Diagnostic studies of PE

(EHJ 2014;35:3033)

• CXR (limited Se & Sp): 12% nl, atelectasis, effusion, ↑ hemidiaphragm, Hampton hump (wedge-shaped density abutting pleura); Westermark sign (avascularity distal to PE)

• ECG (limited Se & Sp): sinus tachycardia, AF; signs of RV strain → RAD, P pulmonale, RBBB, SIQIIITIII & TWI V1–V4 (McGinn-White pattern; Chest 1997;111:537)

• ABG: hypoxemia, hypocapnia, respiratory alkalosis, ↑ A-a gradient (Chest 1996;109:78) 18% w/ room air PaO2 85–105 mmHg, 6% w/ nl A-a gradient (Chest 1991;100:598)

• D-dimer: high Se, poor Sp (~25%); ELISA has >99% NPV ∴ use to r/o PE if “unlikely” pretest prob (JAMA 2006;295:172); cut-off 500 if <50 y, 10× age if ≥ 50 y (JAMA 2014;311:1117)

• Echocardiography: useful for risk stratification (RV dysfxn), but not dx (Se <50%)

• V/Q scan: high Se (~98%), low Sp (~10%). Sp improves to 97% for high-prob VQ. Use if pretest prob of PE high and CT not available or contraindicated. Can also exclude PE if low pretest prob, low-prob VQ, but 4% false (JAMA 1990;263:2753).

• CT angiography (CTA; see Radiology inserts; JAMA 2015;314:74): Se ~90% & Sp ~95%; PPV & NPV >95% if imaging concordant w/ clinical suspicion, ≤80% if discordant (∴ need to consider both); ~1/4 of single & subseg may be false ⊕; CT may also provide other dx

• Lower extremity compression U/S shows DVT in ~9%, sparing CTA