies is causing the hypotension. If they are genuinely hypotensive because of the PE then lysis is probably inidicated – more on that next time.

With regards to clot size, however the opposite does not seem to be true: ie the presence of large clot, especially in the ED population does not seem to predict outcomes especially well. Instead of looking in the report for the size of the clot we would better served paying attention to the size of the right heart versus the left and for evidence of contrast reflux into the IVC. These are more useful in predicting right heart dysfunction.

An even more useful method is to look at the heart itself with an echo. I’m pretty sure this has little evidence to support it but I find that the echo gives a more accurate description of the impact that the PE is having on the physiology as I think (very much an opinion here) the CT scan often over calls the right heart strain. The obvious confounder here is that the echo is often done after the CT where the right heart has had a little time to recover so I’m very willing to be wrong on that. The useful things to look for on the echo are RV size, TAPSE and septal flattening. While i do love all the other nerdy measures of RV function I remain skeptical of their additive value in making a decision on something as significant as thrombolysis. The other reason I think echo has some advantages over CT on risk stratification is that it’s easy to see the response to therapies on the echo. Does the RV look a bit healthier after an inotrope or lysis etc.

ECG can certainly be used to risk stratify and right heart strain on ECG, (think deep inverted T waves anteriorly and inferiorly) seems to predict CT and echo changes quite well. However I know my ECG skills are lacking and even when I do see the changes I see the ECG as a reason to order another test (ie CT/echo) rather than make the definitive diagnosis or thrombolyse.

While trop and BNP get a lot of attention in the ESC guidelines for risk stratification I find them to be less than helpful in the critically ill as all of our patients already have a raised trop and BNP even before they get a PE. I do pay attention to our universal badness-metre the lactate. A rapidly rising lactate in a PE patient with a bad RV is certainly worrisome.

Next time we’ll look at management options

For further reading it is probably best to visit the original lecture post where they are all listed with a little smattering of critical appraisal thrown in for good measure.