AUTHOR=Zhao Rui , Lanéelle Damien , Gao Meiying , Fu Yuwei , Tong Yisha , Scissons Robert , Wen Chaoyang , Mahé Guillaume TITLE=Inter-rater Reliability of 4-Item Arterial Doppler Waveform Classification System for Description of Arterial Doppler Waveforms JOURNAL=Frontiers in Cardiovascular Medicine VOLUME=7 YEAR=2020 URL=https://www.frontiersin.org/journals/cardiovascular-medicine/articles/10.3389/fcvm.2020.584274 DOI=10.3389/fcvm.2020.584274 ISSN=2297-055X ABSTRACT=

Background: Non-invasive Doppler waveform (DW) analysis is a widely adopted method for detecting and evaluating lower extremity peripheral artery disease (PAD). Previous investigations have reported that broad heterogeneity in the description of Doppler waveforms is reduced by using a classification method. The reliability of arterial Doppler classification, however, is unknown.

Purpose: The purpose of this study is to assess the reliability of a 4-category arterial DW classification method among Chinese sonographers.

Methods: During a national ultrasound conference in China attendees were invited to classify thirty arterial Doppler waveforms. After viewing a 4-category (triphasic, biphasic, monophasic, and other) arterial Doppler waveform descriptor presentation, attendees were asked to classify 15 continuous wave (CW) and 15 pulsed wave (PW) Doppler waveforms. Their responses were recorded via mobile phone and the reliability of this 4-category classification was estimated by Fleiss' Kappa inter-rater statistical analysis.

Results: One hundred and seventy-eight attendees participated in the analysis. The Kappa coefficient of Fleiss (κ) for all attendees was 0.522 (p < 0.005) with 95% confidence interval (CI): 0.520–0.523. The reliability of the waveform descriptor triphasic was the highest (κ = 0.621, p < 0.005), and other was the lowest (κ = 0.341, p < 0.005).

Conclusion: The inter-rater reliability of a 4-category arterial Doppler waveform classification by Chinese sonographers is considered weak (κ = 0.522, CI95%: 0.520–0.523, p < 0.005). This study reinforces the importance of assessing DW classification reliability and the development of DW descriptors that are more accurately predictive of clinical hemodynamic events.