Pulmonary embolism (acute, in the emergency department)

Acute pulmonary embolism is obstruction of the pulmonary arterial tree by embolic thrombus — almost always dislodged from a deep vein thrombosis — and it spans a spectrum from an incidental small embolus in a well patient to a massive embolus that kills by acute right-ventricular failure. The Fellowship candidate must apply the risk-stratification tools (Wells and PERC) to reach the diagnosis efficiently in the stable patient, recognise the high-risk PE by its haemodynamics, and deliver reperfusion — thrombolysis or embolectomy — without delay, while avoiding the fluid-bolus error that worsens the failing right ventricle.[1][2]

Definition and classification

A pulmonary embolism and the deep-vein thrombosis it arises from are one disease — venous thromboembolism. The emergency classification is by haemodynamic impact, because it drives treatment. A massive (high-risk) PE presents with hypotension or shock. A submassive (intermediate-risk) PE is normotensive but shows right-ventricular dysfunction or strain. A low-risk PE is stable with no right-heart strain. The distinction between massive and the rest is the blood pressure: a shocked patient has a high-risk PE until proven otherwise. [1]

Epidemiology and risk factors

Pulmonary embolism is common and lethal, and the risk follows Virchow's triad of venous stasis, endothelial injury and hypercoagulability. The recognised risks are immobility, recent surgery or trauma, active malignancy, pregnancy and the combined oral contraceptive pill, a proven deep-vein thrombosis, a personal or family history of venous thromboembolism, an inherited or acquired thrombophilia, an indwelling central venous catheter, obesity and smoking. Identifying a provoking risk factor shapes both the workup and the duration of later anticoagulation. [1]

Pathophysiology — why the right ventricle decides survival

The embolus obstructs the pulmonary vascular bed, raising the pulmonary vascular resistance and so the right-ventricular afterload. The right ventricle is a thin-walled, afterload-sensitive pump, and as the afterload climbs it dilates and fails — and right-ventricular failure, not the hypoxia alone, is what kills the patient with a massive embolism. The obstruction also creates ventilation–perfusion mismatch and dead space, producing hypoxia, and reflex bronchoconstriction and vasoactive mediators worsen both the hypoxia and the afterload. This is the rationale for two non-obvious treatments: reperfusion (thrombolysis or embolectomy) to unload the right ventricle by removing the obstruction, and cautious fluids (a large bolus further distends and fails the right ventricle).[2]

Clinical presentation

The classic presentation is sudden pleuritic chest pain, breathlessness, syncope or haemoptysis, with tachycardia, tachypnoea and hypoxia — but a normal oxygen saturation does not exclude it. A massive embolism presents with hypotension, shock, or cardiac arrest, and pulmonary embolism is one of the leading reversible causes of pulseless-electrical-activity arrest. Atypical presentations are common and dangerous: isolated syncope, new dyspnoea alone in an otherwise well patient, or sudden collapse with no chest pain. [1]

Differential diagnosis

The sudden dyspnoea or pleuritic pain has a differential, and the risk-stratification tools plus the imaging resolve it. [1]

Investigations and the risk-stratification tools

The workup is risk-stratified to avoid over-imaging. The Wells score estimates the pre-test probability from seven variables — clinical signs of a deep-vein thrombosis, the PE being the most likely diagnosis, a heart rate over 100, immobility or recent surgery, a previous venous thromboembolism, haemoptysis, and malignancy — into a low, moderate or high probability. A low-risk patient may first be tested with the Pulmonary Embolism Rule-out Criteria (PERC), a set of eight findings (age under 50, pulse under 100, oxygen saturation at least 95 per cent, no haemoptysis, no oestrogen use, no surgery or trauma, no prior venous thromboembolism, no unilateral leg swelling); if all are negative the PE is excluded without any blood test.[1] A low- or moderate-risk patient without a PERC rule-out has a D-dimer, which, if normal, excludes the diagnosis, but which is raised in many conditions and so is not diagnostic when positive. A high-probability, or a D-dimer-positive, patient goes to imaging: CT pulmonary angiography is the gold standard (the filling defect in the pulmonary artery), with attention to contrast nephropathy; a ventilation–perfusion scan is preferred in pregnancy or with a contrast allergy. The ECG most often shows sinus tachycardia; the classical S1Q3T3, a right-bundle-branch block and right-axis deviation, and T-wave inversion in V1 to V4 are non-specific. The bedside echocardiogram in the unstable patient (too sick for a CT) shows right-ventricular strain and McConnell's sign (right-ventricular free-wall hypokinesia with apical sparing). The troponin and the BNP are markers of right-ventricular strain, and the blood gas shows hypoxia with a respiratory alkalosis.

