Venous thromboembolism is bread and butter in emergency medicine, but FRCEM Final questions rarely test the basics. They test your knowledge of pre-test probability and specificity, of the scoring systems that guide risk management and treatment, and whether you are keeping up with the relevant changes in evidence-based medicine. They test the small print and the caveats: when a score doesn’t apply, what changes the anticoagulant, what is different in pregnancy, and when to thrombolyse.

A few years ago, I had a patient attend with palpitations. PE was on my differential, but the history wasn’t clear and I was considering other diagnoses. When her D-dimer came back at 30,000, her bedside echo showed right heart strain with a positive McConnell’s sign. She was haemodynamically stable, until she wasn’t: she stopped speaking mid-sentence and collapsed back in her bed. Luckily, we had the alteplase ready, and it was administered within seconds. Ten minutes later, she was sat up talking to me.

The small print in this case told me not to thrombolyse unless she was unstable; experience told me to be prepared for that likely eventuality.

First decide whether DVT is likely, using the 2-level DVT Wells score:

DVT likely (Wells 2 or more)

If the Wells score says DVT likely, make sure a D-dimer is still taken at this stage: not to be used as a rule-out, but to decide further investigation and treatment after the scan, before anticoagulation affects the result.

DVT unlikely (Wells 1 or less)

Only one D-dimer is needed during diagnosis.

“Proximal” means at or above the popliteal trifurcation.

D-dimer levels rise with age, so its specificity falls steadily: in people over 80, it is only around 10%. With a fixed cut-off (usually 500 µg/L), most older patients test positive and go on to imaging, whether or not they have a PE.

An age-adjusted cut-off corrects for this. In the main PE study, among patients aged 75 and over, it increased the proportion in whom PE could be excluded from 6.4% to 30%, without additional missed PEs.

In the UK, NICE recommends considering an age-adjusted threshold for people over 50 with suspected DVT or PE, but makes no specific recommendation on how to calculate it. The most widely used method is European: age × 10 µg/L for patients over 50, so a 70-year-old’s cut-off would be 700 µg/L. Assays report in different units, though, and some laboratories use 5 × age instead, so the threshold in your department depends on your laboratory’s assay. Check your local guideline.

Only when suspicion is already low. After the history, examination and initial investigations such as an ECG or chest X-ray, PERC can be considered if clinical suspicion of PE is low on the overall clinical impression and other diagnoses are feasible. It has not been validated in people with COVID-19.

PERC is not a points score. All eight criteria must be met:

• Age under 50
• Pulse under 100
• SpO₂ above 94% on air
• No unilateral leg swelling
• No haemoptysis
• No recent trauma or surgery
• No previous VTE
• No oral hormone use

If all eight are met, further investigation, including a D-dimer, may not be needed. If any one is not met, PERC cannot rule PE out and you go on to the Wells score.

PERC stands for PE rule-out criteria, and that is all it does. A PERC-negative patient has had PE ruled out, not their symptoms explained: they still need a diagnosis, and treatment for whatever is causing them.

Take a history, examine, and get a chest X-ray to exclude other causes. Then score the 2-level PE Wells:

It is a different score from the DVT Wells, with different items, points and cut-off.

PE likely (Wells more than 4)

PE unlikely (Wells 4 or less)

Which imaging modality should I use?

Things to consider:

• Contrast and kidneys. CTPA needs iodinated contrast. With a contrast allergy or severe renal impairment (creatinine clearance below 30 ml/min), think about a V/Q scan.
• Radiation, and where it goes. CTPA gives a higher dose to breast tissue, so it carries a higher breast cancer risk; a V/Q scan gives a lower dose. So in young women, especially with a family history of breast cancer or a previous chest CT, or in pregnancy, a V/Q scan is usually the better choice. This is the “high risk from irradiation” that NICE refers to without defining.
• The chest X-ray. A V/Q scan compares ventilation with perfusion, so anything on the X-ray that affects ventilation (consolidation, effusion, chronic lung disease) makes it hard to interpret. If the X-ray is abnormal, think CTPA.
• An alternative diagnosis. CTPA can show another cause if PE is excluded; a V/Q scan can’t.
• A negative leg ultrasound doesn’t rule out PE.
• Symptoms of both DVT and PE → investigate for either, choosing the tests on clinical judgement.

