Core outcome set developers should consider and specify the level of granularity of outcome domains

Jan Kottner, Dorcas Beaton, Mike Clarke, Susanna Dodd, Jamie Kirkham, Toni Lange, Robby Nieuwlaat, Jochen Schmitt, Peter Tugwell, Paula Williamson

Research output: Contribution to journalArticlepeer-review


In current core outcome set (COS) development practice, there seems to be variation regarding the level of granularity of outcome domains within the same and between different COS. The chosen granularity of domains has consequences for definitions and interpretation of domains and instrument development. To determine what guidance can be given to (COS) developers regarding the granularity of outcome domain extraction and presentation during COS development. Review of recent COS development research and two structured online meetings by members of the Red Hat Group and other experts. All participants agreed that area or discipline specific outcome domain and possibly subdomain classifications are needed and that existing taxonomies should be considered as a starting point. Regarding granularity, participants agreed that standardized outcome extraction procedures might be useful, but that a single set of guidance is not applicable. Regardless, COS developers must be explicit and justify their approach to outcome extraction and the level of granularity must be similar within subsequent consensus procedures, including Delphi studies. Irrespective of the level of granularity, the domains must be described as clearly as possible. The level of granularity of outcome domains should be explicitly planned and explained by COS developers early in the process, and the domains should be described as clearly as possible.
Original languageEnglish
Article number111307
JournalJournal of Clinical Epidemiology
Early online date27 Feb 2024
Publication statusPublished - 01 May 2024


  • outcomes
  • clinical trials
  • core outcome set

Cite this