User Tools

Site Tools


pulmonary embolism (PE)


  • deep venous thrombosis (DVT) & PE are much more common than is usually realised
  • without anticoagulant prophylaxis, acute DVT's are found in (50% of these will have PE):
    • 10-13% of all general medical pts admitted for bedrest > 1wk
    • 29-33% of medical ICU pts
    • 20-26% of pulm. disease pts kept in bed >= 3 days
    • 27-33% of CCU pts post-AMI
    • 48% of asymptomatic pts post-CAG surgery
  • prophylactic heparin given to medical pts admitted via ED reduced overall mortality by 31%
  • having an autopsy tends to double the likelihood that PE will be listed on death certificate
  • PE was found in 18.5% of consecutive autopsies & were responsible for 11% of all deaths
  • PE is 2nd only to AMI as a cause of sudden unexpected natural death at any age & the Dx is especially likely to be missed in older pts:
    • in pts dying of massive PE, correct Dx made in 30% overall, but only 10% in those aged > 70yrs
  • PE is the single most common unsuspected cause of death in pts who die while on ventilator
  • Bedell found a major missed diagnosis in 14% of pts who dies in hospital & had autopsy after CPR:
    • 90% of these major missed diagnoses were PE's or bowel ischaemia
  • at autopsy of pts clinically diagnosed with primary pulm. HT, 57% will have chronic pulm. thromboembolism rather than primary pulm. HT (the Mx is different for each of these diseases!!)
  • The bad news about thromboembolism:
    • most DVTs are asymptomatic
    • most DVTs produce PE
    • most PEs are asymptomatic
    • many PEs are not detectable by current imaging techniques
    • many DVTs are not detectable by current imaging techniques
    • many non-detectable DVTs & PEs will progress, with recurrences that can cause sudden death and chronic morbidity


  • 10% of pts with acute PE will die in 1st 60min
    • presence of shock increases mortality 3-7x, with most deaths occurring within 1hr presentation1)
  • mortality rates in the 1st 3 days following diagnosis depends upon severity of presentation2):
    • cardiac arrest on arrival = ~95%
    • requiring mechanical ventilation or CPR = ~80%
    • syncope = ~40%
    • haemodynamically stable patients who have echocardiographic evidence of right ventricular dysfunction = 15% (and a further 8% mortality within 1 yr if survive to discharge from hospital)3)
    • all “stable” patients who had CT-PA or V/Q = 1-2%
  • of those patients who do not present with massive PE:
    • 1/3rd will be eventually diagnosed & Rx:
      • of these 1 in 12 will die from massive PE or complications of PE
    • 2/3rds will remain undiagnosed (ie. 400,000 cases /yr in USA):
      • of these, 1/3rd will die
  • survivors will be at risk of:
    • recurrent PE's
      • risk is dependent upon:
        • persisting underlying cause for thrombosis (eg. cancer, thrombophilia)
        • “unprovoked” PE
        • elevated D-Dimer after cessation of anticoagulant Rx
        • compliance with anticoagulation Rx
    • chronic thromboembolic pulmonary hypertension (CTEPH)
      • 4% of treated patients by 2 yrs after PE
      • risk depends upon:
        • more than one episode of PE (odds ratio, 19.0)
        • younger age (odds ratio, 1.79 per decade)
        • a larger perfusion defect (odds ratio, 2.22 per decile decrement in perfusion)
        • idiopathic pulmonary embolism at presentation (odds ratio, 5.70)4)



