Joshua B. Ewen, Claudio Babiloni, Gary S. Collins, Lauren E. Ethridge, Jean Gotman, Akio Ikeda, Philippa J. Karoly, William Z. Potter, Stephan Rampp, Margitta Seeck, Sándor Beniczky
Submitted to the IFCN Executive Committee on 15 July, 2024
The primary goal of reporting guidelines is to promote the disclosure of risks of bias when understanding the performance characteristics of a particular test. A secondary goal of clinical-test–focused reporting standards is to compel developers and authors to explicate the methodology used within a validation study in a comprehensive way that can be implemented not only in replications but also when the test is applied in the clinic. The EQUATOR Network has supported and distributed several reporting guidelines that are relevant to the evaluation of the accuracy of clinical tests, including STARD (for diagnostic test-accuracy studies) (Bossuyt et al., 2015) and TRIPOD (for multivariable prediction studies) (Moons et al., 2015). Separately, the EEG and event-related potential (ERP) communities (Ewen & Levin, 2022) have developed recommendations for the reporting of certain types of EEG studies, targeted toward specific clinical populations or EEG metrics, and often inclusive of both group-level mechanistic studies and individual-level biomarker validation studies (Keil et al., 2022; Keil et al., 2014; Picton et al., 2000; Webb et al., 2015).
The gap that the “Guidelines for Reporting EEg/Neurophysiological Biomarker Evaluation for Application to Neurology and neuropsychiatry” (GREENBEAN) guideline is intended to fill is guidance around the reporting of clinical biomarker accuracy studies—not only diagnostic, but also prognostic, real-time monitoring, etc. (FDA-NIH Biomarker Working Group, 2021)—that involve a wide range of EEG technologies and are applicable to a variety of patient groups. Moreover, these guidelines are intended to operate over the various phases of the validation journey, from early exploratory studies to studies demonstrating that the use of the biomarker improves patient outcomes (Table 2) (Ewen et al., 2019).
The current guidelines reflect the consensus of an expert panel convened by the International Federation of Clinical Neurophysiology (IFCN). While we have attempted to make the titles of the items in the checklist as clear as possible so that they “speak for themselves,” we recognize that both EEG methodologies and biomarker-validation experimental designs require a high level of technical expertise. We therefore offer this Explanation and Elaboration document.
We recognize that the range EEG collection and analysis methods, coupled with the range of clinical applications for which biomarkers are intended—so-called Contexts of Use (COU; Table 1) (FDA-NIH Biomarker Working Group, 2021)—means that minimal reporting requirements will vary from study to study.
As with any reporting guideline, the intent is not to set technical standards, though we hope that these guidelines will have the complementary effect of encouraging the field to define areas in which “best-practices research” is needed in EEG.
Please note that the citations included as examples below were lightly edited to manage abbreviations and citations within the quoted text for clarity.