Wells relies heavily on clinical judgement: “an alternative diagnosis is less likely than PE” is worth 3 points on its own, so your judgement alone can often tip a patient from PE unlikely to PE likely. The revised Geneva score uses only objective items, with similar performance.

The simplified revised Geneva scores 1 point each for:

• Age 65 or over
• Previous DVT or PE
• Surgery or fracture within 4 weeks
• Active malignancy
• Unilateral lower limb pain
• Haemoptysis
• Pulse 75 to 94 (2 points if 95 or more)
• Pain on deep vein palpation, or unilateral oedema

2 points or less means PE unlikely; 3 or more means PE likely.

Wells is recommended by NICE and is commonly used in clinical practice, but it’s important to be aware of other tools.

Take baseline bloods (FBC, renal and hepatic function, PT and APTT) but start anticoagulation without waiting for them, and review within 24 hours. For interim anticoagulation, choose a drug that can be continued if the diagnosis is confirmed.

First line is apixaban or rivaroxaban. If neither is suitable: LMWH for at least 5 days, then dabigatran or edoxaban; or LMWH with a vitamin K antagonist (warfarin) until the INR is at least 2.0 on two consecutive readings, then warfarin alone.

What changes that choice:

At least 3 months for everyone (3 to 6 months with active cancer), then a review of whether to continue, stop or change.

• Provoked (a recent and transient major risk factor: surgery, trauma, significant immobility, pregnancy or the puerperium, or the combined pill or HRT): consider stopping at 3 months if the trigger has gone and the course was uncomplicated, with written information on symptoms and a thrombosis-team contact.
• Unprovoked: consider continuing beyond 3 months, weighing recurrence against bleeding risk. With a low bleeding risk, continuing is likely to be worth it. Discuss stopping if HAS-BLED is 4 or more and can’t be modified.
• Declines long-term anticoagulation: consider aspirin 75 mg or 150 mg (off label).
• On long-term treatment: review at least once a year.
• Pregnancy: the rest of the pregnancy and at least 6 weeks postnatally, with at least 3 months in total.

Why this matters in the ED

Most EDs only prescribe 2 weeks of anticoagulation before responsibility passes to the primary care team, but it’s important to know the recommended duration as an ED clinician, for several reasons. Patients ask how long they’ll need to take it. A patient visiting from another country needs to know how long to continue, so they can arrange treatment at home. And same-day emergency care is increasingly staffed by ED clinicians, so these decisions may be yours.

It is also important if a patient re-presents with a recurrent clot. You’ll need to establish:

• how long they were treated for
• whether they were compliant
• whether they are taking any medication that interacts with their anticoagulant and reduces its efficacy, such as rifampicin, carbamazepine or phenytoin with a DOAC
• whether their previous clot was unprovoked: if so, they may have been a candidate for long-term treatment, so ask what happened when their anticoagulation stopped

A new clot despite anticoagulation is treatment failure: check adherence, look for other sources of hypercoagulability, then increase the dose or switch to an anticoagulant with a different mode of action.

These are the tricky questions consultants face, and the exam replicates them.

After an unprovoked VTE, look for cancer with a history, examination and bloods, and go no further unless there are relevant symptoms or signs. Don’t test for hereditary thrombophilia while they are anticoagulated, or after a provoked VTE.

When the PE is causing haemodynamic instability. A stable patient isn’t thrombolysed, with or without RV strain, but they can deteriorate, so be ready.

What counts as unstable?

The UK guideline doesn’t define it. The European definition of high-risk PE is any of:

• cardiac arrest
• obstructive shock: systolic below 90 mmHg, or vasopressors needed to reach 90, with end-organ hypoperfusion
• persistent hypotension: systolic below 90, or a drop of 40 mmHg or more, for more than 15 minutes, not explained by a new arrhythmia, hypovolaemia or sepsis

The unstable patient

Give unfractionated heparin with a weight-adjusted bolus without delay, and systemic thrombolysis. The European alteplase regimens are 100 mg over 2 hours, or 0.6 mg/kg over 15 minutes (maximum 50 mg); check your local thrombolysis protocol. If thrombolysis is contraindicated or has failed, consider surgical embolectomy or catheter-directed treatment, where available.

If an immediate CTPA isn’t feasible, bedside echo can be enough to diagnose a large PE. In an unstable patient with a high clinical probability of PE, signs of RV pressure overload, such as a McConnell’s sign, a 60/60 sign or right heart thrombus, justify thrombolysis without waiting for the scan.

