Statistical methods, outcome structures and research designs in hybrid type II effectiveness–implementation studies

Hybrid type II studies give roughly equal weight to whether an intervention works and whether the strategy used to implement it works. That dual emphasis creates statistical problems: two primary endpoints raise multiple-testing questions, most of these studies are cluster randomized, and designs such as stepped-wedge and sequential multiple-assignment trials produce outcomes that change over time, dose or strategy intensity. Methodological solutions exist, but how often they are applied, and how authors justify their choices, is unclear.

The protocol was published in F1000Research on 13 August 2026 (version 1, awaiting peer review) and registered on the Open Science Framework. Searches and screening are underway.

The review follows JBI scoping-review guidance and will be reported with PRISMA-ScR. Embase, Medline, Scopus and Web of Science are searched for hybrid type II studies and for methodological guidance on their design or analysis, with no limit on publication type. Two reviewers extract data independently; statistical methods are categorized alongside study characteristics such as the implementation strategy, and methodological guidance is summarized narratively.

• A map of the statistical methods, outcome structures and designs in use, against the methods available
• A narrative summary of existing methodological guidance for hybrid type II designs
• Priorities for methodological development, to inform the planning and reporting of future hybrid type II studies

An author group across McGill University, the CTN+ Methods Think Tank, McMaster University, the University of Utah and The Ohio State University, with the lab's medical librarian.

The TIDieR and AACTT secondary analysis is listed because it examines how implementation strategies, the other half of a hybrid design, are specified and reported.