Background: The presence of a large physiologic shunt, defined as >20 left atrial microbubbles within 3 cardiac cycles on transesophageal echocardiography (TEE), is a randomized trial-validated indication for patent foramen ovale (PFO) closure in patients with otherwise cryptogenic ischemic stroke. The frequency with which this information is available to treating physicians from clinical TEE reports has not been well-delineated.
Methods: Among consecutive ischemic stroke patients, clinical TEE report shunt size characterizations were abstracted and compared to transcranial Doppler (TCD) formal shunt grades in the same patients and to central Core Lab TEE quantified assessments.
Results: Among 77 patients, median age was 64 (IQR 56-73), and 33 (43%) female. On TEE, shunt presence was assessed by bubble study in 60 (78%), direct Doppler alone in 5 (7%), and neither in 12 (16%). Among bubble study patient, a right-to-left shunt (RLS) potentially due to PFO was present in 25 (42%). RLS severity was quantified on the clinical report in 4 (16%) patients and only with informal descriptive terms in 21 (84%) - "small/mild/trace" (13 cases), "moderate/medium" (6), and "large" (1). In the 19 patients also undergoing TCD, RLS severity was quantified in all clinical reports. Shunt severity agreement between clinical TEE reports and TCD quantification was 100% (3/3) for formally quantified TEE shunts but poor (3/15, 20%) for the15 TEE reports using informal descriptions. For presence of a large shunt, an indication for PFO closure, clinical TEE with informal descriptions and TCD were incongruent in 5/15 (33%) of patients.
Conclusions: Quantified, evidence-based ratings of PFO shunt severity were present in less than 1 of every 6 TEE reports, and unquantified, informal size estimates correlated poorly with TCD quantification of shunt severity. Patient management would be aided by inclusion of formal PFO shunt size quantification in all clinical stroke patient TEE reports.