Establishing an AHRQ Learning Collaborative, A White Paper
Existing Learning Collaboratives
Largely owing to the complexity of knowledge that exists, collaboration has become the preferred mechanism within and across a number of industries and professions.3 Appendix C details several collaboratives and networks that currently exist across North America, Europe, and Asia with an emphasis on efforts that focus on the fields of health care and health services research. The collaboratives presented in Appendix C appear to have the following characteristics in common:
- There is an explicit mission or charter and, in some cases, clearly articulated objectives and/or goals driving the efforts of the collaborative.
- Collaboratives contain a set of routine activities that are customizable depending on the needs of constituents. Examples of routine activities that are customizable include: continuous learning work groups, peer learning groups, sub/interest groups that convene during larger meetings, "action learning" that meets constituents where they are, and flexible work plans that allow for innovation.
- Relationship-building is emphasized through ongoing communication, which occurs through a variety of mechanisms, including in-person meetings, conferences, and a variety of online resources such as Web sites, list-servs, social networking tools, and other means of communication and sharing of resources.
Several aspects of the existing initiatives are highlighted in Appendix C, including: background and history, mission and goals, tools and mechanisms used, lessons learned, and measures of success. The selected collaboratives were based on Internet searches related to learning collaboratives and networks, input from our partners at AHRQ, recommendations from interviewees, and our own experience and knowledge with respect to learning collaboratives. In our research and conversations, it was suggested that we consider not only the terms "learning collaborative" and "collaborative" but also that we explore the term "network." It seems that in some cases, these terms have clear distinctions, while in others they are used more interchangeably. The information provided in Appendix C is based on available resources, and in some cases, is supplemented by input from leaders of the particular initiative gathered through informal email or telephone conversations. We were not always able to get in contact with leaders of the collaborative, which resulted in some gaps with respect to lessons learned and measures of success.
These collaboratives and networks offer insight into what kind of form an AHRQ Learning Collaborative could take. For example, many of the collaboratives were developed to focus on a specific issue or problem—a key feature of the Institute for Healthcare Improvement's Breakthrough Series model25—and to enhance learning across distinct entities with common practices and interests. For example, the Medicaid Medical Director's Learning Network is a forum for State Medicaid directors to share ideas and best practices with respect to improving access, quality, and costs in Medicaid. While each State program is different, they share some common goals, and they have focused on solving specific common problems, such as reducing C-section rates and improving preventive screenings. Similarly, the National Association of State Medicaid Directors (NASMD) Multi-State Collaborative was formed to support the implementation of Medicaid Transformation Grants with a focus on electronic health records (EHR) and Health Information Exchange (HIE). Again, the goal here is to share best practices across States and focus on targeted areas of interest to all participants, such as implementing meaningful use requirements in Medicaid. Intermountain Health Care emphasizes the importance of these best practices coming not only from the literature, but also from the experiences of practitioners.
Together, these examples offer a model for the AHRQ Learning Collaborative. To the extent that specific issues or areas of interest could be identified that are common to participating T32 training programs, these could be used as a focal point for collaboration. While recognizing that each program is unique, with different strengths, identifying common areas for improvement could provide a useful roadmap for sharing best practices by drawing on both the research literature and the experiences of individual programs and fostering learning across the training programs.
There are also networks that are explicitly focused on advancing research in a certain area. For example, Knowledge Translation Canada is a collaborative to "identify and study solutions to ensure that key stakeholders in the Canadian health care system, have the opportunities, tools and skills necessary to achieve health for Canadians."26 Specifically, they focus on translating research into practice and making the link between health services research and the delivery of health care. Similarly, collaboratives in the pharmaceutical field focus on advancing drug development. Specifically, the NewMeds Collaboration between the pharmaceutical industry and academic institutions focuses on identifying new methods for developing drugs to treat schizophrenia and depression.27 Relatedly, the Centre of Excellence for External Drug Discovery (CEEDD) brings together drug developers to form alliances through the discovery and development process of a drug in order to enhance value for patients.28 In a very specific case, the Pediatric Acute Lung Injury and Sepsis Investigators (PALISI) Network brings researchers from pediatric intensive care together to provide feedback and identify best strategies for serving this specific population through sharing and providing feedback on relevant research.29
In the case of the AHRQ Learning Collaborative, this approach could be a way for trainees to share their research with one another across institutions, perhaps developing smaller working groups that would focus in specific areas of health services. Such a model could not only include presentations of research by trainees, but also webinars presented by experts in the field (e.g., program directors or faculty from participating institutions), which could help foster mentoring relationships across institutions and build networks within specific issue areas of health services research.
