The majority of reporting guidelines development lack conducting the Delphi method: A systematic search of reporting guidelines

Article type
Authors
Banno M1, Tsujimoto Y2, Kataoka Y3
1Department of Psychiatry, Seichiryo Hospital
2Department of Healthcare Epidemiology, Graduate School of Medicine and Public Health, Kyoto University
3Hospital Care Research Unit, Hyogo Prefectural Amagasaki General Medical Center
Abstract
Background: Previous guidance of reporting guidelines recommends incorporating the Delphi method, that is a method to integrate the opinions of a group of experts, for consensus in the development of reporting guidelines.

Objectives: The purpose of this study was to clarify whether reporting guidelines generally perform the Delphi method, what factors may be associated with the employment of Delphi method, and the reporting quality of Delphi method in reporting guidelines.

Methods: We included all reporting guidelines (n=244) in Enhancing the QUAlity and Transparency Of health Research (EQUATOR) Network published after January 1, 2011. We investigated the trends and factors associated with performing Delphi method, using a Cochrane-Armitage test and multivariable logistic regression, respectively. Moreover, we assessed the reporting quality of Delphi method among current reporting guidelines.

Results: Of 244, 62 (25 %) performed Delphi method for consensus development. The proportion of reporting guidelines that conducted Delphi method was less than 10 % in 2011 and 2012, and 29 % in 2019. Year of publication (adjusted odds ratio [AOR] =1.35, [95 % confidence interval (95 % CI) 1.15-1.58]), number of authors (1–19: AOR 3.35 [95 % CI 1.48 to 7.59], 20 or more: 3.96 [95 % CI 1.70 to 9.19]), compared with 1–9 and multiple and simultaneous publications (AOR 2.54 [95 % CI 1.01 to 6.37]) were associated with the use of the Delphi method. The reporting quality of the Delphi method was moderate in most reporting guidelines using Delphi method.

Conclusions: The use of the Delphi method in reporting guidelines is insufficient. Users and reviewers should carefully appraise the consensus development in the reporting guidelines.

Patient or healthcare consumer involvement: None.