Pulmonary Embolism

Pulmonary embolism (PE) is what happens when a blood clot — usually born in the deep veins of the leg — breaks loose, travels through the right side of the heart, and lodges in the pulmonary arteries, suddenly obstructing blood flow to the lungs. It is the third most common cause of cardiovascular death after heart attack and stroke, and it is famously a great mimic: it can present as anything from a trivial twinge of pleuritic pain to sudden collapse and death. What makes PE both dangerous and rewarding to learn is that it is frequently missed and highly treatable — the difference between the two often comes down to whether the clinician thought of it at all.

This page treats PE and deep vein thrombosis (DVT) as two faces of a single disease, venous thromboembolism (VTE). Understand the shared pathophysiology, learn to gauge pre-test probability, use the right test in the right patient, and match the intensity of treatment to the severity of the embolism, and you will manage the great majority of cases confidently.

Learning Objectives

• Explain venous thromboembolism as a unified disease and describe Virchow's triad as its mechanistic foundation.
• Enumerate the major provoked and unprovoked risk factors for VTE.
• Recognise the varied clinical presentations of PE, from subtle to catastrophic.
• Apply a structured diagnostic pathway using pre-test probability (Wells score), D-dimer, and imaging.
• Risk-stratify confirmed PE into high-, intermediate-, and low-risk categories.
• Choose and dose appropriate anticoagulation and know when reperfusion (thrombolysis) is indicated.

Quick Answer

Pulmonary embolism is obstruction of the pulmonary arterial tree, almost always by thrombus embolising from a lower-limb or pelvic deep vein. The unifying mechanism is Virchow's triad: venous stasis, endothelial injury, and hypercoagulability. Classic symptoms are acute dyspnoea, pleuritic chest pain, and tachycardia, but presentation ranges from asymptomatic to shock. Diagnosis follows a probability-based algorithm: assess pre-test probability (e.g. Wells score), use D-dimer to rule out low-probability cases, and confirm with CT pulmonary angiography (CTPA) — or a ventilation/perfusion scan when contrast is contraindicated. Treatment is anticoagulation, most commonly a direct oral anticoagulant (DOAC); haemodynamically unstable ("massive") PE requires thrombolysis or embolectomy. Prognosis depends heavily on right ventricular strain and hypotension.

Where It Came From

For most of medical history, sudden death from a leg clot was invisible — clinicians could see swollen legs and could see people drop dead, but had no way to connect the two. The conceptual leap came from Rudolf Virchow, the German pathologist often called the father of modern pathology. In the 1840s and 1850s, working with the microscope and systematic autopsy, Virchow demonstrated that material found blocking the pulmonary arteries had travelled there from veins elsewhere in the body. He coined the term embolism (from the Greek embolos, a plug or wedge) to distinguish a clot that had migrated from one that formed in place (thrombosis).

The genuine problem Virchow was solving was causation: why do clots form in veins, and why do they kill in the lung? His answer, distilled by later authors into the phrase Virchow's triad, identified three contributing conditions — abnormal blood flow (stasis), vessel wall injury, and altered blood constituents (hypercoagulability). This framework has survived nearly two centuries because it is genuinely mechanistic: every modern risk factor for VTE maps onto one or more limbs of the triad. Immobility and long-haul travel cause stasis; surgery and central lines cause endothelial injury; cancer, pregnancy, and inherited thrombophilias cause hypercoagulability.

The therapeutic story is more recent. Heparin was discovered in 1916 by Jay McLean and William Henry Howell and entered clinical use in the 1930s–40s. The oral anticoagulant warfarin emerged from investigation of "sweet clover disease," a haemorrhagic condition of cattle traced to spoiled hay; the active compound (dicoumarol) was isolated by Karl Paul Link's laboratory at Wisconsin in the 1940s (warfarin is named for the Wisconsin Alumni Research Foundation). For over half a century warfarin was the only oral option, despite its narrow therapeutic window and constant monitoring. The 2010s brought the direct oral anticoagulants (DOACs) — dabigatran, rivaroxaban, apixaban, edoxaban — which now dominate treatment because they need no routine monitoring and have fewer interactions.

