This section focuses on approaches that investigators might take to support research teams and health system partners as they navigate the last part of the trial, before outcomes are known. These suggestions reflect lessons shared by NIH Collaboratory Trial investigators.

For end-of-trial decision-making before trial results are available, consider:

• Planning for this phase as early as possible. While the questions and considerations in this chapter may seem too abstract to entertain in the very beginning, establishing norms, systems, and activities (e.g., a schedule of in-person meetings) to nurture relationships with partnering systems throughout the life of the project can be important first steps. Another possibility may be for investigators to engage with these issues starting in midtrial and to revisit them as the end of the intervention approaches.

• Incorporating a broad participatory approach into trial design. This would involve not only leadership-level collaborators from participating systems, but also, if possible, frontline managers or other staff who are directly affected by or involved with the intervention. Including patients and communities in the process is also an important way of understanding and considering early the varied (and sometimes conflicting or competing) interests relevant to the full range partners.

• Conducting interviews with decision-makers as part of the trial. This method can be used to assess perceptions of intervention effects on staff satisfaction and organizational climate. The information gathered can not only provide important perspectives on the trial but also may reveal health system leaders’ plans and inclination to maintain or abandon the intervention.

• Including quantitative measures of broader outcomes. Outcomes such as staff satisfaction and leadership perception of intervention benefits and burdens can help inform end-of-trial decisions, even if those outcomes are not considered relevant by research sponsors, journal editors, or scientific reviewers (e.g., in the LIRE trial “liking” the intervention likely was related to it being maintained at some sites).