RESUMEN
BACKGROUND: Globally, Multidisciplinary Teams (MDTs) are considered the gold standard for diagnosis and treatment of cancer and other conditions, but variability in performance has led to demand for improvement tools. MDT-FIT (Multidisciplinary Team Feedback for Improving Teamwork) is an improvement programme developed iteratively with over 100 MDTs (≥1100 MDT-members). Complex interventions are often adapted to context, but this is rarely evaluated. We conducted a prospective evaluation of the implementation of MDT-FIT across an entire integrated care system (ICS). METHODS: MDT-FIT was implemented within all breast cancer MDTs across an ICS in England (n = 10 MDTs; 275 medical, nursing, and administrative members). ICS managers coordinated the implementation across the three stages of MDT-FIT: set up; assessment (self-report by team members plus independent observational assessment); team-feedback and facilitated discussion to agree actions for improvement. Data were collected using process and systems logs, and interviews with a purposively selected range of participants. Analysis was theoretically grounded in evidence-based frameworks for implementation strategies and outcomes. RESULTS: All 10 MDTs participated in MDT-FIT; 36 interviews were conducted. Data from systems and process logs covered a 9-month period. Adaptations to MDT-FIT by the ICS (e.g., coordination of team participation by ICS rather than individual hospitals; and reducing time protected for coordination) reduced Fidelity and Adoption of MDT-FIT. However, the Acceptability, Appropriateness and Feasibility of MDT-FIT remained high due to embedding implementation strategies in the development of MDT-FIT (e.g., stakeholder engagement, interactive support). CONCLUSIONS: This is a unique and comprehensive evaluation of the multi-site implementation of a complex team improvement programme. Findings support the imperative of considering implementation strategies when designing such programmes to minimize potentially negative impacts of adaptations in "real world" settings.