Pathophysiology: One Disease, Two Ends

VTE begins as a thrombus in a deep vein, most often in the calf, that may propagate proximally into the popliteal, femoral, or iliac veins. Proximal DVTs are the dangerous ones — they are large enough that, if they embolise, they can obstruct significant pulmonary vasculature. This is why DVT and PE are two ends of one continuum: up to half of patients with a proximal DVT have a silent PE on imaging, and many patients presenting with PE have no leg symptoms at all.

When thrombus lodges in the pulmonary arteries, two problems follow. First, gas exchange fails: ventilated lung is no longer perfused, creating dead space and a ventilation–perfusion mismatch that produces hypoxaemia and a compensatory drop in carbon dioxide (hence the classic finding of hypoxaemia with a low or normal PaCO2). Second, and more lethal, is the haemodynamic insult. A large clot burden acutely raises pulmonary vascular resistance. The thin-walled right ventricle (RV), unaccustomed to high afterload, dilates and begins to fail. As the RV dilates it bows the interventricular septum leftward, impairing left ventricular filling. Cardiac output falls, coronary perfusion of the already-strained RV drops, and a vicious spiral of RV failure, hypotension, and death can unfold within minutes. This is why right ventricular dysfunction is the central prognostic axis of PE, not the size of the clot per se.

Risk Factors

Risk factors are usefully divided into provoked (a transient or persistent trigger is present) and unprovoked (idiopathic), because this distinction drives how long anticoagulation continues.

Category	Examples	Triad limb
Surgery / trauma	Major orthopaedic surgery (hip, knee), abdominal/pelvic surgery, fractures	Stasis + endothelial injury
Immobility	Prolonged bed rest, stroke, long-haul flights, hospital admission	Stasis
Malignancy	Especially pancreatic, gastric, lung, brain; chemotherapy	Hypercoagulability
Hormonal	Oestrogen-containing contraception, hormone replacement therapy, pregnancy and postpartum	Hypercoagulability
Inherited thrombophilia	Factor V Leiden, prothrombin G20210A, protein C/S or antithrombin deficiency	Hypercoagulability
Acquired	Antiphospholipid syndrome, nephrotic syndrome, obesity, smoking, central venous catheters	Mixed
Prior VTE	Strongest single predictor of recurrence	—

A worked point on magnitude: major orthopaedic surgery without prophylaxis carries a very high VTE risk, which is why hip and knee replacements routinely receive pharmacological thromboprophylaxis. Pregnancy raises VTE risk several-fold and the risk is highest in the postpartum period. Cancer both causes VTE and worsens its prognosis, and an unprovoked VTE occasionally unmasks an occult malignancy.

Clinical Presentation

PE has no single reliable picture — this is its defining diagnostic challenge. The commonest symptoms are dyspnoea (often of sudden onset), pleuritic chest pain (sharp, worse on inspiration, from pulmonary infarction irritating the pleura), tachycardia, and tachypnoea. Cough, haemoptysis, and low-grade fever may occur. Leg swelling or pain points to the DVT source.

Think of presentation as a spectrum:

• Small/peripheral PE: pleuritic pain and mild breathlessness, normal vital signs. Easily mistaken for musculoskeletal pain or a chest infection.
• Submassive (intermediate-risk) PE: significant dyspnoea, tachycardia, hypoxaemia, and echocardiographic or biomarker evidence of RV strain, but blood pressure maintained.
• Massive (high-risk) PE: hypotension (systolic below 90 mmHg), syncope, or cardiac arrest. Syncope is an under-recognised presentation — a patient who faints and recovers may have had a large PE.

Case vignette. A 68-year-old woman, ten days after a total knee replacement, develops sudden breathlessness and feels faint while walking to the bathroom. Heart rate is 118, respiratory rate 26, oxygen saturation 89% on room air, blood pressure 128/78. Her calf is mildly swollen. The recent surgery (a strong provoking factor), the sudden dyspnoea, tachycardia, and hypoxaemia make PE highly likely — this patient warrants immediate imaging and empirical anticoagulation while awaiting confirmation.

Diagnosis: A Probability-Driven Pathway

The cardinal rule is that D-dimer and imaging are interpreted through the lens of pre-test probability. Skipping the probability step causes the two great errors: irradiating and over-anticoagulating low-risk patients, and falsely reassuring high-risk ones.

