CONTENTS
- Rapid reference 🚀
- Risk factors and epidemiology
- Clinical presentation of PE
- Individual tests:
- Approach to PE diagnosis
- Treatment of DVT and/or low-risk PE
- Related topics
- Questions & discussion
abbreviations used in the pulmonary section: 6
- ABPA: Allergic bronchopulmonary aspergillosis 📖
- AE-ILD: Acute exacerbation of ILD 📖
- AEP: Acute eosinophilic pneumonia 📖
- AFB: Acid fast bacilli
- AIP: Acute interstitial pneumonia (Hamman-Rich syndrome) 📖
- ANA: Antinuclear antibody 📖
- ANCA: Antineutrophil cytoplasmic antibodies 📖
- ARDS: Acute respiratory distress syndrome 📖
- ASS: Antisynthetase syndrome 📖
- BAL: Bronchoalveolar lavage 📖
- BiPAP: Bilevel positive airway pressure 📖
- CEP: Chronic eosinophilic pneumonia 📖
- CF: Cystic fibrosis 📖
- COP: Cryptogenic organizing pneumonia 📖
- CPAP: Continuous positive airway pressure 📖
- CPFE: Combined pulmonary fibrosis and emphysema 📖
- CTD-ILD: Connective tissue disease associated interstitial lung disease 📖
- CTEPH: Chronic thromboembolic pulmonary hypertension 📖
- DAD: Diffuse alveolar damage 📖
- DAH: Diffuse alveolar hemorrhage 📖
- DIP: Desquamative interstitial pneumonia 📖
- DLCO: Diffusing capacity for carbon monoxide 📖
- DRESS: Drug reaction with eosinophilia and systemic symptoms 📖
- EGPA: Eosinophilic granulomatosis with polyangiitis 📖
- FEV1: Forced expiratory volume in 1 second 📖
- FVC: Forced vital capacity 📖
- GGO: Ground glass opacity 📖
- GLILD: Granulomatous and lymphocytic interstitial lung disease 📖
- HFNC: High flow nasal cannula 📖
- HP: Hypersensitivity pneumonitis 📖
- IPAF: Interstitial pneumonia with autoimmune features 📖
- IPF: Idiopathic pulmonary fibrosis 📖
- IVIG: Intravenous immunoglobulin 📖
- LAM: Lymphangioleiomyomatosis 📖
- LIP: Lymphocytic interstitial pneumonia 📖
- MAC: Mycobacterium avium complex 📖
- MCTD: Mixed connective tissue disease 📖
- NIV: Noninvasive ventilation (including CPAP or BiPAP) 📖
- NSIP: Nonspecific interstitial pneumonia 📖
- NTM: Non-tuberculous mycobacteria 📖
- OHS: Obesity hypoventilation syndrome 📖
- OP: Organizing pneumonia 📖
- OSA: Obstructive sleep apnea 📖
- PAP: Pulmonary alveolar proteinosis 📖
- PE: Pulmonary embolism 📖
- PFT: Pulmonary function test 📖
- PLCH: Pulmonary Langerhans cell histiocytosis 📖
- PPFE: Pleuroparenchymal fibroelastosis 📖
- PPF: Progressive pulmonary fibrosis 📖
- PVOD/PCH Pulmonary veno-occlusive disease/pulmonary capillary hemangiomatosis 📖
- RB-ILD: Respiratory bronchiolitis-associated interstitial lung disease 📖
- RP-ILD: Rapidly progressive interstitial lung disease 📖
- TNF: tumor necrosis factor
- UIP: Usual interstitial pneumonia 📖
strong risk factors (odds ratio >10)
- Orthopedic:
- Hip/leg fracture.
- Hip/knee replacement.
- Surgery/trauma:
- Major general surgery.
- Major trauma.
- Spinal cord injury.
- Cardiac disease:
- Hospitalization for CHF/AF within three months.
- MI within prior three months.
- Prior venous thromboembolic disease (DVT/PE).
moderate risk factors (odds ratio 2-10)
- Arthroscopic knee surgery.
- Paralytic stroke.
- Central venous lines.
- Malignancy:
- Especially: Lung, pancreas, gastric, kidney, prostate, brain, or hematologic).
- Especially: metastatic disease.
- Chemotherapy may increase risk.
- Congestive heart failure or respiratory failure (e.g., COPD).
- Among patients admitted with COPD exacerbation, the incidence of PE may be ~15%. (27522956)
- Postpartum state (especially status post C-section), estrogen-containing oral contraceptives, hormone replacement therapy, in vitro fertilization.
- Age >80 years old.
- Thrombophilia, including:
- Autoimmune diseases.
- Infection (including pneumonia, urinary tract infection, HIV).
- Inflammatory bowel disease.
- Superficial vein thrombosis.
- Erythropoietin use.
weak risk factors (odds ratio <2)
- Bed rest >3 days.
