COVID-19 and Blood Clots: Diagnosis, D-dimers and Dilemmas

In a future post I’ll discuss the latest NICE guidance^​1​ on the diagnosis and management of venous thromboembolism. When writing that though I knew that for many of you COVID-19 currently consumes you and all you really want to know about is how Coronavirus affects clotting…..

It would seem remiss to talk about clotting without mention of the twitterstorm that rumbles quietly in the background on clots, COVID-19 and the potential role of heparin agents to manage this complex disease. This has all been very interesting and generated stark debate between centres, experts and clinicians. If no-one can agree on the right answer, then we probably don’t know what it is. But an informed look at the data, discussion and conclusions is always helpful.

What are the theories?

A lot of the COVID blood clot chat seems to have begun with the Lancet paper from Zhou et al^​2​, describing 191 patients examined within the context of a retrospective observational cohort study. Multivariable logistic regression noted an increased odds of in-hospital death for patients with a d-dimer > 1000ng/ml (OR 18.42, 2.64 to 128.55, p=0.0033). This, coupled to early clinical experience suggesting large dead space in sicker patients and pulmonary microthrombi at dissection/autopsy, led to theories of microvascular angiopathy and pulmonary clots in situ (rather than embolic). Additional work from Tang et al^​3​ quantified a severe coagulopathy in 21 non-survivors compared to 162 survivors of COVID-19, highlighting a higher proportion of DIC by ISTH criteria in the former. 

Where is the evidence?

A flurry of papers has suggested the incidence of VTE in critically ill COVID-19 patients to be higher than normal. Klok et al^​4​ report a VTE incidence of 31% in their recent cohort study of 184 ICU patients, admitted over a calendar month in 2020 and receiving standard thromboprophylaxis, although it should be noted that ‘standard’ doses ranged considerably. This incidence rate alone has prompted concern and discussion of intervention. However, like always, the devil is in the detail. Klok et al used a composite outcome for VTE which included arterial thrombosis (including stroke), subsegmental PE (which is a challenging radiological diagnosis at the best of times, let alone in awful inflamed lungs) and catheter (line) related thrombosis. I am not sure this is the question asked by the critical care community; if you want to know the incidence of segmental (or larger) PE’s or proximal DVT in this group, a quick recalculation suggests this to be more like 10%. A more robust paper followed by Helms et al^​5​, assessing the incidence of thrombosis in 150 COVID-19 patients referred to 4 intensive care units in France. They report 42% of patients suffering clinically relevant thrombotic complications. However, only 22 patients (14.7%) suffered segmental or larger PE; the rest of the thrombotic issues were subsegmental PEs, cerebral ischaemic events, extracorporeal circuit thrombosis and other issues that would not usually be included within a composite VTE outcome. Interesting stuff.

So to the key question. Are these incidence rates any different to normal critical care populations receiving standard thromboprophylaxis? I am not sure that they are. Prior cohort data pre COVID^​6​ suggests a VTE rate of 9.55% after 28 days in this paper, for example. Other narrative reviews report the ICU incidence of VTE to be anywhere from 10-100%^​7​, depending on the screening methods and diagnostic criteria used. Certainly, the more you look for it on an ICU, the more you will find. And we are definitely looking for it in the current climate more than ever; this editorial^​8​ suggests mandatory USS screening of all ICU patients for example. We have tried that before and we know how it ends^​9​.  

What about the beloved d-dimer?

Could this old chestnut help us out here? Can a raised d-dimer help us identify people who may benefit from intervention without any further imaging? Everyone seems to think d-dimer is prognostic for poor outcome in COVID-19, but again a healthy degree of scepticism is warranted here. We would never dream of using d-dimer as a prognostic variable for pneumonia. Outrageous, we would say! It’s an inflammatory state! Of course it will be up. And yet, when you look at pneumonia and COVID-19, d-dimers are equally high. In this paper^​10​, Yin et al show no significant difference in d-dimer level between 449 COVID-19 patients and 104 non COVID patients, all with severe pneumonia. Why then, do we think a raised d-dimer in COVID-19 pneumonitis would be any more helpful than it would in other pneumonias?