Step 1 — Estimate pre-test probability. The Wells score for PE is the most widely used tool. It assigns points for clinical signs of DVT (3), PE being the most likely diagnosis (3), heart rate over 100 (1.5), immobilisation or surgery in the past four weeks (1.5), previous DVT/PE (1.5), haemoptysis (1), and malignancy (1). It stratifies patients into "PE likely" or "PE unlikely" (or low/intermediate/high in the three-tier version).

Step 2 — Use D-dimer selectively. D-dimer is a fibrin degradation product; it is highly sensitive but poorly specific. In a patient with low/unlikely pre-test probability, a normal D-dimer effectively excludes PE and no imaging is needed — its value lies in ruling out, not ruling in. It rises non-specifically in infection, malignancy, pregnancy, trauma, and old age, so an elevated D-dimer only means imaging is required. Age-adjusted thresholds (age times 10 in patients over 50) improve specificity in the elderly. In high-probability patients, skip D-dimer and go straight to imaging — a normal result does not overturn a strong clinical suspicion.

Step 3 — Confirmatory imaging.

• CT pulmonary angiography (CTPA) is the first-line test in most centres: fast, widely available, and it visualises the clot directly while also showing alternative diagnoses. Drawbacks are contrast nephrotoxicity and radiation.
• Ventilation/perfusion (V/Q) scan is preferred when contrast is contraindicated (renal impairment, contrast allergy) and in young patients and pregnancy to reduce breast radiation; it reports probability rather than showing the clot directly.
• Compression ultrasound of the legs can confirm VTE indirectly: if a proximal DVT is found in a symptomatic patient, treatment proceeds without further chest imaging.
• Echocardiography does not diagnose PE but is invaluable at the bedside in the unstable patient, showing RV dilatation and strain and supporting a decision to thrombolyse.

Supporting (non-diagnostic) tests: the ECG classically may show sinus tachycardia (most common), or the S1Q3T3 pattern and right heart strain in larger PEs; arterial blood gas may show hypoxaemia with respiratory alkalosis. The PERC rule (Pulmonary Embolism Rule-out Criteria) can, in a genuinely low-risk patient, exclude PE without even a D-dimer if all eight criteria are negative.

Risk Stratification and Anticoagulation

Once PE is confirmed, stratify severity because it dictates treatment intensity.

• High-risk (massive): haemodynamic instability (shock/hypotension). Mortality is high.
• Intermediate-risk (submassive): normotensive but with RV dysfunction (on echo or CTPA) and/or raised cardiac biomarkers (troponin, BNP). The PESI or simplified sPESI score helps quantify 30-day mortality risk.
• Low-risk: normotensive, no RV strain, low PESI. Suitable for early discharge or outpatient management in selected patients.

Anticoagulation is the foundation of treatment for all but the highest-risk patients and should be started empirically as soon as PE is strongly suspected, before confirmation, unless bleeding risk is prohibitive.

• DOACs are now first-line for most patients. Apixaban and rivaroxaban can be started as single-drug regimens (with a higher loading dose for the first days/weeks). Dabigatran and edoxaban require an initial 5+ days of parenteral heparin ("lead-in") before switching.
• Low-molecular-weight heparin (LMWH) is preferred in pregnancy (DOACs and warfarin cross the placenta / are teratogenic) and was long standard in cancer-associated VTE, though DOACs are now often used there too (with caution in luminal GI/genitourinary cancers because of bleeding).
• Warfarin remains an option, especially in severe renal impairment, antiphospholipid syndrome, and mechanical valves; it needs bridging with heparin until the INR is therapeutic (target 2–3) and regular INR monitoring.

Reperfusion for high-risk PE: haemodynamically unstable patients need rapid clot removal. Systemic thrombolysis (e.g. alteplase) dissolves clot and is the standard where there is no major bleeding contraindication. Alternatives include catheter-directed thrombolysis/thrombectomy and surgical embolectomy. Thrombolysis is not routine in stable intermediate-risk PE because bleeding risk (including intracranial haemorrhage) usually outweighs benefit — these patients are watched closely and rescued with thrombolysis if they deteriorate.

