PE pathway

1. Clinical probability. Use a validated tool (Wells, revised Geneva) or gestalt. Low <15%, intermediate 15–50%, high >50%. This tool maps three-tier Wells low (<2) to <15%, moderate (2–6) to 15–50%, and high (>6) to >50%; the guideline gives Wells <2 as the example of <15%.

2. Low (<15%) → PERC. All 8 criteria met → very low probability, no further testing. Any criterion failed → D-dimer.

3. Intermediate, or low with PERC failed → D-dimer + YEARS (COR 2a). 0 YEARS items and D-dimer <1000 ng/mL → PE excluded. ≥1 YEARS item and D-dimer <500 ng/mL or below the age-adjusted threshold → PE excluded. Otherwise image. Age-adjusted D-dimer alone (age × 10 µg/L FEU, age >50) is the stated alternative (COR 2a).

4. High (>50%) → imaging (COR 1). Do not use D-dimer to exclude PE.

Wells two-tier. ≤4 PE unlikely, >4 PE likely. In the Christopher Study strategy, "likely" went straight to CT. The 2026 pathway instead uses three probability levels, so a Wells of 4.5–6 is intermediate (D-dimer allowed) here but "likely" (straight to CT) under the two-tier strategy. The tool flags this.

Pregnancy-adapted YEARS (COR 2b; Artemis). With signs of DVT, do compression ultrasound first. If it shows DVT, treat without CTPA. Otherwise use YEARS thresholds of 1000 / 500 ng/mL, with no age adjustment and no PERC. If imaging is needed and the chest x-ray is normal, low-dose CTPA is reasonable over low-dose perfusion scintigraphy (COR 2a). In Artemis, CTPA was avoided in 65% of patients in the first trimester and 32% in the third.

Three-tier: low <2, moderate 2–6, high >6. Two-tier: unlikely ≤4, likely >4.

PERC (all must be met; use only when gestalt probability is <15%): age <50 · HR <100 · SpO₂ ≥95% · no hemoptysis · no estrogen use · no prior DVT/PE · no unilateral leg swelling · no surgery/trauma requiring hospitalization within 4 weeks.

YEARS items: clinical signs of DVT · hemoptysis · PE the most likely diagnosis.

The 500 / 1000 ng/mL and age × 10 thresholds come from studies using assays that report fibrinogen-equivalent units (FEU). 1 ng/mL = 1 µg/L; 1 µg/mL = 1 mg/L = 1000 ng/mL.

D-dimer units (DDU) are not interchangeable with FEU, and the conversion is assay-specific. If your lab reports DDU, the tool does not apply these thresholds. Use the cutoff your laboratory has validated.

Respiratory modifier R (e.g., C3R, D2R). For C: SpO₂ <90%, RR ≥30, or any supplemental O₂ need. For D: >6 L nasal cannula or a non-rebreather mask. For E: hypoxemic or ventilatory failure, meaning noninvasive or invasive positive-pressure ventilation. The guideline says R "can be added to a subcategory or designated its own subcategory." It gives no rule for moving between C, D and E on respiratory status alone, so the tool shows a forced warning instead of upgrading. It does upgrade one case: a patient with sPESI 0 who needs supplemental O₂ meets the Hestia oxygen criterion (Hestia ≥1 = elevated severity score), so moves from A/B to at least Category C.

Categories are assigned by the most severe clinical, laboratory, or imaging indicator. Reassess them over time.

COR = class of recommendation. 2b* = benefit unclear. 2b = may be considered. 3-NB = no benefit. Lysis = systemic thrombolysis, only with an acceptable bleeding risk. CDL = catheter-directed thrombolysis. MT = mechanical thrombectomy. Surg = surgical embolectomy. For E2, VA-ECMO is reasonable where resources allow (2a).

High: hemodynamic instability, meaning cardiac arrest, obstructive shock (SBP <90 mm Hg or vasopressors needed despite adequate filling, with end-organ hypoperfusion), or persistent hypotension (SBP <90 or a drop ≥40 mm Hg for >15 min, not caused by new arrhythmia, hypovolemia or sepsis). Troponin and PESI are not needed.

Intermediate-high: PESI III–V or sPESI ≥1, with both RV dysfunction on TTE/CTPA and elevated troponin.

Intermediate-low: PESI III–V or sPESI ≥1, with one or neither.

Low: PESI I–II or sPESI 0, with no RV dysfunction or troponin elevation if assessed. An ESC footnote places patients with a low PESI/sPESI but RV dysfunction or troponin elevation in the intermediate-risk category. It does not say whether that means intermediate-high or intermediate-low.

ESC "high risk" requires hypotension. Normotensive shock (2026 D2) and transient hypotension (D1) have no separate ESC class.

• PERC misuse. PERC applies only when gestalt probability is already <15%. A negative PERC in a moderate- or high-suspicion patient excludes nothing.
• D-dimer in high probability. A negative result does not exclude PE. Go to imaging.
• Don't wait to anticoagulate. ESC 2019 advises starting anticoagulation in patients with high or intermediate clinical probability while the workup is in progress. The 2026 AHA/ACC guideline makes this 2a when suspected PE looks like Category C2 or higher, bleeding risk is low, and imaging is delayed.
• When CTPA isn't possible (e.g., contrast contraindication, or a kidney-function concern under local contrast policy): V/Q, preferably SPECT, is reasonable over contrast MRA (2a). CTPA is otherwise preferred over V/Q (1).
• Pregnancy. Use the pregnancy-adapted YEARS pathway, not PERC or age adjustment. DOACs and VKAs are contraindicated in pregnancy; LMWH is the agent.
• Already anticoagulated. D-dimer strategies were validated in patients not on therapeutic anticoagulation.
• Not diagnostic. Echo is not recommended to confirm or refute PE (3-NB). After a negative CTPA or normal V/Q SPECT, leg duplex adds nothing (3-NB).