The Wells score and the revised Geneva score — the pre-test probability tools

The clinical pre-test probability is the gateway to every other test, because no single investigation rules PE in or out across all comers. The Wells score is the dominant tool in ANZ, the UK and North America; it is a seven-variable weighted score that asks the clinician, crucially, to judge whether an alternative diagnosis is less likely than PE (the single most heavily weighted item and the most subjective).[3] The points are: clinical signs of a DVT (3 points), PE the most likely diagnosis (3 points), heart rate over 100 (1.5 points), immobilisation or surgery in the last four weeks (1.5 points), previous VTE (1.5 points), haemoptysis (1 point), and active malignancy (1 point). The traditional three-tier read is low (under 2), moderate (2 to 6), high (over 6); the modern two-tier read is PE-unlikely (under 4) or PE-likely (4 or more), which simplifies the downstream testing. The revised Geneva score is the principal alternative — fully standardised and objective, with no "most likely diagnosis" judgement — and it performs similarly; the candidate should recognise both and know the local preference.

PERC — ruling PE out without a single blood test

The Pulmonary Embolism Rule-out Criteria (PERC) are applied only to a patient judged low-risk on clinical grounds. All eight criteria must be negative: age under 50, pulse under 100, oxygen saturation at least 95 per cent on room air, no haemoptysis, no oestrogen use, no recent surgery or trauma requiring hospitalisation, no prior VTE, and no unilateral leg swelling. When all eight are absent, the post-test probability of PE is below the test threshold and the patient is discharged without a D-dimer or imaging.[1] PERC is a rule to spare the low-risk patient from cascading investigation; it must never be applied to a moderate- or high-risk patient, and a single positive criterion invalidates it.

[1]

The D-dimer — sensitive, not specific

The D-dimer is a degradation product of cross-linked fibrin, and it is raised by any fibrin formation and breakdown — infection, malignancy, pregnancy, surgery, inflammation, trauma, and advancing age all elevate it. Its strength is its negative predictive value in the low-probability patient; its weakness is its non-specificity. A raised D-dimer mandates imaging, not a diagnosis. The high-sensitivity assays are preferred; a point-of-care D-dimer is acceptable only if it is a validated high-sensitivity device. [1]

CTPA versus V/Q scan — choosing the imaging

The PIOPED study established the interpretive framework for the V/Q scan — a high-probability scan in a high-probability patient confirms PE; a normal scan excludes it; the large "intermediate" group remained the limitation, and it is this non-diagnostic zone that drove the shift to CTPA.[5]

The ECG in PE — non-specific but informative

The ECG is never diagnostic of PE, but it carries prognostic weight and excludes mimics. The classical S1Q3T3 (a deep S in lead I, a Q in III, and inverted T waves in III) is famous, specific when present, but insensitive — it appears in only about 1 in 10. The commoner and more useful findings are sinus tachycardia (the near-universal default), an incomplete or complete right-bundle-branch block, right-axis deviation, S waves in I and aVL, a QR in V1, and T-wave inversion in the right precordial leads V1 to V4 (the pattern that most strongly correlates with RV strain and a worse prognosis). A right-heart-strain pattern on ECG, alongside a raised troponin or BNP, flags the intermediate-risk patient for closer monitoring. [1]

[1]

The bedside echo — McConnell's sign and the RV

The focused echo is decisive in the unstable patient who cannot reach the CT. The right-ventricular strain pattern is a dilated, hypokinetic right ventricle with an RV-to-LV end-diastolic diameter ratio over 0.9, paradoxical septal bowing into the left ventricle in systole, a reduced tricuspid annular plane systolic excursion (TAPSE under 1.6 cm), and an elevated estimated pulmonary-artery systolic pressure. McConnell's sign — hypokinesia of the RV free wall with sparing of the apex — is characteristic though not pathognomonic (it also appears in RV infarct and ARDS). A completely normal RV in a haemodynamically unstable patient argues against a massive PE as the cause and redirects the search. [1]