Other measures: an inferior vena cava (IVC) filter is reserved for patients with acute VTE who cannot receive anticoagulation (e.g. active major bleeding) — it prevents further embolisation but does not treat existing clot and should be retrieved once anticoagulation is safe.

Duration of therapy hinges on provocation: a PE provoked by a transient major risk factor (e.g. surgery) is typically treated for three months; an unprovoked PE or one with a persistent risk factor (active cancer, antiphospholipid syndrome) generally warrants extended/indefinite anticoagulation, weighing recurrence risk against bleeding risk.

Real-World Applications

• Prophylaxis is where most PEs are prevented, not treated. Every hospital admission triggers a VTE risk assessment; mechanical (compression stockings, intermittent pneumatic compression) and pharmacological (LMWH) prophylaxis have dramatically reduced hospital-acquired PE.
• Travel medicine. Advising at-risk long-haul travellers to hydrate, mobilise, and consider compression stockings is everyday preventive practice.
• Contraception and HRT counselling. Prescribers routinely weigh personal and family VTE history before starting oestrogen-containing therapy.
• Outpatient PE management. Low-risk patients (validated by tools such as the Hestia criteria or sPESI) can increasingly be treated at home on a DOAC, sparing admission — a major operational shift enabled by risk stratification.

Common Mistakes

1. Ordering a D-dimer in a high-probability patient. Why it's wrong: a normal D-dimer cannot exclude PE when clinical suspicion is high — the negative predictive value only holds at low pre-test probability. Correction: in "PE likely" patients, go straight to CTPA and start anticoagulation while waiting.

2. Treating a normal D-dimer as reassuring in the wrong setting. Why it's wrong: D-dimer is a rule-out test only when probability is low; using it as a general "PE screen" in unselected patients causes missed diagnoses and pointless imaging. Correction: always calculate pre-test probability first.

3. Waiting for imaging confirmation before anticoagulating. Why it's wrong: a large PE can kill in the hours it takes to arrange a scan. Correction: if suspicion is high and bleeding risk acceptable, anticoagulate empirically and confirm afterwards.

4. Thrombolysing every submassive PE. Why it's wrong: routine thrombolysis in stable intermediate-risk patients increases major and intracranial bleeding without a clear mortality benefit. Correction: reserve thrombolysis for haemodynamic instability (or deterioration), and monitor intermediate-risk patients closely.

5. Stopping anticoagulation at three months for an unprovoked PE without reassessment. Why it's wrong: unprovoked VTE has a high recurrence rate off treatment. Correction: explicitly weigh extended anticoagulation.

Comparison and Connections

Feature	Deep Vein Thrombosis (DVT)	Pulmonary Embolism (PE)
Location	Deep veins, usually leg/pelvis	Pulmonary arteries
Typical symptoms	Unilateral leg pain, swelling, warmth	Dyspnoea, pleuritic pain, tachycardia
First-line diagnosis	Compression ultrasound	CT pulmonary angiography
Immediate danger	Embolisation to lung	RV failure, shock, death
Shared treatment	Anticoagulation	Anticoagulation (± thrombolysis)

Distinguish PE from its mimics. Acute coronary syndrome also causes chest pain and dyspnoea but classically gives central crushing pain and different ECG/troponin dynamics. Pneumonia gives fever, productive cough, and consolidation. Pericarditis gives positional pleuritic pain with diffuse ST changes. Aortic dissection gives tearing pain radiating to the back. The overlap is real — which is exactly why a structured probability-and-testing approach matters. Note also that PE can cause a modestly raised troponin (from RV strain), a useful prognostic marker rather than evidence of a heart attack.

Practice Questions

Recall

Q: Name the three components of Virchow's triad. A: Venous stasis (abnormal blood flow), endothelial/vessel wall injury, and hypercoagulability (altered blood constituents).

Understanding

Q: Why does a normal D-dimer exclude PE in a low-probability patient but not in a high-probability one? A: D-dimer is highly sensitive but non-specific, so a normal value has a high negative predictive value only when the prior probability of disease is already low. When pre-test probability is high, the residual (post-test) probability after a normal D-dimer remains too high to safely rule out PE, so imaging is still required.