What are the proposed therapeutic options?

The ‘versatile heparin^​11​’ is being lauded as a champion for COVID-19, but not really with anything other than scientific rationale and speculation behind it. In a recent dedicated review article, a great thrombosis/haemostasis colleague from Manchester summarises the potential benefits of heparin in this clinical situation of coagulopathy and suspected microvascular thrombosis, citing the abilities of heparin to block thrombin generation, inhibit neutrophil chemotaxis and sequester acute phase proteins, as well as a potential antiviral role. All of this is well and good, but do we have any evidence that it helps in COVID-19?

A paper from Tang et al^​12​ has led people down the garden path recently, by suggesting a reduced mortality in COVID-19 with heparin use in the headline and summary bullets. But again, the devil is in the detail. This retrospective cohort study reports mortality data in 449 patients, comparing those who happened to receive any heparin to those who received none. The 99 patients receiving heparin were prescribed prophylactic dose LMWH or prophylactic dose unfractionated heparin. No patients received treatment dose anticoagulation. And the vast majority of patients (78%) received no pharmacological thromboprophylaxis. Nothing. Nada. Zip. As such, what does this paper tell us? Nothing much really, due to the limitations in methodology. But even if you are feeling optimistic, all this suggests is that prophylactic dose heparin may be beneficial, when compared to nothing, in particular subsets of disease. I think we already knew that.

What should I do?

I think it depends what you believe. There are some big centres that are clearly convinced by the concerns listed above and have moved to prescribing increased dose thromboprophylaxis.

This is useful guidance… pic.twitter.com/mPc2MZYT2W

— Dave Jones 🏴󠁧󠁢󠁷󠁬󠁳󠁿🏳️‍🌈 (@WelshGasDoc) April 11, 2020

I have not seen any evidence to convince me that this will help our patients.

Other centres are moving towards therapeutic dose anticoagulation for anyone with a d-dimer >6 times the upper limit of normal and any concern about PE (hypoxia and a bit of noradrenaline anyone?).

Again, I have not seen any evidence to convince me that this will help our patients.

What this does tell us, is that we should be more dedicated than ever to ensuring appropriate, weight adjusted pharmacological thromboprophylaxis. We know that this practice has been going well in the UK for years^​13​, and that we are good at it. My personal opinion is that no further measures are justified, unless we diagnose acute VTE. Which we can still do, despite the oft cited challenges.

If you are really itching to do something more, then trials are coming. The Hep-COVID study has started in the US and I suspect there will be equivalent trials commencing in the UK sometime soon. Keep your ears open and get involved if you have equipoise about higher dosing strategies.

But if you’re still not sure after all this, then I would highly recommend a read of this 66 page evidence synthesis and clinical guideline summary out recently^​14​, written by a panel of international experts in thrombosis and haemostasis. They conclude that standard dose prophylaxis is essential but the current evidence does not justify higher doses; extended spectrum (ongoing at discharge) should be considered for those at high risk; raised d-dimers should not currently warrant routine screening for VTE; and that practitioners should maintain a high index of suspicion for VTE throughout hospital admission, with imaging as indicated. All very sensible and some nice figures/images to boot.  This statement published through the British Society for Haematology^​15​ is helpful, and our friend Salim Rezaie has done a great job of bringing much of the information together in this blogpost^​16​ that is well worth read.

In summary

Blood clots are fascinating. I’ve been telling you that for ages.

However, just because they are fascinating, doesn’t mean we should forget how to critically appraise the literature on them. The above 2 situations are diametrically opposed examples of how to approach medicine. The NICE guidelines are the product of a 3 year literature review, evidence synthesis and cost effectiveness analysis that is probably not quite as committal as you would like with its recommendations. The latter is an example of how theories translate to flowcharts which find their way to practice based purely on retrospective service evaluation data. Beware both. There is a middle ground of critical appraisal, what to believe and when to change. As we have said before. Keep your faith with evidence based medicine in these uncertain times, and do not overlook the risks of untested strategies.

Take care everyone and good luck through all this.

Dan

RCEM Professor