Troponin and BNP — the biomarkers of RV strain

A raised cardiac troponin (I or T) and an elevated BNP or NT-proBNP are not diagnostic of PE — they are markers of right-ventricular myocardial strain and microinfarction. Their value is in risk stratification: a normotensive patient with RV dysfunction on imaging AND a raised troponin or BNP is intermediate-high-risk and warrants admission to a monitored bed, anticoagulation, and a low threshold for escalation if they deteriorate. A normal troponin and BNP identify the intermediate-low-risk patient with a far more benign course.[10]

[1]

Immediate management — risk-stratified, and the massive PE first

Stabilise the airway and the breathing with oxygen, and stratify by the blood pressure. [1]

[1]

[1]

For anticoagulation, a direct oral anticoagulant is now first-line for most low- and intermediate-risk PE and needs no initial parenteral agent — apixaban 10 mg twice daily for seven days then 5 mg twice daily, or rivaroxaban 15 mg twice daily for twenty-one days then 20 mg daily. Low-molecular-weight heparin (enoxaparin 1 mg per kilogram twice daily) followed by warfarin or a DOAC is the alternative for the cancer and the pregnant patient, and intravenous unfractionated heparin is chosen when rapid reversal may be needed or in renal failure. The duration is about three months for a provoked event and indefinite for an unprovoked or cancer-associated one. The absolute contraindications to thrombolysis are the same as for the acute coronary syndrome — a prior intracranial haemorrhage, a recent ischaemic stroke, an intracranial lesion, and active bleeding — and in their presence an embolectomy is the route to reperfusion. A pulmonary embolism suspected behind a cardiac arrest warrants thrombolysis during the resuscitation. [1]

Massive PE — the resuscitation that unloads the right ventricle

The shocked patient with a PE is dying of right-ventricular failure, and every intervention must be judged by its effect on the right ventricle. Oxygen relieves the hypoxic pulmonary vasoconstriction that is adding to the afterload; a vasopressor (noradrenaline first-line, with vasopressin or adrenaline as alternatives) restores the coronary perfusion pressure to the right ventricle, which is failing partly because its own wall is underperfused; a cautious fluid bolus (250 mL aliquots, reassessing) may help a genuinely underfilled ventricle but a large bolus pushes the distended RV past the flat part of the Frank-Starling curve and into worsening failure; and reperfusion (systemic thrombolysis, or embolectomy when thrombolysis is contraindicated) is the only treatment that removes the obstruction and unloads the pump. [1]

[1]

Thrombolysis for PE — the dose and the decision

Systemic alteplase is the standard fibrinolytic for the massive PE. The original dose was 100 mg over 2 hours; the contemporary ESC-favoured dose is 50 mg over 2 hours (or a weight-based 0.6 mg per kilogram bolus, capped at 50 mg), which retains efficacy with less bleeding. Tenecteplase, a single-bolus fibrinolytic, is an alternative with logistic advantages and comparable outcome data. The candidate must know the absolute contraindications — prior intracranial haemorrhage, ischaemic stroke within 6 months, intracranial or spinal lesion, active bleeding, recent brain or spinal surgery, recent major trauma — and that a relative contraindication is weighed against the lethality of the untreated massive PE. In a cardiac arrest with suspected PE, thrombolysis is given during CPR and, if the arrest reverses, the resuscitation continues with full post-ROSC care. [1]

Catheter-directed and surgical reperfusion

When systemic thrombolysis is contraindicated, has failed, or the bleeding risk is prohibitive, catheter-directed thrombolysis (ultrasound-assisted or standard, delivering a fraction of the systemic alteplase dose directly into the clot) and surgical pulmonary embolectomy are the alternatives. Catheter-directed therapy achieves clot lysis with lower systemic fibrinolytic exposure and is the emerging standard in centres with the capability; surgical embolectomy, on cardiopulmonary bypass, is reserved for the patient in whom catheter therapy is unavailable or has failed, or for a large central clot. An inferior-vena-cava filter is placed when anticoagulation is absolutely contraindicated or has failed (recurrent PE despite therapeutic anticoagulation); it is a bridge to the resumption of anticoagulation, not a substitute for it, and a retrievable filter is preferred. [1]

