Evaluation of AHRQ's Partnerships for Quality Program

You Are Here: AHRQ Home > About AHRQ > Mission & Budget > Evaluation of AHRQ's Partnerships for Quality Program > Chapter VI

Chapter VI. Contribution of Partnerships and Other Key Factors to Project Success and Sustainability

Partnerships are promoted to address health problems because they can often achieve what no organization can do on its own. Diverse partners, with different strengths and networks, can increase resources to address a problem, broaden the reach of interventions, and persuade others to adopt innovations. The power of partnerships comes not just from combining resources, but capitalizing on each partner's strengths, capacities, and influence with different audiences to create synergy (Lasker, et al., 2001). They can help create the tipping point that leads to widespread adoption of innovations and ideas (Gladwell, 2000).

The assumption behind the PFQ program, which we built into the evaluation framework, was that the relationships among the lead grantee organizations, key collaborators and target organizations or providers would be critical for achieving buy-in to evidence-based changes for improving health care quality, safety, and security. Strong support from each project's key collaborators and target organizations, as framed in the rationale for the PFQ program, was key to the implementation and sustainability of health care improvements.

This chapter examines the composition and structure of the partnerships created in the 20 PFQ grant projects, assesses the elements of effective partnerships, and discusses other important factors that contributed to the projects' success and sustainability. It concludes with a set of lessons for AHRQ about how to structure effective partnerships to translate research into practice on a large scale.

A. Variation in Partnership Structure and Composition

AHRQ provided relatively little guidance in the RFA on the structure of the partnerships, or who should be involved. The agency recognized that the diversity of organizations targeted to achieve improvements, and the specific types of changes proposed to translate evidence-based research into practice, required flexibility in selecting the most appropriate partners and deciding how they would work as a group.

Partnership structure and composition differed across the projects first and foremost by their grant focus, as shown in Box 3. The bioterrorism and emergency preparedness projects generally formed partnerships with target organizations that were looser and more informal than those focused on clinical quality or safety improvements. This may reflect the fact that the first set of projects sought to assess needs and develop tools, whereas the second set was more likely to seek change within the targeted organizations.

Box 3. PFQ Project Partnership Models

Partnerships with Provider Organizations & Practitioners

Direct Relationship between Leadership Team and Target Providers

American College of Physicians.
American Medical Association.
American Medical Directors Assn.
Assn of California Nurse Leaders/CalNOC.
Catholic Healthcare Partners.
Child Health Corp of America.
International Severity Info Systems.
Lehigh Valley Hospital & Health Network.
New York State Dept. of Health/RDHHAR.
Physician Micro Systems, Inc./MUSC.
Research Triangle Institute.
VNSNY (Phase I diabetes collaborative).

Intermediaries heavily involved in work with Target Providers

American Academy of Pediatrics.
American Hosp Assn/HRET.
VNSNY (Phase II acute care hospitalization reduction).

Partnerships with Health Care Purchaser, using target organizations as study participants

HealthFront.
The Leapfrog Group.

Partnerships using Target Organizations as Advisors or Study Participants

Altarum Institute.
Texas A&M Univ.
JCAHO.
CT Dept of Health/Yale New Haven Health System.

For example, the bioterrorism preparedness projects led by JCAHO and Connecticut Department of Health/Yale New Haven Health System used target organizations as participants and subjects in studies and training courses. Target organizations also provided information, data, and lessons for studies on bioterrorism preparedness, or participated in modeling exercises and case studies, conducted by Altarum Institute and Texas A&M University, and on the value of performance measurement for JCAHO's other study.

Partnerships formed around the two purchaser-led projects also reflected the role that payers play in the health system. Both HealthFront and The Leapfrog Group worked closely with local coalitions of employers, large health plans, and large companies. While their ultimate quality improvement targets were physicians and hospitals, respectively, the two project teams had little communication or collaboration with providers, other than as survey participants. When they wished to communicate with providers, the most common model was to use them in a convening role. For instance, HealthFront, Altarum Institute, and Texas A&M University organized and held seminars with target organizations to present their preliminary or final results, and discuss how the results could be used in practice.

Partnership structure differed in the 14 projects that focused on clinical quality and safety improvements, usually seeking close working relationships with target organizations. Project leadership teams worked directly with provider organizations or practitioners in the design, implementation, and assessment of the effects of interventions to translate research into quality improvements, though the strength of the relationship differed. These projects typically had three partnership components, which varied in the regularity of their communication:

The Leadership Team, consisting of PIs, co-PIs, and project directors or managers, who communicated at least weekly, and sometimes daily during certain periods, on tasks as diverse as grant management and reporting, provider training, advisory group consultations, research design, data collection and analysis, and target organization relations.
Structured Relationships between the Leadership Team and Target Organizations, through such mechanisms as annual or semi-annual training workshops, learning collaborative sessions, site visits, and conference calls with leadership team members and other intermediaries and support organizations.
Ancillary Support through Linkages between the Leadership Team and Advisors, whose support could be organized into formal advisory groups that met at the start of the project, and occasionally after that, or as an informal group, with advisors providing expertise and input into the design of the intervention as needed.