  • apart from gynae. surgery pts & major trauma pts, lower extremity venous thrombosis virtually always starts in the calf veins, and propagates above the knee in 87% of cases with symptomatic DVT
  • in pts with PE, 70% have DVT evident on venograms (?50% on US) although up to 90% of all PEs arise from pelvic or lower DVTs
  • risk of embolisation from DVTs:
    • ileofemoral: ~100%
    • popliteal-femoral: 60-80% (half of these PEs will be asymptomatic) - degree of proximity not important
    • calf veins: 33-46%
      • 35% of pts with PE
      • 25% of lethal PEs
      • 33% of “serious” PEs (NB. all PE's are clinically significant as minor ones may just be sentinal events!)
non-thrombotic PEs:
  • amniotic fluid embolism:
    • most commonly occurs near end of 1st stage of labour, but also in abortion or immediate post-partum period
    • Mx:
      • empty the uterus
      • Rx DIC which contributes to the 80% mortality rate
  • fat embolism:
    • most commonly occur after fractures of long bones
    • fat globules pass through pulm. circulation to enter systemic circulation where end-organs are affected via obstruction of end-capillaries causing petechiae, CNS symptoms (headache, irritability, convulsions, coma), ARDS, thrombocytopenia
    • Mx:
      • high-dose steroid Rx
      • heparin is C/I as theoretical risk of increasing production of toxic fatty acids in the lungs.
  • air embolism:
    • generally iatrogenic, but may be from pressurised underwater breathing apparatus, air-powered drills, vaginal insufflation (in orogenital sex, esp. in pregnancy)
    • often venous (causing circulatory collapse) and arterial (causing CNS impairment)
    • clinical syndrome:
      • loud churning murmur over precordium
      • air bubbles within heart may be seen on echo
      • circulatory collapse may result from as little as 5ml/kg of air forming an air lock in RV or pulm. circulation
    • Mx:
      • left lateral position in an attempt to trap air within RA
  • other embolism:
    • bile:
      • post-percutaneous transhepatic drainage for Ca pancreas
    • tumour:
      • ~20% of embolism deaths in pts with solid cancers
    • bone marrow:
      • assoc. with # long bones or sternum/ribs in CPR
    • IV drug preparations
    • broken catheters
    • parasites
    • bullets
    • cardiac vegetations

effects of PE:

  • blockage of one portion of pulm. vasculature causes increased pulm. vasc. resistance
    • ⇒ increased pulm. arterial pressures (if > 25% pulm. vasculature blocked)
      • ⇒ increased RV pressures
      • ⇒ if > 50% blocked then:
        • signif. pulm. HT results with cor pulmonale
          • in extreme cases, with PABP > 40mmHg, TI results
        • haemodynamic collapse may occur ⇒ death
    • ⇒ V/Q mismatches ⇒ nonperfused lung converted into physiologic dead space
      • ⇒ increased minute ventilation (~1% increase per 1% decrease in perfused alveoli)
        • eg. 15% increase via RR 14to16/min or 150ml increase in tidal volume ⇒ subtle!
    • pain/anxiety ⇒ further hyperventilation
    • after 24hrs, localised alveolar oedema may occur ⇒ rales
    • release of neurohumoral factors may ⇒ pulm. vasoconstriction & bronchoconstriction ⇒ rhonchi
  • small recurrent PE's:
    • pulm. arterial endothelium is relatively incapable of lysing thrombus so most emboli remain there permanently!!
      • V/Q & angio. abn. often resolve but this is by organisation & recanalisation
        • decreased pulm. vascular distensibility, thus any attempt to increase cardiac output results in increased pulm. vasc. resistance
          • ⇒ cardiac output can only increase if PABP increases enough to overcome this additional resistance
            • ⇒ pulmonary HT ⇒ chronic cor pulmonale

Clinical features of thromboembolic PE:

Predisposing factors

  • clinically apparent factor(s) present in 90% of pts with PEs:
    • past DVT/PE - single strongest risk factor for recurrence (RR 15-30x)
    • immobilisation
    • malignancy
    • stroke
    • recent surgery
    • pregnancy & post partum:
      • recognised thromboembolism occurs in 3-4/1000 births
      • accounts for 20-50% of maternal mortality
      • 75% of clinically recognised PEs occur antepartum, > 50% in 1st 15wks gestation, HOWEVER, it is likely that incidence is equal in each of the trimesters, it just gets harder to diagnose later in pregnancy
      • gradient compression stockings are strongly recommended for all pregnant pts
      • NB. post-partum endometritis:
        • risk of septic pelvic vein thrombophlebitis ⇒ septic pulmonary emboli in 40% if untreated
        • thus consider anticoagulant Rx in pts with significant post-partum endometritis
      • NB. ovarian vein thrombosis:
        • diagnosis is rarely clinically apparent & usually made laparoscopically or at C.S. or on CT
        • thrombus may extend into vena cava in which case a vena caval filter is indicated
  • cardiac failure, AMI
  • haematological factors (usually not clinically apparent)
  • obesity (> 20% more than ideal)
  • see also: haemostasis pathophysiology