Anticoagulation — the DOACs and the exceptions

For the haemodynamically stable PE (low- and intermediate-risk), a direct oral anticoagulant is first-line globally, with no requirement for initial parenteral heparin. Apixaban 10 mg twice daily for 7 days then 5 mg twice daily, and rivaroxaban 15 mg twice daily for 21 days then 20 mg daily, are the two single-drug DOAC regimens; both were validated against conventional enoxaparin-warfarin in large non-inferiority trials (AMPLIFY for apixaban, EINSTEIN-PE for rivaroxaban) with less major bleeding.[11][9] The principal exceptions — where a low-molecular-weight heparin (enoxaparin 1 mg per kilogram twice daily) or unfractionated heparin is preferred — are cancer (LMWH historically, though the DOACs now rival it), pregnancy (LMWH throughout), severe renal impairment (UFH, intravenous and titrated), antiphospholipid syndrome (warfarin, because of the DOAC failure signal), and the patient in whom thrombolysis or embolectomy may be needed (UFH, because it is rapidly reversible). The duration is 3 months for a provoked event, 6 months for an unprovoked event, and indefinite for recurrent unprovoked or cancer-associated VTE.

[1]

Pulmonary embolism in pregnancy — the radiation-aware pathway

Pregnancy is a prothrombotic state (increased factors II, VII, VIII, X and fibrinogen, reduced protein S, venous stasis from the gravid uterus, and a roughly five-fold rise in VTE risk), and PE remains a leading cause of maternal death. The workup is distorted: the D-dimer is physiologically raised and unhelpful; the Wells and Geneva scores perform well but the threshold to image is lower; a V/Q scan is preferred over CTPA when the chest X-ray is normal (lower radiation to maternal breast and to the fetus, and comparable accuracy); and bilateral leg compression ultrasonography may identify a DVT and, if positive in the right context, obviate the need for pulmonary imaging. The anticoagulant is weight-based low-molecular-weight heparin — enoxaparin 1 mg per kilogram twice daily — throughout pregnancy and for six weeks postpartum (a minimum of three months total); unfractionated heparin is used around delivery or if rapid reversal is needed. Warfarin is avoided throughout pregnancy because it is teratogenic (first trimester) and causes fetal and placental bleeding (second and third trimester and at delivery); it may be used postpartum and is safe in lactation. A massive PE in pregnancy warrants systemic thrombolysis (the maternal benefit outweighs the fetal bleeding risk, which is modest) and, where time and expertise allow, catheter-directed lysis or surgical embolectomy as alternatives that limit systemic fibrinolysis. [1]

[1] [1] [1]

Subtypes and special scenarios

A massive PE in cardiac arrest — typically a pulseless-electrical-activity rhythm — warrants thrombolysis during the cardiopulmonary resuscitation. Pregnancy distorts the workup (a D-dimer is unhelpful, a V/Q scan is preferred to a CT for the radiation, and weight-based low-molecular-weight heparin is the anticoagulant, with a DOAC deferred). The cancer patient needs long-term anticoagulation. The post-operative patient carries a high bleeding risk that complicates thrombolysis. An isolated subsegmental PE is contentious; it is anticoagulated when symptomatic or in a patient with risk factors. [1]

Complications and pitfalls

The complications are right-ventricular failure and arrest, recurrence, pulmonary infarction, the later chronic thromboembolic pulmonary hypertension, bleeding from anticoagulation or thrombolysis, and heparin-induced thrombocytopenia. The pitfalls are the dangerous inverse of the workup: misreading a normal oxygen saturation or D-dimer as excluding the PE; over-relying on the ECG; missing a massive PE behind an unexplained hypotension; giving a large fluid bolus to the shocked patient; delaying thrombolysis in an arrest; and not anticoagulating or not investigating the cause after a provoked event. [1]

A specific later complication is chronic thromboembolic pulmonary hypertension, which presents weeks to months afterwards with progressive dyspnoea and right-heart failure and is treated, in selected centres, by pulmonary endarterectomy; the implication for the emergency physician is to warn the discharged patient to return with new or progressive breathlessness. A second pitfall worth stating plainly is the failure to investigate an unprovoked PE for an underlying cause — an occult malignancy or a thrombophilia — because the first presentation may be the only clue, and the workup (a focused cancer screen and, in selected patients, thrombophilia testing weeks after the anticoagulation has settled) changes the long-term management. [1]

Prognosis and disposition

The mortality of a massive untreated pulmonary embolism is high — up to 30 to 50 per cent — and falls sharply with prompt reperfusion; a low-risk PE is managed with anticoagulation and, in selected patients, discharge. The patient is admitted for anticoagulation, the duration of which is about three months for a provoked event and longer for an unprovoked or a cancer-associated one; an unprovoked PE prompts a search for an occult cancer or a thrombophilia; and an inferior-vena-cava filter is reserved for the patient in whom anticoagulation is contraindicated. [1]