- Prolonged air/car travel.
- Laparoscopic surgery.
- Obesity.
- Pregnancy, antepartum.
- Varicose veins.
- Traditional vascular risk factors (diabetes, hypertension).
- Older age, yet <80 years old.
hospitalized patients are generally high-risk
- Large studies of unselected ICU patients usually show that ~10% have venous thromboembolic disease.
- DVT prophylaxis reduces risk by ~50%, but doesn't eliminate risk. (23782973) Don't be fooled into believing that a patient can't have a clot because they've been on prophylaxis.
Pulmonary embolism may cause nearly any pulmonary symptom other than purulent sputum production (e.g., dyspnea, cough, hemoptysis, fever, wheeze, chest pain). Thus, listing every potential symptom may simply cause bewilderment. It's more helpful to consider a few constellations of symptoms that should be recognizable:
(1) massive/submassive PE
- Presentation may be dominated by cardiovascular sequelae:
- Syncope or presyncope.
- Bradycardia.
- Cardiac arrest (pulseless electrical activity or bradyasystolic arrest).
- Anginal substernal chest pain may occur (due to ischemia to the right ventricular myocardium).
- Diaphoresis may result from an outpouring of endogenous epinephrine. Epinephrine can also cause patients to feel and look awful (grey, ashen).
- Basic radiological studies:
- POCUS usually shows right ventricular dilation.
- Chest radiograph is generally clear (clot is located in the central pulmonary vasculature).
- Effusion usually doesn't occur.
(2) large central PE
presenting symptoms are often:
- Acute-onset dyspnea.
- Hypoxemia.
basic radiological studies:
- Chest radiograph is often clear (clot is located in the central pulmonary vasculature).
- Effusion may occur in ~25% of patients. (9377961)
(3) pulmonary infarction
basics of pulmonary infarction:
- Pulmonary infarction results from embolization of a clot into the more distal pulmonary arteries, leading to obstruction and necrosis of a bit of lung tissue.
- Pulmonary infarction can result from a smaller PE, or it may occur as a large central PE breaks apart and migrates distally. In either case, pulmonary infarction usually reflects smaller pulmonary emboli – which are usually not an immediate life-threat.
- Overall, pulmonary infarction occurs in ~1/3 to 1/6 of all patients with PE. (Murray 2022)
symptoms of pulmonary infarction
- Hemoptysis may result from necrosis of lung tissue. Hemoptysis is due to capillary oozing (not arterial hemorrhage), so it's generally mild and not life-threatening. Importantly, as a general rule patients with PE and hemoptysis should be anticoagulated (in the context of PE, mild hemoptysis is not a contraindication to anticoagulation).
- Pleuritic chest pain may result from pleural irritation.
- Low-grade fever may result from tissue necrosis. Fever is usually <100F (37.7C). Fever is <103F (39.4C) in 99.3% of cases. (10631196)
- Isolated dyspnea may be the only symptom.
radiology of pulmonary infarction
- Anatomic distribution:
- Infarction creates a wedge-shaped opacity that is always pleural-based. However, the opacity may be based in the fissure or mediastinal pleura, so it won't always appear wedge-shaped on chest radiograph.
- Lesions are usually present in the lower lobe.
- Multiple lesions can be present (which can mimic multinodular lung disease).
- CT scan appearances of pulmonary infarction may include:
- Consolidation (which may represent necrotic lung tissue).
- Ground-glass opacities (which may represent local pulmonary hemorrhage).
- Reverse halo sign 📖 may occur (a rim of consolidation, containing a central area of ground-glass opacity; figure below).
- (Cavitation may occur, but it's rare.)(Walker 2019)
- Evolution: Over ~3-5 weeks, the infarct slowly decreases in size. On CT scan, the resolving infarct may appear to have a “melting ice cube sign:”
- Pneumonia resolves in a patchy fashion.
- Pulmonary infarct decreases in size while maintaining a homogeneous appearance – like an ice cube.
radiology of pulmonary infarction: accompanying pleural effusion
- Pulmonary infarction produces a small pleural effusion in ~60% of patients. (9377961)
- Effusion is usually ipsilateral to the pulmonary infarct.
- The pleural effusion is often unimpressive and easily overlooked (e.g., causing mild blunting of the costophrenic angle).
- (More on pleural effusion due to pulmonary embolism: 📖)
(4) DVT (deep vein thrombosis)
DVT may occur by itself, or in combination with pulmonary emboli.
symptoms of DVT
- Half of DVTs are asymptomatic.
- Pain is the most common symptom.
- Other symptoms may include swelling, redness.
signs of DVT may include
- Tenderness may be present.
- Asymmetric pitting edema.
- Discoloration (erythema or cyanosis).
- Dilation of superficial veins.
- Warmth.
sensitivity of D-dimer
- D-dimer is very sensitive for acute PE/DVT (~98% sensitive).