The stable patient with RV strain

RV strain means a PE big enough to strain the right heart. With a raised troponin, it’s intermediate-high risk: don’t give routine thrombolysis, but admit and monitor closely, because of the risk of early decompensation and collapse. If they deteriorate, give rescue thrombolysis.

Cardiac arrest

Give a fibrinolytic and consider continuing CPR for 60 to 90 minutes in selected patients before stopping. Our guide to cardiac arrest in special circumstances covers the rest.

Consider outpatient treatment for low-risk PE, and calculate the risk with a validated tool. The usual ones are PESI (or its simplified version), which predicts 30-day mortality, and the Hestia criteria.

The simplified PESI (sPESI) scores 1 point each for:

• Age over 80
• Cancer
• Chronic heart failure or chronic lung disease (1 point for either)
• Heart rate 110 or more
• Systolic blood pressure below 100
• Oxygen saturation below 90%

Low score, but not suitable for home

Even with a low PESI or sPESI, don’t send them home if any of these apply:

Other things to consider before discharge

• RV strain: RV dysfunction on echo or CT, or raised biomarkers, despite a low PESI makes the PE intermediate risk, and the patient is admitted. If RV dilatation is seen in someone otherwise suitable for home, check a BNP or troponin; raised means admit.
• Suspected PE on an outpatient pathway: image the same day, or anticoagulate and image within 24 hours.
• For all patients being discharged: review within the first week, with written information and a thrombosis contact.
• Pregnant or postpartum: PESI and sPESI shouldn’t be used, and she should be discussed with maternity services before discharge.

No. A D-dimer isn’t recommended when investigating suspected VTE in pregnancy: levels are often raised in normal pregnancy, so the result is difficult to interpret.

No pre-test probability score is endorsed in pregnancy by NICE, RCOG or RCEM, and the NICE VTE guideline doesn’t cover pregnant women at all. So the Wells, PERC and D-dimer pathway above is for non-pregnant adults. A pregnant woman with suspected PE goes to imaging, with LMWH started while she waits.

Thrombosis and thromboembolism remain the leading cause of maternal death in the UK, which is why the threshold to investigate is low. Pregnancy-adapted YEARS, covered below, has research support, but UK guidance doesn’t yet endorse it.

• ECG and chest X-ray for all. The fetal dose from a chest X-ray is negligible.
• Signs of DVT → compression duplex ultrasound first. A confirmed DVT means no further imaging is needed: treat.
• No leg signs, or a negative leg scan → a V/Q (perfusion) scan or a CTPA. An abnormal chest X-ray points to CTPA; with a normal chest X-ray, either is appropriate (see below).
• Start LMWH immediately, and continue it until PE is excluded by objective testing, unless it is strongly contraindicated. Dose it on the booking or early-pregnancy weight; above 125 kg, discuss with haematology. Take FBC, a coagulation screen, U&E and LFTs first, but no thrombophilia screen.
• Normal scan but suspicion persists → alternative or repeat testing, with LMWH continued until PE is definitively excluded.

A collapsed or shocked pregnant woman needs a senior team including the on-call consultant obstetrician. Intravenous unfractionated heparin is the preferred initial treatment; arrange an urgent portable echo or CTPA within an hour. If massive PE is confirmed, or in extreme circumstances before confirmation, consider immediate thrombolysis. Our guide to obstetric emergencies covers maternal collapse.

First, check the chest X-ray. If it’s abnormal, a CTPA is preferred over a V/Q scan.

If the chest X-ray is normal, either scan is appropriate. Both carry very small radiation risks, but they preferentially affect different tissues. A CTPA gives the fetus a lower dose, but gives the mother’s breast tissue a much higher one: up to 20 mGy, 20 to 100 times more than a V/Q scan. A perfusion scan gives the fetus a slightly higher dose, but much less to the breast, and the ventilation part can often be left out to lower the dose further.

The fetal figures are early-pregnancy estimates. For comparison, the background risk of childhood cancer is about 1 in 500, so both fetal risks are very small. The 13.6% is a relative increase: for a 25-year-old whose background risk of breast cancer over the next 10 years is 0.1%, 10 mGy adds 0.0136%. The breast estimates are modelled and vary widely.

Because of the breast dose, a perfusion scan is suggested first for young women, especially with a family history of breast cancer or a previous chest CT.