For these projects, the relationships with target audiences were critical to changing behavior. While all of these grantees partnered with the target groups, they differed in how heavily they relied on intermediary partners to support the targets. Twelve projects had direct relationships between grantee leadership teams (PIs, co-PIs, and other key collaborators) and target organizations, and used other individuals or organizations to provide training and technical assistance. These projects typically targeted fewer provider organizations, with the exception of PMSI/MUSC, which targeted over 100 primary care practices, but conducted site visits and conference calls with a smaller subset.

In the other strategy—used by AAP and AHA/HRET plus VNSNY in the second phase of its project—intermediary organizations played a stronger role in the partnership in order to: 1) increase the amount of training and support to a larger number of providers, and 2) build capacity to support and train providers independent of the lead grantee.

For example, AAP worked with more than 180 pediatric practices. To do so, it involved state AAP chapters in recruiting pediatricians, organizing training workshops, and providing on-going training and technical assistance. AHA-HRET's palliative care unit expansion strategy used partnerships with six exemplary palliative care programs, which served as learning labs for 60-70 hospital teams that made site visits and provided some post-site visit support to those teams. VNSNY described its project evolution as a switch from a "retail" strategy in its first learning collaborative project on diabetes care, where it worked directly with home health agencies, to a "wholesale" strategy in its second collaborative project, where it is working with 10 collaborating QIOs in order to reach almost 70 home health agencies to reduce acute care hospitalization among home health patients. In all three projects, a secondary but key goal was to build capacity of the intermediaries to carry on the work on their own, as part of a strategy to assure sustainability.¹⁷

Return to Contents

B. Factors Behind Successful Partnerships

Certain characteristics and processes appear to contribute to effective partnerships in PFQ projects, based on themes that emerged from interviews with project PIs and their partners. This analysis primarily concerns the 15 projects that tested quality interventions. It excludes those that used partnerships primarily to produce knowledge—the bioterrorism preparedness projects and the quality improvement study projects led by RTI and JCAHO.

1. Position of Lead Organizations and Intermediaries

AHRQ expected lead organizations to be well-situated and capable of influencing directly the target organizations that were the focus of quality improvement efforts. Most grantee agencies, or others in the leadership team, were well-positioned to influence target organizations by virtue of being national or state associations representing the target organizations. According to one PI, "Having the credibility of the [national association] behind our work was helpful." In two cases, grantees were health systems that owned or were affiliated with the target providers (CHP and Lehigh Valley). VNSNY is a recognized leader in the home health field, giving it credibility among its peers. One of the home health agency staff in its project said, "Because of the size of [VNSNY] and the work they've done, agencies are very proud that it's one of our agencies that really spearheaded this... there's a sense of credibility to that."

In one instance, the lead organization had existing regulatory relationships with the organizations targeted for project interventions. This held certain advantages. It made it easier to recruit target organizations because they felt that they could not refuse. "When [they] ask something of us, it's not a good idea to say no," said one participating organization. It also gave the lead organization a chance to turn their historically adversarial relationship with providers to a more helpful one. The downside is that regulators still wield power over the target organizations, so the latter felt obligated to take on more than they could handle. Had the target organizations felt comfortable enough expressing this to the project leadership team early on, the project design could have been modified to improve the intervention's success and sustainability.

2. Experience and Skill in Managing Partnerships

Despite having strong potential for influencing target organizations, not all grantee agencies or leadership team partners had experience or skills in managing partnerships. Several national association PIs admitted that this was their first attempt to create working relationships on quality improvement activities with members, and considered it a great success just to show they could implement the partnership. But implementation is not the same as effective management, and some were better than others at building consensus, defining structures and processes for work to progress, developing leadership and joint ownership of the project, resolving conflicts, and finding ways to maximize each partner's strengths and contributions.

Partnership management takes time. The projects with more partners, more partnership groups, and more intense levels of collaboration with providers or target organizations had to spend more time managing the interactions and communication among all the partners. Sometimes, there was not enough time to do all the partnership management that some believed necessary to make the project work better. According to one PI, "I might have tried to do more one-on-one with everyone in the group [to gain consensus and work through problems] to supplement the monthly calls." Another project ran into similar problems in creating a partnership at the national level. According to one PI, "National partnerships need a lot of care and feeding, constant reminders and tasks. You need to keep up the momentum, [and] I think this project probably caught on to that a little late." One project limited the demands of partnership management by delegating responsibility and money to partnerships at the local level. The grantee organization communicated with local pilot projects to get progress reports, and assess their need for technical assistance; but the pilots rather than the national organization assumed most of the partnership management function. These experiences suggest the need for projects involving partnerships to build in adequate time for partnership management, and to consider the costs and benefits of creating partnerships at different levels.

3. Partners' Prior History in Working Together

Some projects had the advantage of starting with an existing partnership to which they could add new quality improvement targets or approaches. Projects led by The Leapfrog Group, the American Academy of Pediatrics, Lehigh Valley, CHCA, California Nurse Outcomes Coalition, and Catholic Healthcare Partners had distinct advantages in this regard. Their intended target organizations or intermediary partners were already organizational members or affiliated providers, making both the task of recruiting them easier and minimizing the need to start from scratch in defining common goals. According to one respondent, the project leadership team "has been together for so long. We are all equal in the design process, and having an effective team that has been together for so long has been invaluable."

