Conducting and reporting the Delphi method in traditional Chinese medicine syndrome diagnosis research: A cross-sectional analysis

Background: The Delphi method has been extensively used to reach a consensus in traditional Chinese medicine (TCM) syndrome diagnosis research when subjective judgment is not uniform and objective evidence is lacking. The conduct and reporting of the Delphi method in TCM syndrome diagnosis research have never been critiqued. Our study aims to explore the consistency of using this technique and assess the reporting quality. Methods: A cross-sectional study was employed to scope articles reporting the conduct of the Delphi method in TCM syndrome diagnosis research. We searched the PubMed, Web of Science, CNKI, VIP, Wanfang and SinoMed databases with the restriction of Chinese and English language from their inception to March 25, 2023. A standardized extraction form was designed to collect demographics and methodological processes reflecting the rigor and transparency in TCM syndrome diagnosis research. Results: A total of 1832 studies were screened, and 50 were included. The median number of panels was 30 (IQR 20-34.5) and only 12 (24.0 %) studies were with a heterogeneous sample of panels. Two rounds was most common (37/50; 74.0 %), followed by three (7/50; 14.0 %), and only 13 (26.0 %) studies determined the number of rounds a priori. The reporting quality varied, with 18.0 % (9/50) reporting anonymity, 30.0 % (15/50) describing the controlled feedback, 20.0 % (10/50) reporting the procedure duration (7.14 +/- 3.29 months) and 26.0 % (13/50) predefining the consensus. Conclusion: The Delphi method is inconsistently conducted and nontransparently reported in TCM syndrome diagnosis research. Standardized criteria are urgently needed for best practices in future research.