Abstract TMP65: Robot-assisted Transcranial Doppler Versus Transthoracic Echocardiography For Right To Left Shunt Detection: “Real World” First-look

Abstract

Background: Right to left shunt (RLS), including patent foramen ovale (PFO), is a recognized risk factor for stroke. RLS/PFO diagnosis can be made by transthoracic echocardiography (TTE) or transcranial Doppler (TCD). A multicenter, prospectively enrolled device trial of robot-assisted TCD (raTCD) vs TTE for RLS detection (BUBL, NCT04604015) was completed. This demonstrated raTCD detecting all and large RLS at approximately 3 times the rate of TTE (primary outcome, any RLS: raTCD 64% vs TTE 20% [absolute difference 43.4% (95% CI 34.3%-52.5%), p < 0.001]). It is unknown if these results are generalizable to routine practice. Methods: We conducted a multi-site retrospective review of prospectively collected “real-world” clinical data of raTCD for RLS diagnosis. All patients referred for a clinical study at the participating sites who underwent raTCD were included. raTCD was performed with standard TCD bubble study technique at all sites. Outcomes included demographics, rate of RLS detection, all and large (Spencer Logarithmic Scale ≥3), as well as “no window” rate. These results were compared to the results of BUBL with descriptive statistics and 2-sample proportion tests. Results: There were 350 patients who underwent clinical raTCD across three participating sites. In this population, the mean age was 56 ± 13 years compared to 59 ± 14 years in BUBL (p = 0.03, 95% CI 0.20-5.49). Both cohorts were 46% female. The clinical population had any RLS on raTCD in 54% (189/350) compared to 64% in BUBL, which was not a significant difference (p= 0.14, 95% CI -0.02-0.18). Large RLS was detected in 26% (91/350) of the clinical population as compared to 28% in BUBL, which was also not a significant difference (p = 0.89, 95% CI -0.08-0.11). The “no window” rate in the clinical population was 5% (18/350), compared to 8% in BUBL (p = 0.39, 95% CI -0.03-0.08). Conclusions: The same raTCD device that was safe and 3 times more likely to diagnose RLS than TTE in a device trial detected all and large RLS in clinical practice at a rate similar to that demonstrated in the device trial. These results further support the notion that raTCD may allow providers to achieve the known sensitivity of TCD for RLS and PFO detection without the need for an experienced operator, even in a “real world” clinical setting.