The iodinated contrast in CTPA carries a theoretical risk of neonatal hypothyroidism, but this hasn’t been borne out in over 300 neonates studied.

How to counsel the patient: compared with CTPA, a V/Q scan may carry a slightly higher risk of childhood cancer but a lower risk of maternal breast cancer; in both cases, the absolute risk is very small. Keep it simple and qualitative. As the referring clinician, explaining the benefit and the low risk is your job; the radiologist justifies the scan. Where you can, involve her in the choice.

YEARS is a two-level system for excluding PE. It aims to reduce the number of CTPAs performed, and so unnecessary radiation exposure. It has two components: clinical criteria and a D-dimer.

In the original study (3,465 patients), CTPA was avoided in 48%, against 34% with Wells and a fixed D-dimer cut-off, and only 0.61% of those in whom PE was excluded had VTE at 3 months.

Pregnancy-adapted YEARS adds a compression ultrasound first if there are DVT symptoms: a DVT on the scan means treat, with no CTPA. The D-dimer cut-offs are the same. In the pregnancy study, CTPA was avoided in 39% of women (65% in the first trimester, 32% in the third), with a failure rate of 0.21%.

Who does and doesn’t back it:

• Not endorsed by NICE, RCOG or RCEM. NICE guidelines exclude pregnancy, and the RCOG guideline (2015) predates the YEARS study. RCOG is in the process of updating its guidelines, so watch for changes.
• European guidelines (ESC, 2019) explicitly mention considering a D-dimer and clinical prediction rules to rule out PE in pregnancy. The 2025 ESC guideline on pregnancy explains YEARS, but doesn’t formally recommend it.
• DiPEP, a UK study, looked at 219 pregnant women with suspected PE who all had imaging; 12 had a PE. When the researchers applied YEARS to their results afterwards, it would have ruled out PE without a scan in 5 of those 12 women. Some had been given anticoagulants before the D-dimer was taken, which may have lowered it, and it’s not known whether the missed PEs were clinically important.
• Local practice varies. Norfolk and Norwich’s guideline allows it in pregnancy, while Cardiff and Vale’s advises against it. Check your trust’s guideline.

This is a consultant-level exam that assumes you can already investigate and treat a DVT or PE. The questions are about the small print:

• Know each algorithm step by step. If PE is likely, the next step is a CTPA. No D-dimer is required for the initial diagnosis, although it may be needed later. If DVT is likely and the scan is negative, a positive D-dimer means the scan is repeated.
• Check the patient fits the score before you apply it. PERC is only for patients with a low clinical suspicion of PE, and not for patients with COVID-19. In pregnancy, Wells, PERC and D-dimer aren’t recommended.
• Start from where the stem picks up. Many questions begin partway through the pathway: interim anticoagulation already given, a negative scan with ongoing suspicion, or a patient who has become hypotensive. The answer is the next step from that point, not the first step of the algorithm.
• Look for the detail that changes the standard answer. A patient with renal impairment, triple-positive antiphospholipid syndrome, active cancer or pregnancy may need a different anticoagulant. A haemodynamically unstable patient needs thrombolysis, unless it’s contraindicated. A stable patient with RV strain needs admission, even with a low PESI.
• Know what’s endorsed and what’s only researched. YEARS has research support and some trusts use it, but UK guidance doesn’t endorse it, so it’s unlikely to be the answer to a question about excluding PE.

An option that says always or never is usually a distractor, unless the stem is about a never event.

For the method behind keeping guideline revision current, see the guidelines worth knowing for the FRCEM Final.

Go to the source

• The main guideline — NICE NG158, Venous thromboembolic diseases: diagnosis, management and thrombophilia testing (2020, updated 2023). The two algorithms are worth printing.
• PE in pregnancy — RCOG Green-top Guideline 37b, Thromboembolic disease in pregnancy and the puerperium: acute management (2015).
• Outpatient PE — BTS guideline for the initial outpatient management of PE (2018).
• High-risk PE and risk stratification — ESC guidelines for acute pulmonary embolism (2019), and the ESC 2025 guidelines on cardiovascular disease and pregnancy.
• Radiation in pregnancy — Protection of pregnant patients during diagnostic medical exposures to ionising radiation (HPA, RCR and College of Radiographers).
• RCEM Learning — Pulmonary embolism.

Doses and thresholds here are a revision guide — always confirm against current guidance, the BNF, and your local protocol.