Purpose: Proximal left coronary artery (LCA) lesions are associated with poor prognosis. Transthoracic Doppler echocardiography (TTDE) can be performed in a short time and combined with routine echocardiography. Further, the accelerated coronary flow velocity detected by TTDE reflects the presence of proximal LCA lesions with high specificity. Confirming its reproducibility is crucial for quality control. We aimed to retrospectively investigate machine and sonographer dependence for detecting coronary flow at the site of the proximal LCA.
Subjects and Methods:
1)Machine dependency We enrolled 325 consecutive patients who underwent two rounds of TTDE using two different echo machines (Vivid 7 and Vivid E9; GE Vingmed, Horten, Norway) by the same sonographers between 2010 and 2016.
2)Sonographer dependency We enrolled 100 consecutive patients who underwent three TTDE examinations using the same echo machine (Vivid 7) by three sonographers with different levels of TTDE experiences (sonographer 1 and 2: 5 years, sonographer 3: 2 years) between 2010 and 2016. In both studies, the definition of detecting coronary flow at the proximal LCA was the successful digital image acquisition of coronary flow via the pulse Doppler method at that site. If coronary flow was detected, the time for the detection was also analyzed. Patients with revascularization during the period of TTDE examinations were excluded.
Results: The detection rate of coronary flow using the Vivid E9 was significantly better than that of the Vivid 7 (68% vs 60%, p=0.01). However, the detection time was not different between the two machines. Also, sonographer dependency for the detection rate (sonographer 1: 77%, sonographer 2: 70%, sonographer 3: 65%, p=0.0486) and the detection time (sonographer 1: 57 sec, sonographer 2: 65 sec, sonographer 3: 86 sec, p<0.01) was observed, which might be associated with TTDE experience.
Conclusion: The detection of coronary flow at the proximal LCA may depend on the echo machine and sonographer. Additionally, a learning curve may exist for its detection.
View full abstract