Special populations

Pregnancy uses the radiation-aware pathway — a ventilation–perfusion scan over a CT, weight-based low-molecular-weight heparin (enoxaparin 1 mg per kilogram twice daily, withheld at the time of delivery), and a DOAC deferred until after delivery — and reserves systemic thrombolysis for a life-threatening massive PE. Cancer warrants long-term anticoagulation. Renal failure dictates unfractionated heparin and careful dose-adjustment. The elderly more often present atypically and carry a higher bleeding risk with anticoagulation and thrombolysis alike, so the risk-stratification and the reperfusion decision weigh the bleeding risk explicitly. [1]

Pivotal trials and the evidence base

Evidence and regional guidelines

The contemporary framework is the 2019 ESC acute-PE guideline; the rule-out evidence is the PERC criteria.[1] The high-risk-stratification and the reperfusion evidence is summarised in the high-risk-PE outcome literature.[2] The DOACs are first-line globally; the thrombolysis dose and the pregnancy pathway are regional and protocol-specific.

Exam practice

[1] [1]

Exam pearls

• Massive PE = shock → thrombolysis (alteplase 50 mg over 2 hours) + anticoagulation, or embolectomy if contraindicated.
• Wells → D-dimer (low-risk) → CTPA; PERC to rule out without a blood test.
• A normal SpO₂ and a normal ECG do not exclude a PE.
• ECG: S1Q3T3, right-bundle-branch block; echo: right-ventricular strain, McConnell's sign.
• Cautious fluids in the shocked PE — a large bolus fails the right ventricle.
• DOAC first-line; pregnancy = V/Q scan + low-molecular-weight heparin.
• A pulmonary embolism is a leading reversible cause of pulseless-electrical-activity arrest — thrombolysis during CPR if suspected. [1]

Additional high-yield clinical pearls

[1] [1] [1] [1] [1] [1]

Red flags

[1] [1]

References

1. [1]Kline JA, Courtney DM, Kabrhel C, et al. Prospective multicenter evaluation of the pulmonary embolism rule-out criteria J Thromb Haemost, 2008.PMID 18318689
2. [2]Ebner M, Böttler M, Røsjø E, et al. Outcome of patients with different clinical presentations of high-risk pulmonary embolism Eur Heart J Acute Cardiovasc Care, 2021.PMID 34125186
3. [3]Wells PS, Anderson DR, Rodger M, et al. Derivation of a simple clinical model to categorize patients probability of pulmonary embolism: increasing the models utility with the SimpliRED D-dimer Thromb Haemost, 2000.PMID 10744147
4. [4]van Belle A, Buller HR, Huisman MV, et al. Effectiveness of managing suspected pulmonary embolism using an algorithm combining clinical probability, D-dimer testing, and computed tomography JAMA, 2006.PMID 16403929
5. [5]The PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism. Results of the prospective investigation of pulmonary embolism diagnosis (PIOPED) JAMA, 1990.PMID 2332918
6. [6]Righini M, Van Es J, Den Exter PL, et al. Age-adjusted D-dimer cutoff levels to rule out pulmonary embolism: the ADJUST-PE study JAMA, 2014.PMID 24643601
7. [7]Konstantinides S, Geibel A, Heusel G, et al. Heparin plus alteplase compared with heparin alone in patients with submassive pulmonary embolism N Engl J Med, 2002.PMID 12374874
8. [8]Meyer G, Vicaut E, Danays T, et al. Fibrinolysis for patients with intermediate-risk pulmonary embolism N Engl J Med, 2014.PMID 24716681
9. [9]Buller HR, Prins MH, Lensin AW, et al. Oral rivaroxaban for the treatment of symptomatic pulmonary embolism N Engl J Med, 2012.PMID 22449293
10. [10]Konstantinides SV, Meyer G, Becattini C, et al. 2019 ESC Guidelines for the diagnosis and management of acute pulmonary embolism developed in collaboration with the European Respiratory Society (ERS) Eur Heart J, 2020.PMID 31504429
11. [11]Agnelli G, Buller HR, Cohen A, et al. Oral apixaban for the treatment of acute venous thromboembolism N Engl J Med, 2013.PMID 23808982