- D-dimer will fall over time, following acute PE/DVT. Patients presenting two weeks after the acute DVT/PE may have a false-negative D-dimer.
specificity of D-dimer
- An elevated D-dimer is highly nonspecific. Causes of elevated D-dimer include:
- Infection (e.g., COVID).
- Inflammatory states (e.g., pancreatitis).
- Malignancy.
- Surgery/trauma.
- Pregnancy (D-dimer tends to elevate over time as pregnancy progresses).
- Older age.
- Cirrhosis.
- Among patients admitted to a medical ICU who don't have venous thromboembolic disease, ~80% will have an elevated D-dimer. (10708176) Thus, in the context of critical illness, D-dimer has extremely low specificity.
D-dimer units & cutoffs
- D-dimer is typically expressed in FEU (fibrin-equivalent units), but some labs use assays that express D-dimer in terms of DDU (D-dimer units). FEU are roughly twice as high as DDU:
- FEU = ~2(DDU)
- Typical cutoff values for D-dimer are:
- FEU <500 ug/L, or FEU <0.5 ug/mL.
- DDU <230-250 ug/L, or DDU <0.23-0.25 ug/mL.
- The conversion between FEU and DDU is controversial. This conversion becomes clinically relevant because if your lab uses DDU, then the conversion becomes essential to utilize age-adjusted or risk-adjusted D-dimer values. (For further discussion, see this post by Tim Soo 🌊)
age-adjusted D-dimer
- Age-adjusted D-dimer: (23645857)
- <50 years old: Use standard D-dimer cutoff values:
- FEU: <500 ug/L (or FEU <0.5 ug/mL).
- DDU: <230-250 ug/L (or DDU <0.23-0.25 ug/mL).
- >50 years old:
- FEU: Cutoff is age x 10(in FEU).
- DDU: Cutoff is roughly age x 5(in DDU).
- <50 years old: Use standard D-dimer cutoff values:
- Numerous studies have demonstrated that an age-adjusted D-dimer cutoff can increase the specificity of D-dimer, without reducing the sensitivity. This improves the diagnostic performance of D-dimer, causing it to trigger fewer unnecessary CT scans.
role of D-dimer in PE diagnosis
- If D-dimer is negative and index of suspicion for PE is low or moderate, this generally is sufficient to exclude PE.
- D-dimer is less likely to be useful among patients known to have systemic inflammation (e.g., septic shock).
- ⚠️ D-dimer should be obtained only if you are otherwise fully intending to get a CT scan.
- Arterial blood gas lacks adequate sensitivity or specificity for the diagnosis of pulmonary embolism. (11112122)
- It is possible for patients with a saddle pulmonary embolism to have a completely stone cold normal arterial blood gas (central emboli may not affect ventilation-perfusion matching, so they don't necessarily cause hypoxemia).
- ⚠️ Do not use arterial blood gas to evaluate for pulmonary embolism.
performance of DVT ultrasound to evaluate for PE
- DVT ultrasound has a sensitivity of ~40% for detecting lower extremity thrombus among patients with PE. (31504429)
- The specificity of a positive DVT scan is nearly 100% for detecting venous thromboembolic disease (DVT and/or PE). Identifying a DVT doesn't necessarily prove that the patient actually has a PE (there could be an isolated DVT). However, if positive, the treatment of DVT is generally the same as that of PE (anticoagulation). Therefore, a positive DVT study may facilitate expedited therapy, while avoiding a CT scan.
- An important exception to this is that if there is consideration for the possibility of a submassive/massive PE that may require more aggressive therapies, then a CT scan is often still required (e.g., to evaluate clot burden and differentiate massive PE from chronic pulmonary hypertension).
PE causes abnormal chest radiograph predominantly in the context of pulmonary infarction
- Pulmonary infarction may cause wedge-shaped opacities as well as pleural effusion.
- The clinico-radiographic syndrome of pulmonary infarction is discussed further above. ⚡️
PE usually doesn't cause an abnormal chest radiograph
- Usually, pulmonary embolism doesn't cause substantial abnormalities on chest radiograph.
- A patient with marked dyspnea or hypoxemia and an unremarkable radiograph should suggest the possibility of PE (especially if other possibilities have been excluded, such as bronchospasm).
sensitivity of CT angiography for PE
- Performance depends on the technical adequacy of the study, which should be assessed on an individual basis. Thus, published numbers won't apply to an individual patient. Some numbers are shown below to provide a general concept of the performance, however.
- Sensitivity for central or segmental PE is >95%.
- Sensitivity for subsegmental PE is lower.
- A modern, high-quality CT scan is probably adequate to exclude a clinically significant PE.
- 💡 When in doubt, review the images yourself and discuss with a radiologist to determine the quality of the scan and how reliable the imaging data is. Sensitivity and specificity will vary between patients, depending on the quality of the scan!
