At least 264 Recovery Colleges have been identified in 31 countries, yet the RECOLLECT tools used to define and assess the model originated in a high-income Anglophone context. A two-round Delphi used cross-cultural de-centring: 54 panellists from 20 countries rated importance and cultural difficulty, with free text examined through corpus-informed linguistic analysis. Core principles were endorsed, but terms including co-production, recovery, professionals, and shifting the balance of power were interpreted inconsistently. After revision, 41 panellists from 15 countries reached consensus on all 11 items for understandability and cultural appropriateness.
Key findings
- The background identified 264 Recovery Colleges in 31 countries across six continents, motivating tools not anchored only in one Anglophone setting.
- Core principles were widely endorsed, but co-production, recovery, professionals, and shifting the balance of power were interpreted inconsistently and carried assumptions about equality, autonomy, and participation.
- Round 1 included 54 panellists from 20 countries; Round 2 retained 41 panellists from 15 countries, or 75.9%.
- All 11 revised items reached consensus for understandability and cultural appropriateness, producing Global RCM v1 and Global RFM v1 while retaining interoperability with the originals.
Why this matters globally
The study offers a de-centring method that goes beyond literal translation by surfacing source-context assumptions and revising them across settings while preserving comparability with original tools. The approach may transfer to other mental-health instruments, but wording consensus is not psychometric validity, implementation feasibility, or service impact. Translated versions and real-world use still require testing.
Thai researcher contribution
Peanchanan Leah of Ratchasuda Institute, Faculty of Medicine Ramathibodi Hospital, Mahidol University, is the Thai-affiliated author in the 21-member team. Thai representation contributes to de-centring beyond Western settings, but the public article does not attribute specific feedback to Thai panellists or detail each author's individual task.
Limitations to consider
The authors note that participating countries were limited to settings with Recovery Colleges or accessible expertise, so not all world regions are represented, and representation was uneven. English-language consultation may favor people familiar with Anglophone recovery discourse. Round 2 fell from 54 people in 20 countries to 41 in 15. Consensus addressed wording understandability and cultural appropriateness, not translated versions, implementation, or psychometric performance in each context. Participants were already engaged with Recovery Colleges, leaving wider stakeholders underrepresented.