I'm not going to reinvent the wheel -- see some of the fantastic analyses of PESIT (in no particular order) at:
St. Emlyns - Simon Carley
EM Lit of Note - Ryan Radecki
EMNerd at EMCrit - Rory Spiegel
One common thread is that the patients who had PEs seemed to be patients who we would think had PEs, rather than some occult finding we need to hunt for in all of our syncope patients.
Just look at Table 2, emphasis mine, which looks a lot like their Table 1, which is (gasp!) the Wells Score:
Sure, prolonged immobility and recent trauma/surgery don't reach frequentist significance, but they're close, and there just aren't a lot of people in either of those groups.
Literally the only non-Wells factors they find are tachypnea and hypotension.
You cannot make this up:
no surprise: rate of PE in US syncope patients <1% in JAMA IM