- A high-quality scan should have a density >200 Hounsfield units within the main pulmonary artery. (Walker 2019)
- 💡 MIP projections (maximal intensity projection) may be especially useful for visualizing small pulmonary emboli.
specificity of CT angiography for PE
- Specificity is high for segmental or larger pulmonary emboli.
- For an isolated subsegmental PE, there may be inter-reader variability. Thus, this finding may represent a false-positive result in some patients.
- Alternative causes of filling defect on CT scan are discussed in the section below.
role of CT angiography in PE diagnosis
- CT angiography is generally the definitive test of choice.
- Recent evidence suggests that contrast nephropathy probably doesn't exist. 📖 Thus, CT angiography shouldn't be avoided due to fears regarding renal dysfunction.
- Historically, CT angiography was sometimes avoided in pregnant women to limit radiation exposure to breast tissue. However, modern imaging techniques have reduced radiation exposure, so the lifetime cancer risk is negligible (<1/3,000). (31504429)
CT angiography followed by CT venography
- This consists of performing a CT angiography, followed by a CT scan of the pelvis and thighs to evaluate for venous clot. The same amount of IV contrast is involved, but there is greater radiation exposure.
- CT venogram generally has similar performance to ultrasonography for detection of DVT, but may be superior in some situations:
- Obese patients, who are difficult to ultrasound.
- Evaluation for very proximal clot (e.g. iliofemoral DVT).
- This may be considered in the following context:
- Older patients (who are unlikely to be harmed by additional radiation exposure).
- Logistic or technical barrier to performing compression ultrasonography (e.g., morbid obesity).
The differential diagnosis of a filling defect on CT angiography includes several possibilities. Examination of the filling defect in multiple perspectives (e.g., axial and coronal images) may help clarify the etiology.
acute pulmonary embolism
- Central filling defect surrounded by contrast.
- “Railway track” sign – endoluminal clot is surrounded by a rim of contrast on both sides.
- Peripheral filling defect that makes an acute angle with the arterial wall (in longitudinal cross-section).
- Complete vessel occlusion (although this is nonspecific and may occur in chronic thrombosis as well).
CTEPH (chronic thromboembolic pulmonary hypertension)
- Discussed here: 📖
in situ thrombosis
the most common cause is advanced pulmonary hypertension
- Advanced pulmonary hypertension of any cause may cause massive pulmonary artery dilation with subsequent development of chronic clot lining the central vessel walls.
- This in situ thrombosis is nonobstructive, so the ventilation-perfusion scan may remain normal (differentiating it from CTEPH). Thus, identification of thrombus doesn't necessarily indicate that thrombosis was the original cause of the patient's pulmonary hypertension. (Murray 2022)
other causes of in situ thrombosis
- Postsurgical (lobectomy, pneumonectomy).
- Pulmonary artery stenosis.
- Chronic atelectasis. (Walker 2019)
macroscopic pulmonary tumor embolism
- Macroscopic tumor embolization can rarely involve large, proximal pulmonary arteries.
- Macroemboli tend to result from cancers that invade large veins:
- Hepatocellular carcinoma.
- Renal cell carcinoma.
- Sarcomas.
- Right atrial myxomas.
- Radiographic features:
- Tumor may be suggested by a rounded and smoothed shape, which can appear a bit like a series of balls lodged in the pulmonary artery.
- Serial imaging may show interval tumor growth.
- Patients with large, central tumor emboli could be candidates for interventional radiology embolectomy. Embolectomy may be both therapeutic and diagnostic (if it yields tumor specimens for pathological examination).
pulmonary artery sarcoma
clinical features
- Pulmonary artery sarcoma usually occurs between ~45-55 years old, being unlikely in patients <30 years old. (24493514, 29192821)
- This tends to cause a gradual onset pulmonary hypertension, with right ventricular failure. Symptoms may include cough, dyspnea, chest or back pain, and hemoptysis. (24493514)
- Clinical clues that the patient doesn't actually have a PE:
- 🚩 Symptoms may be relatively unimpressive, in comparison to the CT scan.
- 🚩 Gradual onset of symptoms over months.
- 🚩 Constitutional symptoms may be present (e.g., fever, weight loss, and night sweats). (31699232)
- 🚩 Failure to respond to anticoagulation (following an initial misdiagnosis as a PE).
imaging features on CT scan
- Most common clues:
- Lobulated or nodular appearance is often seen.
- Single, central lesion tends to occur (unlike a pulmonary embolism, which frequently causes multiple, bilateral filling deficits).
- Filling of the entire arterial lumen (“eclipsing vessel sign”).
- More specific features that may occasionally be seen:
- Vessel expansion due to mass effect.
- Invasion of lung parenchyma or mediastinum.
- Hematogenous metastasis causing randomly distributed nodules in the lung parenchyma. (29192821)
- Calcification may occur; if encountered this argues against PE (although calcification may occur in the context of chronic thromboembolic pulmonary hypertension).