Clinical presentations:

  • PE is known as “the great masquerader”
  • pleuritic chest pain
  • “pneumonia”
  • “asthma”
    • “All that wheezes is not asthma” (Chevallier Jackson). PE may appear as asthma when diffuse wheezing results from the release of vasoactive and bronchoactive mediators. 10-20% of PE patients present with wheezing.
  • “pleurisy”
  • “angina” (4-30% present & misdiagnosed as angina)
  • “AMI” (PE is 10% of deaths in pts Dx with AMI)
    • conversely, AMI is the most common cause of death when clinical Dx of PE is incorrect!
  • paradoxical embolism (usually from calf DVTs which tend have smaller emboli):
    • 27% adults have patent foramen ovale
  • pulmonary infarction mimicking “bronchogenic carcinoma” or if cavitates, a “lung abscess” on CXR or V/Q
  • “psychosomatic hyperventilation” or “anxiety disorder”
  • collapse, cardiac arrest with PEA and death - within 2 hrs of onset of massive PE

Historical features:

  • dyspnoea (73%)
  • pleuritic pain (66%)
  • cough (37%)
  • leg oedema (28%)
  • leg pain (26%)
  • haemoptysis (13%)
  • palpitations (10%)
  • wheeze (9%)

Examination features:

  • tachypnoea (70%)
  • rales (51%)
  • tachycardia (30-44%)
  • elevated neck veins (31%)
  • S4 (24%)
  • loud S2 (23%)
  • cyanosis (11%)
  • sweating (11%)
  • rhonchi (5%)
  • pleural rub (3%)

non-specific investigations:


  • increased A-a gradient (75-90%):
    • see under hypoxia
    • normal A-a gradient:
      • in air:
        • < 10mmHg in young adults
        • < 20mmHg in elderly
        • < 4+ age/4 on air (Skorodin)
      • < 10% pIO2
    • A-a gradient = alveolar pO2 - arterial pO2
      • alveolar pO2 = (760-47)xFiO2 - arterial PCO2/R, where R usually = 0.8
      • may be normal in 23% of pts with PE
      • if abnormal, is more likely to be due to other pulmonary pathology rather than PE
      • thus although better than pO2 or pCO2, it has poor predictive value
  • pO2:
    • has zero predictive value (as does SaO2)
    • 20% have pO2 > 80mmHg in air & 5% have pO2 > 100mmHg in air
    • in pts with suspected PE (ie. probability of PE ~50%):
      • the incidence of PE is higher in the group with pO2 higher rather than lower than an arbitrary pO2 cutoff value (whether 65, 70 or 80mmHg)
  • low pCO2:
    • poor marker - 35% PE's vs 33% no PE!


  • abnormal in 70% but non-specific abnormalities due to right heart strain:
  • acute PE:
  • chronic PE:
    • RAH (7%)
    • RVH (6%)
    • T wave inversion (esp. III, V1-2) (40%) - if T inv. in both ant. & inf. leads this is highly specific for PE (75%) - Marriott
    • pseudoinfarction (11%)


  • usually abnormal (84%) but mostly subtle abnormalities:
    • atelectasis or parenchymal abnormality (68%) - may take 3 days to develop
    • pleural effusion (48%)
    • pleural based opacity (35%)
      • eg. Hampton's hump (rare):
        • pleural based density/lung consolidation with a rounded border pointing to hilum
    • elevated hemidiaphragm (30%)
    • reduced pulmonary vasculature (21%)
    • prominent central pulmonary arteries (14%)
      • Westermark's sign (7%)
        • dilated pulmonary outflow tract on side of embolism with area of decreased perfusion distal to it
    • pulmonary oedema (4%, BUT 13% pts without PE had it!!)
  • useful in diagnosis of other conditions (eg. pneumonia, APO, pneumothorax)
  • if “normal” in setting of hypoxaemia then suggestive of PE

lung scans for Dx of PE

V/Q scan


  • PIOPED II study: sens 83% and spec 96% for PE thus positive likelihood ratio (LR) = 0.83/(1-0.96) = 21
pe.txt · Last modified: 2018/10/10 16:53 (external edit)