Application

Q: A 30-year-old woman, 32 weeks pregnant, presents with sudden pleuritic chest pain and dyspnoea; creatinine is normal. What is the preferred confirmatory imaging and why? A: A V/Q scan (or perfusion-only scan) is often preferred over CTPA in pregnancy to reduce radiation dose to breast tissue, though leg compression ultrasound is a reasonable first step if there are leg symptoms. If a proximal DVT is confirmed on ultrasound, treatment can begin without chest imaging. LMWH is the anticoagulant of choice in pregnancy.

Analysis

Q: Two patients have confirmed PE and normal blood pressure. Patient A has a normal RV on CTPA and normal troponin; Patient B has RV dilatation and a raised troponin. How does management differ, and why? A: Patient A is low-risk and may be suitable for outpatient DOAC treatment. Patient B is intermediate (submassive)-risk: normotensive but with RV strain and biomarker positivity, indicating higher short-term mortality. Patient B should be admitted and monitored closely for deterioration, with thrombolysis held in reserve (rescue therapy) rather than given routinely. The difference reflects that RV dysfunction, not clot size alone, drives prognosis.

FAQ

Is every leg clot dangerous? Not equally. Distal (calf) DVTs are less likely to embolise than proximal (popliteal/femoral/iliac) ones. Proximal DVTs and PEs are the clinically dangerous end of the spectrum and always warrant treatment; isolated distal DVTs are sometimes monitored or treated depending on symptoms and risk.

Can you have a PE with normal oxygen saturations? Yes. Especially with small emboli, saturations and even the chest X-ray can be normal. Normal oxygen levels never exclude PE — this is a classic trap.

Why is troponin, a "heart attack" marker, checked in PE? A large PE strains the right ventricle enough to leak troponin. In PE it is not a sign of a coronary event but a marker of RV strain and worse prognosis, helping to identify intermediate-risk patients.

How long will I need to take blood thinners? At least three months. If the PE was provoked by a temporary factor like surgery, three months is often enough. If it was unprovoked or you have an ongoing risk such as cancer, longer or indefinite treatment is usually recommended, balanced against your bleeding risk.

Are DOACs really as good as warfarin? For most patients with PE, DOACs are at least as effective and cause less major bleeding (notably less intracranial bleeding), with no routine monitoring. Exceptions where warfarin or LMWH are preferred include antiphospholipid syndrome, mechanical heart valves, severe renal impairment, and pregnancy.

What is a saddle embolus? A large thrombus lodged at the bifurcation of the main pulmonary artery, straddling both branches. It sounds catastrophic and can be, but the patient's clinical stability (blood pressure and RV function) matters more for management than the dramatic appearance on the scan.

Quick Revision

• VTE = DVT + PE; one disease, driven by Virchow's triad (stasis, endothelial injury, hypercoagulability).
• Classic PE: sudden dyspnoea, pleuritic pain, tachycardia — but presentation ranges from silent to shock.
• Diagnose by pre-test probability first (Wells), then D-dimer (rule-out in low-probability only), then CTPA (or V/Q if contrast contraindicated / pregnancy).
• D-dimer is sensitive, not specific: use to exclude, never to confirm.
• Risk-stratify: high-risk = hypotension; intermediate = RV strain/raised troponin but normotensive; low = neither.
• Anticoagulate all; DOACs first-line; LMWH in pregnancy; start empirically if suspicion is high.
• Thrombolysis only for high-risk (unstable) PE; IVC filter only when anticoagulation is contraindicated.
• Duration: 3 months if provoked by transient factor; extended/indefinite if unprovoked or persistent risk.

Related Topics

Prerequisites

• Pulmonology Overview
• Cardiovascular and pulmonary physiology (see ../../2._Physiology/index.md)

Related Topics

• Deep vein thrombosis and the coagulation cascade (see ../../31._Hematology/index.md)
• Anticoagulant pharmacology (see ../../5._Pharmacology/index.md)

Next Topics

• Pulmonary hypertension and chronic thromboembolic disease (see ../index.md)
• Shock and haemodynamic support (see ../../24._Critical_Care_Medicine/index.md)