- Lesions may have low attenuation. (31699232) Heterogeneous contrast enhancement may occur (although this won't be captured on a CT angiogram). (Walker 2019)
diagnosis
- If sarcoma is suspected, then MRI and/or PET scan may help establish this diagnosis more firmly. (31699232, 26203610)
- Interventional radiology may be able to obtain an endovascular biopsy. Alternatively, if the lesion is resectable then tissue may be obtained at the time of operative intervention. (26203610)
foreign body emboli
- Cement emboli.
- Catheter embolus.
- Iodinated oil (used for lymphangiography or transcatheter oil chemoembolization of hepatocellular carcinoma).
- Brachytherapy seeds used for prostate cancer.
pulmonary artery echinococcal cyst 📖
- Radiographic clues:
- (1) Appears as a circumscribed, round lesion.
- (2) Lesion is stable over time (e.g., will fail to resolve with anticoagulation).
- This is rare, but should be considered in people who have lived in endemic regions. The disease may further be suggested by anti-echinococcal antibodies and/or peripheral eosinophilia. (34022024)
artifacts of various types
- Beam-hardening artifact (dense contrast in the superior vena cava can overlie the right pulmonary artery and medial upper lobe pulmonary arteries). (Shepard 2019)
- Slab or misalignment artifact (when examining coronal images, there is a linear “filling defect” that extends perpendicularly across the vessel). (Shepard 2019)
- Motion artifact (may be best appreciated on lung windows).
- Flow-related artifacts (margins of filling defect are poorly defined).
- In areas of consolidation or atelectasis, hypoxic vasoconstriction may cause sluggish blood flow, which causes unenhanced or poorly opacified vessels. (Shepard 2019)
- Poorly opacified pulmonary vein or mucus-plugged bronchus may be misidentified as a pulmonary artery. (Shepard 2019)
for greatest yield from a VQ scan, the following conditions should ideally be met
- The lung parenchyma should be reasonably normal (e.g., clear chest radiograph).
- There is no known chronic PE.
- The patient needs to be able to perform a breath-holding maneuver.
- The patient must be stable enough to transport to radiology without being intubated.
interpretation of a VQ scan result is nuanced
- High-probability of PE:
- Likelihood ratio for PE is ~18.
- This should usually be considered as diagnostic for PE.
- Intermediate probability:
- Likelihood ratio for PE is ~1.2.
- This is nonspecific, providing little information one way or the other.
- Low-probability of PE:
- Likelihood ratio for PE is ~0.4
- Among patients with a low pretest probability, this may be sufficient to exclude PE.
- Among patients with intermediate or high pretest probability, this doesn't exclude PE.
- Normal VQ scan:
- Likelihood ratio for PE is ~0.05
- This largely excludes pulmonary embolism.
use of VQ scan?
- There is nearly no utility for VQ scanning in the diagnosis of acute pulmonary embolism, since essentially all patients may be evaluated using CT angiography:
- Patients with renal insufficiency should receive CT angiography. 📖
- Patients with contrast allergy may receive CT angiography following steroid pretreatment.
- Reasons to avoid VQ scanning:
- VQ scan requires that patients can hold their breath, which is difficult for most critically ill patients (although some centers may be able to perform an isolated perfusion scan at the bedside). (Fishman 2023)
- VQ scans are inconclusive in about half of cases. (34481601) Overall, the combination of clinical probability and VQ scan results either confirms or excludes PE in <30% of patients. (ERS handbook 3rd ed.)
- VQ scans are unable to evaluate the thorax globally for alternative diagnoses (unlike CT scan).
- An isolated perfusion scan may occasionally be considered to exclude PE in pregnant women with normal chest radiographs, with a goal of limiting radiation exposure to breast tissue. If this is done, placement of a Foley catheter and aggressive hydration will accelerate excretion of radiotracer (and avoid radiotracer from accumulating in the bladder, adjacent to the fetus). The approach to PE diagnosis in pregnancy is discussed further here: 📖
- ⚠️ VQ scan is a niche test that is rarely appropriate for evaluation of PE. In most cases, CT angiography is preferred over VQ scan.
test performance
- Pulmonary angiography is traditionally considered the “gold standard” test, so by definition the sensitivity is high. However, this suffers from circular logic.
- Overall, pulmonary angiography may offer similar diagnostic accuracy as compared to modern CT angiography. (31504429) For example, there may be substantial interobserver variability at the level of subsegmental PE with either study.
- Invasive angiography isn't immune to problems of variable image quality and inter-rater variability. For example, flow artifacts may falsely suggest the presence of pulmonary embolism. (Murray 2022)
procedural risks
- The procedure may carry a mortality rate as high as ~0.5% (31504429)
- There is a 1.5% risk of major complications (including myocardial infarction, hypotension, renal failure, or cardiac perforation).
utility of invasive pulmonary angiography?
