Toward best practice methods for delirium biomarker studies: An international modified Delphi study.

Background: Delirium is a serious and distressing neurocognitive condition common in people with advanced illness. The understanding of delirium pathophysiology is limited and largely hypothetical. To accelerate empirical understanding of delirium pathophysiology, robust scientific methods for conducting and reporting delirium biomarker studies are urgently needed. The aim of this study was to develop international consensus on the core elements of high-quality delirium biomarker studies.

Methods: A three-round modified Delphi survey was conducted from February to August 2019. Participants were international researchers experienced in conducting delirium studies from a range of settings (hospital, university, research centres). Round one commenced with open-ended questions developed from results from a prior systematic review and the REMARK (REporting recommendations for tumour MARKer prognostic studies) checklist. Responses were qualitatively analysed, and closed statements were developed. Participants then ranked the importance of these statements using a 5-point Likert scale in rounds 2 and 3. A priori consensus was defined as ≥70% participant agreement. Descriptive statistics for each item were computed including the mean Likert scores, SD and median participant scores.

Results: Twenty-eight participants completed survey round one, 16 completed round two and 19 completed the final round. Consensus was achieved for a total of 60 items.

Conclusion: The Delphi survey identified items that expert researchers agreed were important in the conduct of delirium biomarker studies. These reporting items provide a strong platform for improved methodological quality and opportunities to synthesise future delirium biomarker studies.