The models for communication offer some insight into the potential structure for building relationships across partners in a collaborative. First and foremost, Mindtree, Inc. stresses that "communities of practice" should have face-to-face interactions as a venue to solve problems, share best practices, and brainstorm.14,21 However, many of these collaboratives, recognizing the limitations of geography, acknowledge the necessity of virtual forms of communication. For example, the Patient-Centered Medical Home/Meaningful Use Collaborative combines in-person meetings with virtual forms of communication.30 Regular communication takes a number of forms, but importantly begins with an in-person meeting (four in total) between which three "activity periods" occur. Activity periods are punctuated by ongoing webinars, webinettes, one-on-one coaching, conferences, and e-sharing of resources. All of the collaboratives listed in Appendix C used some blend of these components. Finding the balance between in-person meetings and virtual forms of sharing and communication will be a critical component of the AHRQ Learning Collaborative.
Another important component of effective collaboration is developing and communicating a structure for the activities of the collaborative, specifically a work plan that clearly communicates responsibilities, deadlines, and deliverables. For example, the NewMeds Collaboration developed 10 "workpackages," which are focused around a particular sub-component of the collaborative's activities with clear objectives and two leaders—an academic leader and a deputy.27 Each workpackage fits into the project structure as a whole. Similarly, Knowledge Translation Canada features four thematic research projects that individually examine the methodological and conceptual issues unique to the research project and then connect these issues to the objectives of the collaborative.26 Knowledge Translation Canada also features working groups, which appears to be a common, though less formal, structure characterizing collaboratives, including the NASMD (continuous learning workgroups), AHRQ Primary Care Practice-Based Research Networks (PBRNs; peer learning groups), and the PALISI Network (subgroups).
To provide structure to the activities of the AHRQ Learning Collaborative, it is important to establish:
- Clear objectives and deliverables:
- What specifically will be accomplished and what evidence will be provided?
- By whom?
- By when?
Once the formal project structure is determined, it is likely that more informal and flexible arrangements will emerge to guide the process itself.
The Institute for Healthcare Improvement (IHI) is a leader in helping organizations establish collaboratives. One of their experts notes, "To make a collaborative successful, you need a well-structured problem, a group of interested people, and a process that engages people" (Andrea Kabcenell, Vice President, IHI, personal communication, March 30, 2011). Through communication with IHI and other individuals with expertise in establishing collaborations and consortia at a number of organizations, the following best practices emerged:
- The more well-defined a problem is, the more amenable it may be to the collaborative approach. The collaborative approach may not be suitable for problems with many moving parts.
- Groups should be smaller and include a mix of strangers and people who know each other. To overcome communication barriers, at least one (maybe two) face-to-face communication session is necessary and should be supported by virtual communication (e.g., webinars, webinettes, and other interactive tools).
- The benefits of collaborating must outweigh the sacrifices. The efforts of the collaborative must be clearly linked to mutual interests and eventual benefits for every participant.
- For each collaborative, a team charter should be established, followed by a work plan with a clear set of objectives to be accomplished.
- Scheduling meetings, rotating locations, and planning for followup communication are required—the work of the collaborative will not just happen on its own.
- Continuous evaluation of the collaborative, particularly at the outset, can foster success and better align the tasks of the collaborative with the needs of the participants.
- Effective leadership is essential, to engage the group, schedule meetings, and cultivate a climate of trust. More importantly, rotating leaders will deepen the commitments.
- Learning collaboratives succeed when there is support from "top executives" and an expectation that there will be productive collaboration among people across programs. In this case, AHRQ could play an important role in investing, building, and expecting social relationships across programs.