Ix of suspected pulmonary embolism (PE)

see also:

• pulmonary embolism (PE)
• Mx of confirmed PE
• deep venous thrombosis (DVT)
• Mx of suspected DVT
• V/Q ventilation-perfusion lung scans
• template to assist in the discharge documentation of suspected PE SSU admits
• MJA 2007: simplifying the Dx of PE
• AFTB lectures - venous thromboembolism
• Two-level Well's score for DVT/PE
• 2017: PE and syncope - the maths - D-Dimer or no D-Dimer?
• see AHA 2026: 2026 AHA/ACC/ACCP/ACEP/CHEST/SCAI/SHM/SIR/SVM/SVN Guideline for the Evaluation and Management of Acute Pulmonary Embolism in Adults: A Report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines

• investigating a patient for suspected pulmonary embolism to one of the major common conundrums facing emergency doctors who must use their experience, knowledge and judgement to:
  • avoid the harm and costs of embarking upon high radiation exposure investigations and subsequent potentially unnecessary anticoagulation for a false positive when risk of patient having PE is too low
    • it is considered acceptable to miss 2% of significant PEs and this prevalence rate correlates with a Wells score < 4 with negative d-Dimer ¹⁾
    • the risk of CTPA or V/Q generally equal the benefits of doing the test if positive PE rates exceed 15%
    • asymptomatic PEs generally should not be treated and risk of investigating outweighs benefits ²⁾
    • if aortic dissection is in the differentials and a CTPA is negative for PE but unable to exclude aortic disease, a CT aortogram may still be needed which further increases radiation and contrast risks
  • avoid the harm of not investigating and not treating significant PEs when the risk of patient having a significant PE is too high

• the advent of newer imaging technologies allows one to detect even smaller PE's but these may not present significant harm to the patient even if left untreated
• even finding a “positive PE” on the newest technologies may have a very high “false positive” rate of over 50% if the pre-test probability was low, thus in this group, perfoming a V/Q or CT-PA, not only exposes the patient to the risks of these tests, but then commits them potentially unneccesarily to the substantial risks of 12 months of anticoagulation with warfarin.
• not anticoagulating a patient with a substantial PE may cause harm!
• see also: first, do no harm

• do NOT expose your patient to the potential risks of V/Q, CT-PA or anticoagulation without good cause
  • ONLY perform these investigations on those patients with a reasonable pre-test probability of PE (this should probably be around 15% or higher to consider investigating and treating PE)
• THUS, if stable, the 1st aim is to try to assess the pre-test probability using the Well's score, PERC rules +/- D-Dimer if appropriate
• THEN, one should assess the potential risk of long term anticoagulation and whether the patient would be at high mortality risk from that than the potential PE
• THEN, one needs to discuss with the patient the pros and cons of whether they would accept the risks of the investigations and anticoagulation
• patient's with a pre-test probability of PE of less than 20% or a high risk of mortality from anticoagulation, should probably be advised to accept the chance of having a “missed PE” as a necessary “evil” but safer than the “evils” of radiologic investigation and anticoagulation. If they re-present with more obvious evidence of PE then Ix and Rx options should be re-assessed in light of changed pre-test probabilities.
• only once the above are considered should one be embarking on V/Q or CT-PA

• this is primarily a clinician gestalt assessment with some assistance from Well's score and PERC rule

• Wells scoring for pre-test probability of PE
• MDCalc online calculator
• see MJA 2007 - utility of the Well's score in Australia

• 3pts = clinical signs and symptoms of DVT (minimum of leg-swelling and pain with palpation of the deep veins)
• 3pts = an alternative diagnosis is less likely than PE (ie. clinician gestalt for PE is >10% probability)
• 1.5pts = heart rate greater than 100 beats/min
• 1.5pts = immobilisation (complete bedrest for ≥ 3 days in the 4 weeks before presentation) or surgery in the previous 4 weeks
• 1.5pts = previous objectively diagnosed DVT or PE
• 1.0pts = haemoptysis
• 1.0pts = malignancy (receiving treatment, treated in the last 6 months or receiving palliative care)

• score < 2 points = low pre-test probability = 3-4% chance of PE
• score 2-6 points = moderate pre-test probability = 9-25% chance of PE
• score > 6 points = high pre-test probability = 40-95% (mean ~67%) chance of PE

• score < 4 points = PE unlikely
• score 4 pts or more = PE likely

• Revised Geneva:
  • Low: 0-3
  • Intermediate: 4-10
  • High: ≥11
• Simplified Revised Geneva:
  • Low: 0-1
  • Intermediate: 2-4
  • High: 5-7
  • Unlikely: 0-2
  • Likely: 3-7