The 14 CHCA members who had worked together under the "Child Health Accountability Initiative" banner had some experience and success in joint quality improvement projects before they began the PFQ project, and therefore had a head start in working together. Based on their early successes, the rest of the CHCA members wanted to join their efforts. But integrating into the project was challenging. Even though the new partners were already members of CHCA, they had not previously been exposed to the QI concepts and approaches or data collection requirements of the project. Getting them up-to-speed on the core partners' values and mode of operation took almost a year, slowing down the project's momentum. However, the PI believes that in the long-run, the time invested to integrate these organizations into their quality improvement efforts will have a large pay-off in expanding the number of children's hospitals involved in more rigorous and measurable QI activities.

Other projects began with little or no history of partnerships between the lead agencies and the target organizations, so they had to spend time building trust and a common vision to be successful. For example, the AMDA Foundation had prior relationships with the medical directors of nursing homes, but not with the staff most responsible for quality improvement in these facilities—directors of nursing. AMDA Foundation staff therefore had to build relationships with these individuals. VNSNY and ISIS also had to quickly establish partnerships with provider groups; they did so by holding semi-annual meetings and regular conference calls, which rapidly created group cohesion and facilitated an open exchange of ideas and lessons.

4. Involvement of Target Organization Administrators and Staff in Decision Making

According to emerging health care organization theory on partnerships, partners' roles in decision making and partnership governance are critical factors in partnership effectiveness. (Mitchell and Shortell, 2000; Shortell, et al., 2002, Bolda, et al. 2006). The experience of the PFQ projects provides some support for this theory. Partnerships that involved partners in making collective decisions on the project's intervention were more successful in gaining buy-in and long-term commitment to the intervention. Partnerships that used partners to advise and legitimize the efforts of the lead organization seemed to have less success in gaining target organizations' commitment to adopt or sustain the intervention.

Involving administrators from participating organizations is critical, according to some of the PFQ project partners. "You've got to have administrative buy-in to support this," according to one PI. Even in a large health system such as Catholic Health Partners, there are limits to the "command and control" approach. "The HF advocates that have been very successful have had complete buy-in from [their managers]... this just shows that if you are starting something like this, you have to have commitment from administration." While involving target organizations in project decision making may take more time to achieve consensus on goals, strategies, or tactics, it may create stronger buy-in in the end and appears to result in greater commitment of resources and long-term organizational change.

Some of the most successful projects involved people at all levels of the target organizations deciding how to adapt the intervention to their organizations, which helped produce tangible improvements and fostered better teamwork. ISIS and VNSNY, for example, not only involved administrators and nursing directors, but also nursing assistants and home health aides. ACP invited teams of physicians, nurses and office managers to their practice-based, team-oriented training programs on diabetes care improvement. According to one of the partners, "What's remarkable is that, in terms of process, the office administrators are saying [the ACP training] is helping them feel like they're more part of the care process, and now they understand how they can fundamentally improve care. This has opened up dialogue between physicians and staff in how they can improve quality and makes the practice feel like they have social value."

Meetings among staff from the participating organizations to share and learn from each other were also important factors in success. According to one respondent, "The interactions we had with other facilities [in the study] were great. Our meetings with [them] helped us to develop best practices." In another project that had prior relationships but had not met in person before the PFQ project, one respondent said, "My partners' involvement contributed to the project's progress. The ability to meet with the partners through in-person interactions in a concentrated, focused way has led to interesting work, and I've learned a lot." Another interviewee claimed that, "Creating a learning network has helped us move forward. Everyone having the opportunity to say, 'here's what I learned this week, here's what's working and here's what's not working,' that's an enabler."

5. Partners who Can Promote Sustainability and Broader Diffusion

In several PFQ projects, partners changed over the life of the project, depending on their strengths and connections. Some partners are better suited to test an approach, while others are needed to take an intervention to scale. The Leapfrog Group, for example, selected a small group of regional purchasers from its membership to test different approaches to quality incentive programs in the six pilot projects. But for broader diffusion, Leapfrog is working with a larger number of its employer coalition members for its "regional roll-out" initiative. Similarly, VNSNY worked with a small group of eight agencies willing to test the use of the IHI rapid cycle quality improvement learning collaboratives in the home health setting. But for its wider diffusion efforts, VNSNY (and ISIS in a follow-on project) are involving quality improvement organizations (QIOs) in different parts of the country to take their approaches to scale. To the extent that VNSNY can build capacity in QIOs to carry on rapid-cycle quality improvement in the home health care setting, it will expand this approach to a larger group of home health agencies than it could in the first phase of the project.

^17. ISIS is pursuing a similar strategy in its work with six QIOs to replicate the "real-time," computerized care process documentation system in 30 more nursing homes, using digital pens or facility IT systems. This work is supported by a separate AHRQ Health Information Technology grant.

Return to Contents
Proceed to Next Section