- In the modern era of advanced-generation CT angiography, there is little reason to perform invasive pulmonary angiography solely for diagnosis of pulmonary embolism.
- Invasive pulmonary angiography is usually performed as a component of an interventional radiology procedure (e.g., embolectomy).
start with the right question
- When approaching a patient with cardiopulmonary dysfunction, be sure to ask the correct question:
- Wrong initial question: Does this patient have a PE?
- Right question: Why does this patient have respiratory dysfunction?
- Chest radiograph and POCUS can be performed rapidly at the bedside. Often this will detect an alternative diagnosis that explains the patient's symptoms, rendering the PE question moot (e.g., pneumothorax, pneumonia, lobar atelectasis, heart failure).
Wells Score 🧮
- This is a risk-stratification tool to help determine how likely the patient is to have a PE.
- ⚠️ Wells criteria have been primarily validated among patients in the emergency department. Some studies support its use among hospitalized patients, but exercise caution with this application.
- Summate the following:
- Clinical signs & symptoms of DVT: 3 points.
- PE is as likely, or more likely, than alternative diagnoses: 3 points.
- Heart rate >100: 1.5 points.
- Surgery or immobilization (for at least 3 days) within one month: 1.5 points.
- History of DVT or PE: 1.5 points.
- Hemoptysis: 1 point.
- Active malignancy within 6 months: 1 point.
- Interpretation: (36805291)
- 0-1 = Low risk (1-3% likelihood of PE)
- 2-6 = Intermediate risk (~10% likelihood of PE)
- >6 = High risk (~40% likelihood of PE)
PERC rule (PE rule-out criteria) 🧮
- The PERC rule may be used to exclude pulmonary embolism. It is valid only among patients who present to the emergency department and are deemed low-risk of PE (e.g., Wells score 0-2).
- ⚠️ The PERC criteria shouldn't be utilized among hospitalized patients.
- PE may be excluded if all of the following criteria are met:
- Age <50.
- Heart rate <100 b/m.
- Saturation >94%.
- No hemoptysis.
- No hormone use (oral contraceptives, hormone replacement, or estrogenic hormone use in male or female patients).
- No history of surgery or trauma requiring treatment with general anesthesia within 4 weeks.
- No prior PE/DVT.
- No signs of DVT.
The following algorithm may be used to evaluate for possible PE in a patient who is stable. (For a patient in shock, see the approach to shock: 📖)
concept of the test threshold
- The testing threshold is a bedrock concept in diagnostic medicine, which is particularly relevant to the evaluation of pulmonary embolism. (7366635)
- The testing threshold is the likelihood of disease above which testing is beneficial. In the case of pulmonary embolism, the test threshold is 2%. (36194215)
- If the patient's likelihood of having a PE is <2%, then diagnostic testing is more likely to cause harm than benefit (e.g., due to complications from diagnostic tests, or false-positive results).
- It's worth noting that the likelihood of PE among patients with low-risk Wells scores (0-1 points) is ~1-3%, which is really close to the test threshold.
optimal miss rate
- Stated another way, the target “miss” rate for PE diagnosis should be 1-2%.
- If zero pulmonary emboli are missed, that's not an achievement – it implies that harm is being done due to excessive testing. Indeed, increased use of CT scanning for PE evaluation has likely caused overdiagnosis of pulmonary emboli (diagnosis of clinically insignificant emboli, leading to excessive therapy with iatrogenic harm).
(⚠️ Treatment of submassive or massive PE are discussed here: 📖)
avoid empiric anticoagulation
- When possible, it's often wise to avoid empiric anticoagulation for patients with possible PE.
- Anticoagulation carries a small, yet real risk of hemorrhage. The risk of hemorrhage generally isn't justified by the benefit of starting anticoagulation a few hours earlier (among the small subset of patients who actually will have a PE).
- If there is concern that CT scan will be delayed, a reasonable option is often to perform a bedside POCUS scan of the legs for DVT.
- If DVT is found, this is diagnostic of venous thromboembolic disease, so anticoagulation should be initiated.
- If no DVT is detected, this may suggest that the patient is at relatively low risk of immediate deterioration – and may await a CT scan within the next few hours.
low molecular-weight heparin (LMWH) is the preferred anticoagulant for stable patients
- RCTs and Cochrane analysis have shown that LMWH is generally superior to unfractionated heparin (LMWH has greater efficacy, with fewer bleeding complications). (28182249)
- Heparin is perceived to be “safer” because it can be easily turned off. This is not actually borne out in the evidence. Thus, LMWH should usually be preferred as the default option.
- Contraindications to LMWH include:
- Planned or possible procedures in the near future.
- Patient with unusually high bleeding risk (e.g., recent gastrointestinal hemorrhage).
- Renal dysfunction (e.g., GFR <30 ml/min).
