“SDM:HOSP”- a generic model for hospital-based implementation of shared decision making

BACKGROUND: Shared decision making (SDM) is a core element in the meeting between patient and healthcare professionals, but has proved difficult to implement and sustain in routine clinical practice. One of five Danish regions set out to succeed and to develop a model that ensures lasting SDM based on learnings from large-scale real-world implementation initiatives that go beyond the ’barriers’ and ’facilitators’ research approach. This paper describes this process and development of a generic implementation model, SDM:HOSP. METHODS: This project was carried out in the Region of Southern Denmark with five major hospital units. Based on existing theory of SDM, SDM implementation, implementation science and improvement methodology, a process of four phases were described; development of conceptual elements, field-testing, evaluation, and development of the final implementation model. The conceptual elements developed aimed to prepare leaders, train SDM teachers, teach clinicians to perform SDM, support development of patient decision aids, and support systematic planning, execution and follow-up. Field testing was done including continuous participant evaluations and an overall evaluation after one year. RESULTS: Data from field testing and learnings from the implementation process, illustrated the need for a dynamic and easy adjustable model. The final SDM:HOSP model included four themes; i)Training of Leaders, ii) Training of Teachers and Clinicians, iii) Decision Helper, and iv) ‘Process’, each with details in three levels, 1) shared elements, 2) recommendations, and 3) local adaption. CONCLUSIONS: A feasible and acceptable model for implementation of SDM across hospitals and departments that accounts for different organizations and cultures was developed. The overall design can easily be adapted to other organizations and can be adjusted to fit the specific organization and culture. The results from the ongoing and overall evaluation suggest promising avenues for future work in further testing and research of the usability of the model.