• a simple assessment for low pre-test probability (after Kline et al³⁾):
  • patient < 50yrs age
  • heart rate < 100
  • no unexplained hypoxia (ie. SaO2 > 95% on room air)
  • no PH DVT or PE
  • no haemoptysis
  • no calf swelling 
  • no proven DVT
  • not on exogenous oestrogen Rx such as OCP or HRT
  • no recent general anaesthetic or trauma in past 4wks
• if all of the above are satisfied (ie. “PERC negative”) and Well's < 2 then:
  • probability of a PE or DVT is < 1% if the doctor also has a gestalt feeling that PE or DVT is unlikely⁴⁾, in which case even a D-Dimer is not really needed to exclude a PE.
  • a normal D-Dimer will be reasonably adequate to exclude PE (Kline et al) and V/Q or CTPA will not be needed.
• if any of the above are NOT satisfied, then do a D-Dimer if no C/I

• in general, DO NOT ORDER a D-Dimer if either:
  • false positives likely due to other conditions that cause a raised D-Dimer:
    • recent surgery
    • high C reactive protein (CRP)
    • pleural effusion or ascites
    • inflammation / infection (including cellulitis of the leg - D-Dimer will be positive in ~75% even in absence of DVT)
    • haemorrhage / known clots
  • very low probability PE
    • if the patient has Wells < 2 and is “PERC negative” and if your gestalt is that PE is unlikely - in these patients, PE investigation is NOT indicated as risks will outweigh the benefits!
    • even a positive D-Dimer result, which has a positive LR of about 3 will raise your pre-test probability from 1% to about 3%, still far short of our test threshold of around 15% for doing a V/Q or CTPA
  • mod-high probability PE
    • these patients should be imaged even if D-Dimer is negative
  • D-Dimer not validated
    • pregnant patients

• exclude other causes
• assess pre-test probability primarily using clinician gestalt +/- PERC / Well's score
  • difficulties arise when PERC or D-Dimer cannot be applied
  • most of the Well's score is based on clinician gestalt but this component is not graded and is either a score of 0 or 3
  • if an elderly patient requires admission for the episode of syncope:⁵⁾
    • if clinical features of PE (eg. RR > 20, BP < 110, HR > 100), known active cancer, or probable or PH DVT then test for PE as per moderate probability PE
    • note that patients with alternative causes for syncope still had a 12% probability of having a PE causing the syncope and their PEs they tend to be more proximal and life-threatening
    • if no clinical features PE/DVT/active cancer, then probability of “occult PE” is around 5% and if you believe the mortality benefit of anticoagulation is 3.2% or more which is probably the case, then we should consider testing with D-Dimer as per low probability PE
  • very low probability PE
    • if PERC negative and Well's < 2 (ie. low or very low probability gestalt)
    • ⇒ NO further Ix
  • low-intermediate probability PE
    • Well's 2-4 or Well's < 2 but PERC positive:
      • do D-Dimer:
        • if D-Dimer “negative” as per YEARS ⇒ NO further Ix
        • if D-Dimer “positive” as per YEARS then V/Q or CTPA
  • moderate or high probability PE
    • Well's > 4 points
      • examples:
        • most patients with risk factors for DVT and unexplained chest pain, SOB, hypoxia or persistent tachycardia would probably fall into this group
    • DO NOT bother with D-Dimer!
    • ⇒ anticoagulate, do V/Q or CTPA if stable, otherwise consider ECHO
      • if high probability and scan negative, consider USS legs or repeat imaging if ongoing symptoms are unexplained

• if critically ill and high clinical probability:
  • CTPA is not needed as PE can be confirmed by echocardiographic evidence of RV dysfunction
  • absence of RV dysfunction on echo in a critically ill patient excludes PE as a cause of the instability⁶⁾
  • if PE is confirmed, consider urgent intervention - either:
    • thrombolysis
    • catheter embolectomy, or,
    • surgery
• only perform a D-Dimer if low pre-test probability and no reason for a raised D-Dimer (see above)

• a low pre-test probability with -ve D-Dimer gives a probability of PE < 2% and consequences of further imaging in these patients in terms of false positive CTPA or V/Q, or radiation/nephrotoxic effects of the imaging outweigh the benefits ⇒ DON'T do CTPA or V/Q.

• patients with a high pre-test probability should be started on heparin or enoxaparin whilst Ix is proceeding as EARLY anticoagulation decreases overall mortality from 30% to lower than 10%

• current 64-slice CTPA has a very low non-diagnostic rate which should minimise the need for serial imaging

• BUT if imaging IS non-diagnostic and pre-test probability is mod or high, then further imaging or serial imaging should be considered such as Doppler USS of leg(s) to detect DVT

• if there is a marked discrepancy between pre-test probability of PE and imaging results, then further imaging should be considered.

• pulmonary angiography is no longer performed in Australia even though it once was the gold standard for Dx of PE

• NB. when requesting CXR:
  • ask for well inspired PA film and if R pleuritic pain then R. lateral or if L. pleuritic pain then L. lateral
  • ensure fetus is shielded if pregnant
• NB. CT-PA:
  • contra-indications:
    • allergy to contrast
    • ? pregnancy
  • relative C/I:
    • renal failure, metformin or dehydration
    • breast feeding
    • young patient
    • V/Q available (do V/Q as 1st line Ix usually)
    • out of hours in stable patients as takes long time to send images to radiologist and potential errors in off-site interpretation
  • Ideally, the positive CTPA result should be ~20% of all CTPA tests ordered
  • lower rates of positive results suggests excessive ordering with increasingly higher chance that the +ve result is actually a false +ve and exposes patients not only to radiation risk but anticoagulation risks for 12 months - hence the critical importance of careful patient selection before ordering a CTPA
  • subsegmental PEs on CTPA may not require Rx as many believe these are not clinically significant however, Mx of these may require further clinical studies.