- Additional information regarding low molecular-weight heparin is here: 📖
direct oral anticoagulants (DOACS)
- DOACS have been demonstrated to be safe and effective therapies for DVT and/or PE.
- DOACS are often undesirable among patients who are admitted to the hospital, since they are difficult to reverse (if the patient should require a procedure).
- A HESTIA score of zero 🧮 is generally regarded as the primary tool to determine if outpatient management is reasonable. (36805291) Of course, this must be combined with clinical judgement.
- Appropriately selected patients may be discharged home on an oral Xa inhibitor (e.g., apixaban, rivaroxaban).
- Calf vein thrombosis itself isn't dangerous, but it can propagate proximally to cause a DVT.
- For patients without risk of bleeding, anticoagulation is reasonable.
- For patients with a risk of bleeding, it may be reasonable to hold anticoagulation and perform serial ultrasonography Q48-72hr. If the clot extends proximally, then anticoagulation is indicated. However, if the clot is stable then ongoing observation is acceptable. This strategy may be useful for patients who are temporarily at higher risk of bleeding (e.g., postoperative patients).
definition
- ⚠️ Different studies have used various definitions, often exacerbating confusion. When evaluating studies, look very carefully at how they are defining isolated subsegmental PE.
- For the purpose of the following discussion, an isolated subsegmental PE is defined as meeting the following criteria:
- A single PE identified in a single subsegmental pulmonary artery.
- Ultrasonography of the legs is negative for DVT.
- There is no evidence of DVT elsewhere in the body (e.g., asymmetric arm edema with a PICC line that could suggest an upper extremity DVT).
initial approach to subsegmental PE
- The first step is to review the CT images with a thoracic radiologist. Subsegmental PE is misdiagnosed at a surprisingly high rate. An expert radiologist may overturn this diagnosis.
- There are numerous potential causes of a filling defect on CT angiography, including a variety of different artifacts. These are discussed further above: 📖
- Reconsider the pretest probability. Ultimately, the result of any test must be interpreted in light of the pretest probability. If the pretest probability is low, then an isolated subsegmental PE may be less likely to represent true disease.
- Obtain a DVT study (if there is a PE plus DVT, this generally warrants therapy). Consider obtaining upper and lower extremity studies. (Fishman 2023)
- If the CT scan is of suboptimal quality, consider repeating the scan.
- Check a D-dimer:
- A normal D-dimer supports the possibility that the sub-segmental PE is either a radiological artifact or a chronic remnant of a prior PE. In either case, this could argue against the need to anticoagulate.
- An elevated D-dimer may very weakly support the diagnosis of PE. However, D-dimer is quite nonspecific. Any positive result should be considered in the context of other causes of D-dimer elevation (as discussed above).
treatment of an isolated, subsegmental PE
- There is no simple treatment algorithm, but rather treatment needs to be tailored to each individual patient.
- Available evidence suggests that on average, treatment doesn't affect outcomes. (29498138)
- Recent ACCP (American College of Chest Physicians) guidelines suggest using clinical surveillance rather than anticoagulation in patients with subsegmental PE and no DVT and who have a low risk of recurrent venous thromboembolism (2C recommendation, indicating weak recommendation and low-quality evidence). (26867832, 31532966)
- Primary considerations include:
- Pretest probability of PE.
- Review of the PE scan & discussion with a radiologist.
- D-dimer value.
- DVT study results.
- Risk of hemorrhage with anticoagulation.
- Risk factors for future PE/DVT.
- Other indications for anticoagulation (e.g., atrial fibrillation).
- Patient's values and preferences (informed decision-making).
- If a decision is made to withhold anticoagulation, consider performing serial leg ultrasonography (e.g., 5-7 days later). This may provide an added layer of safety, to ensure that ongoing thrombosis is less likely.
- Discussed here: 📖
To keep this page small and fast, questions & discussion about this post can be found on another page here.
Guide to emoji hyperlinks
- = Link to online calculator.
- = Link to Medscape monograph about a drug.
- = Link to IBCC section about a drug.
- = Link to IBCC section covering that topic.
- = Link to FOAMed site with related information.
- 📄 = Link to open-access journal article.
- = Link to supplemental media.
References
- 24493514 Shigemitsu H, Orjioke N, Shriki JE, Varras J, Koss MN. A middle-aged man with cough, chest pain, and pulmonary artery filling defect: primary pulmonary artery sarcoma. Chest. 2014 Feb;145(2):407-410. doi: 10.1378/chest.13-1488 [PubMed]
- 26203610 Qamar A, Siddiqi HK, Rossman MD. Worth a second look. A 49-year-old woman with progressive dyspnea. Ann Am Thorac Soc. 2015 Jul;12(7):1093-6. doi: 10.1513/AnnalsATS.201502-072CC [PubMed]
- 29192821 Revelo AE, Chaudhari P, Khan G, Shakil F, Maguire G, Epelbaum O. Direct Connection: A Man with Lung Nodules and Filling Defects in the Pulmonary Arterial Tree. Ann Am Thorac Soc. 2017 Dec;14(12):1844-1848. doi: 10.1513/AnnalsATS.201706-511CC [PubMed]
- 31504429 Konstantinides SV, Meyer G, Becattini C, Bueno H, Geersing GJ, Harjola VP, Huisman MV, Humbert M, Jennings CS, Jiménez D, Kucher N, Lang IM, Lankeit M, Lorusso R, Mazzolai L, Meneveau N, Ní Áinle F, Prandoni P, Pruszczyk P, Righini M, Torbicki A, Van Belle E, Zamorano JL; ESC Scientific Document Group. 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 Jan 21;41(4):543-603. doi: 10.1093/eurheartj/ehz405 [PubMed]
- 31532966 Fernandes A, Connors JM, Carrier M. Anticoagulation for Subsegmental Pulmonary Embolism. N Engl J Med. 2019 Sep 19;381(12):1171-1174. doi: 10.1056/NEJMclde1907665 [PubMed]
- 31699232 Ziatabar S, Rabii K, Oron MN, Shah V, Altschul E, Ishikawa O, Gabra NI, Mahmoud OH, Mina BA. A 44-Year-Old Man With Dyspnea and a Pulmonary Artery Filling Defect. Chest. 2019 Nov;156(5):e103-e106. doi: 10.1016/j.chest.2019.06.026 [PubMed]
- 34022024 Ucar EY, Araz O, Kerget B, Tuna ME, Yıldızeli B, Alper F, Akgun M. A 19-Year-Old-Woman With Backache, Hemoptysis, and Hypereosinophilia. Chest. 2021 Apr;159(4):e231-e235. doi: 10.1016/j.chest.2020.10.061 [PubMed]
- 34481601 Sessums MT, Ritchie C. 37-Year-Old Woman With Shortness of Breath. Mayo Clin Proc. 2021 Sep;96(9):2480-2485. doi: 10.1016/j.mayocp.2020.12.042 [PubMed]
- 35793208 Kahn SR, de Wit K. Pulmonary Embolism. N Engl J Med. 2022 Jul 7;387(1):45-57. doi: 10.1056/NEJMcp2116489 [PubMed]
- 35953217 Trott T, Bowman J. Diagnosis and Management of Pulmonary Embolism. Emerg Med Clin North Am. 2022 Aug;40(3):565-581. doi: 10.1016/j.emc.2022.05.008 [PubMed]
- 36194215 Freund Y, Cohen-Aubart F, Bloom B. Acute Pulmonary Embolism: A Review. JAMA. 2022 Oct 4;328(13):1336-1345. doi: 10.1001/jama.2022.16815 [PubMed]
- 36344135 Murphy CG, Goldstein JM, Besharati S, Kobsa S, Salvatore MM, Rosenzweig EB, Ingham M, Del Portillo A, Takeda K, Chandra S, Furfaro D. A 52-Year-Old Man With Chest Pain and Dyspnea. Chest. 2022 Nov;162(5):e259-e264. doi: 10.1016/j.chest.2022.05.020 [PubMed]
- 36805291 Westafer LM, Long B, Gottlieb M. Managing Pulmonary Embolism. Ann Emerg Med. 2023 Sep;82(3):394-402. doi: 10.1016/j.annemergmed.2023.01.019 [PubMed]
Books:
- Shah, P. L., Herth, F. J., Lee, G., & Criner, G. J. (2018). Essentials of Clinical pulmonology. In CRC Press eBooks. https://doi.org/10.1201/9781315113807
- Shepard, JO. (2019). Thoracic Imaging The Requisites (Requisites in Radiology) (3rd ed.). Elsevier.
- Walker C & Chung JH (2019). Muller’s Imaging of the Chest: Expert Radiology Series. Elsevier.
- Palange, P., & Rohde, G. (2019). ERS Handbook of Respiratory Medicine. European Respiratory Society.
- Rosado-De-Christenson, M. L., Facr, M. L. R. M., & Martínez-Jiménez, S. (2021). Diagnostic imaging: chest. Elsevier.
- Murray & Nadel: Broaddus, V. C., Ernst, J. D., MD, King, T. E., Jr, Lazarus, S. C., Sarmiento, K. F., Schnapp, L. M., Stapleton, R. D., & Gotway, M. B. (2021). Murray & Nadel’s Textbook of Respiratory Medicine, 2-Volume set. Elsevier.
- Fishman's: Grippi, M., Antin-Ozerkis, D. E., Cruz, C. D. S., Kotloff, R., Kotton, C. N., & Pack, A. (2023). Fishman’s Pulmonary Diseases and Disorders, Sixth Edition (6th ed.). McGraw Hill / Medical.