• consider performing DVT ultrasound studies as these are not invasive, and, if:
  • positive ⇒ treat as PE
  • negative ⇒ CXR with lead shielding:
    • if CXR normal and no PH lung disease including asthma then half-dose V/Q scan:
      • if V/Q scan not diagnostic then consider helical CT PA for definitive Dx
    • if CXR abnormal or PH lung disease incl. asthma then V/Q likely to be non-diagnostic thus consider helical CT PA
• if there are no better explanation for the symptoms than PE, then most would proceed to a V/Q scan if CXR is clear
  • NB. comparative radiation doses:
    • background radiation over 9mths pregnancy = 100mrad
    • V/Q scan = 88mrad
    • CT PA = 40-80mrad depending on trimester
    • pulm. angio. = 88mrad
    • fetal radiation: V/Q scan = 0.11-0.22mGy; CTPA = 0.01-0.06mGy
    • maternal breast radiation: V/Q scan = 0.25mGy; CTPA = 35mGy per breast 
    • incidence of childhood cancers due to fetal radiation = 1 in 16,000 per mGy thus fetal risks are minimal with half-dose V/Q or CTPA 
    • breast cancer due to radiation: a single 10mGy exposure to breast in women < 35yrs increases lifetime risk of breast cancer by ~13% ie. 1.1 per 200 women instead of 1 per 200 women; ⁷⁾
      • ⇒ avoid CTPA in pregnancy if possible (and it also exposes mother & baby to iodinated IV contrast and thus neonatal hypothroidism may need to be excluded post-natally)
• ref: BMJ vol 334 Feb 2007.

• preferred Ix is normal dose V/Q scan
  • patient should continue to express milk, and can keep it in fridge for at least 24 hrs to allow time for the technetium to decay, and then can be safely given to baby
• if V/Q is not emergently available, then either:
  • treat as for PE then V/Q next day, or,
  • CTPA although higher radiation dose to breast tissue

• determine pre-test probability of PE, if pre-test probability is:
  • moderate or high pre-test probability:
    • consider immediate anticoagulation
    • if no C/I to CTPA then:
      • if age > 40 yrs and CTV available then do CTPA with follow through CTV (include pelvic CTV if pelvic disease or gynae op)
      • otherwise do CTPA and bilat. doppler US legs
    • if CTPA C/I then perform V/Q scan, if 
      • V/Q result is high and high pre-test probability ⇒ treat as PE
      • any other result ⇒ DVT studies which if:
        • +ve ⇒ treat as PE
        • -ve or uncertain ⇒ seek advice
  • low pre-test probability:
    • if no clinical evidence of DVT then:
      • D-Dimers if available:
        • if negative then PE very unlikely thus allow home with close follow up
        • if positive then further Ix
    • otherwise perform V/Q scan or helical CT PA, 
      • if V/Q result is normal ⇒ send home with close follow up
      • if V/Q result is any other result ⇒ DVT studies which if:
        • +ve ⇒ treat as PE
        • -ve or uncertain:
          • post-test probability PE still too high to justify abandoning Ix, thus
            • consider helical CT PA for definitive Dx
• if PE diagnosed then:
  • see Mx of confirmed PE
  • evidence of RV dysfunction is an indication for hospital admission
  • evidence of RV dysfunction with raised troponin indicating injury is usually an indication for ICU admission &/or thrombolysis
  • if R&L ventricular ratios > 1.0 or large clot burden (eg. pulm. obstruction > 50% or large DVT) then
    • standard anticoagulation therapy plus troponin levels
    • consider urgent Echo with 12-24hrs
    • monitor SaO2, ECG, NIBP
    • consider thrombolysis if deteriorates
    • otherwise standard anticoagulation therapy and consider early discharge

• an option for those in whom a CTPA is not desirable (esp. young adults with normal CXRs)
• CXR should be performed FIRST to exclude obvious pathology which would make a V/Q more difficult to interpret
• perfusion only scans
  • these may be required if potential infection transmission to other patients in ventilation phase (eg. positive Covid-19, active TB, etc)
  • 80-85% sensitive if normal CXR ⁸⁾
  • a preferred option for young adults with normal CXRs
• perfusion and ventilation scans
  • more sensitive and specific than perfusion only scans HOWEVER, the ventilation phase may be superfluous in young adults with normal CXRs ⁹⁾

• risks of radiation and contrast
• inadequate technical quality can be an issue in which case only larger PEs could be excluded and repeat scans may be indicated (or a V/Q scan)