Chapter VII. Perspectives on the Collaborative's Infrastructure and Process (continued)
Evaluation of a Learning Collaborative's Process and Effectiveness to
D. The Contribution of Sponsors/Support Organizations
1. Being Part of a Sponsored Collaborative
The fact that the Collaborative was sponsored by an important federal agency (the Agency for Healthcare Research and Quality) and a major health foundation (the Robert Wood Johnson Foundation) enhanced the credibility of the Collaborative in the eyes of participating firms. One senior executive said, for example:
"Having the Collaborative out there and especially sponsored by the federal government sends a signal: we want you to measure and attempt to lower disparities. It is an important signal... without it, public perception will be that firms are doing it [collecting racial and ethnic data] for reasons of economic or racial profiling, not quality improvement."
"A group like this is prestigious—it's got important players around the table and a national scope that has elevated, for anyone who may have doubted, the importance of the issue (of disparities) beyond the altruism that comes from the philosophy of a single organization."
As a third said, "The sponsors [AHRQ and RWJF] are certainly recognized by our board...the involvement of (such) substantive organizations gives [the Collaborative] 'cachet.'" While the existence of more than one sponsoring organization has the potential for conflict, most participants generally perceived that the two sponsoring organizations worked well together and with the support organizations. However, the visibility of the sponsors with participating firms varied across Collaborative participants. For all firms, sponsors were most appreciated for their support and the cachet they brought, as well as for their flexibility. Firms, realistically, viewed the support organizations as responsible for handling the day-to-day work of supporting the Collaborative. Several firms mentioned that the Collaborative involved well-respected support organizations (paid for by the sponsors) helped keep the effort focused.
2. Firm Perspectives on Support Organizations and Overall Support for the Collaborative
Although support organizations made great efforts to coordinate with one another, participants made distinctions among them. Consistent with the way the Collaborative was structured, firms viewed CHCS and RAND as the two main support organizations. They saw CHCS as the facilitator of the process and "glue" that held the Collaborative together. RAND, in contrast, was respected for the substantive knowledge it brought and particularly for the help in supporting the firm's work on geocoding/surname analysis. Both the network analysis and our interviews show strong support among firms for the contribution of each organization to the Collaborative. In contrast to the two main support organizations, IHI staff working with CHCS staff were less visible in the Collaborative, which is consistent with their more limited role. GMMB was well regarded for its support to communications goals, as discussed in the previous chapter.
Positive Overall Assessment. Firms were positive about the contribution the support organizations made to the success of the Collaborative. They were especially appreciative of sponsors' willingness to make resources available to support their needs. As mentioned in the interim report, some firms viewed the large number of organizations sponsoring and supporting the Collaborative—including MPR as the evaluator, a function they did not necessarily distinguish from the others—as imbalanced, since in some meetings it meant there felt there were "more of them than us." Yet, the firms were very positive on the contributions made by staff from support organizations during meetings. While we heard that there was a "complicated leadership structure with too many constituents that don't all seem to be aligned," firms provided little in the way of concrete examples of problems generated as a result. The main exception appears to be through effects on the diverse demands made of them by different organizations, each seeking information for their own purposes in ways that were not always coordinated. In addition, while less visible to firms, the leadership structure also meant that efforts at coordination consumed a significant share of the resources available to support the Collaborative, including regular calls of an "operational committee" including all partners and sponsors.
Strengthening the Evidence Focus. Firms were very positive about the contribution of the Collaborative and the support organizations. The most universally perceived gap in what firms received from the Collaborative was in the area of information addressing firms' interest in interventions that might reduce disparities. While they may not have been interested in pursuing particular interventions and may not always have felt this was what they sought from the Collaborative, they were hoping for evidence, insight, and support on reducing disparities. They expressed desire for more information on "evidence-based practice," "existing tools," "interventions that work" and "best practice sharing." Perhaps because firms received more substantive help with geocoding/surname analysis, they were more interested in understanding what the research showed about ways they could effectively intervene. However, learning more about how to directly collect race/ethnicity data was also noted as a desire.
One senior executive articulated the desire most broadly by saying, "What I'm wrestling with at the end of the day is, what's the big lever to close the gap? What will move the needle [in terms of disparities]. Right now, I'm flying blind and just selecting interventions and approaches on face validity."
One national firm participant noted that implementing interventions based on face validity was appropriate, and cautioned that the search for evidence could lead to a "research bottleneck" delaying action. Yet, most sought more information about disparities, and were concerned that both too little was known about effective interventions and that the Collaborative did not focus enough on sharing what was known. As noted above, this was the main area of disappointment expressed by firms.
Reporting Requirements. The most contentious issue involved the structure the Collaborative sought to impose and the requirements it made of firms. As we discussed in the interim report after the second round of interviews, a number of firms objected to the demands they felt the Collaborative made on them to report progress or move in a specific direction. The core principles in the early Memorandum of Understanding—such as common measures and sharing—were more meaningful to AHRQ staff than to firms, none of which mentioned any such explicit overall agreement.
Reporting requirements and other demands for information were a particular concern among firms. Firms perceived the cumulative number of requests from sponsor-affiliated groups to be burdensome. They did not want to respond to periodic requests from the support organizations for formatted feedback on their work or to provide common measures. They were confused when a request to respond to a survey about common measures (from CHCS) was closely followed by a request from MPR for feedback on the relationships among participants and to set up interviews. Responding to requests was particularly challenging because it often required coordination across many staff and organizational entities within the firm. Generating a response not only required many people, but also the coordination time of the lead contact who had many other responsibilities.
At the final meeting in Chicago in September 2006, firms' rejection of externally imposed reporting requirements was explicit—they said they wanted to define any measures of progress that would be used in Phase II, rather than have sponsor or support organizations define measures for them. They also expressed considerable unease about ways in which efforts could be monitored. Although these views are clear, one firm that strongly opposed such requirements admitted in our final interviews that having timelines did push it to respond, and others noted that external demands helped to generate action both from participants ("Type A people who want to be Number One") and from organizations (since the external demands of the Collaborative helped them garner resources). However, these qualifications applied to concrete activities, not to specific reporting requirements or documentation that the Collaborative might require.
Model for Collaboration. As discussed previously, firms did not like the expectation that they would all pursue similar strategies. One interviewee observed that the Collaborative was "trying to fit everything into one shoebox and that's just not the way to proceed," as firms move at different paces and use different approaches to reducing disparities. Another suggested that AHRQ and other organizations perhaps had "become a little omnipotent in believing that without them the firms wouldn't be doing these things [focusing on disparities]." In our final round of interviews, one participant suggested that perhaps the sponsors of the Collaborative had inadvertently established negative dynamics through their interest in using an IHI-like collaborative model developed around providers rather than plans. Another said that rather than meeting firms where they were, the model imposed its own structure and expectations. Commercial firms were also slightly distrustful of the Medicaid focus of CHCS's prior work. Even at the end of the Collaborative, a few still referred to the "Santa Monica meeting" (that is the second meeting of the Collaborative in March 2005) where such a structure was rejected, using that as one explanation for why the request for common measures got so little attention from many of them.
Logistics. Firms participating in the Collaborative were mostly positive about the value of meetings though they perceived some room for improvement in the agendas to help them be able to share and learn more, with more substantive focus on issues. Some expressed concern over scheduling, which they felt provided too little advance notice to allow them to lock in the dates and times and also were timed with too much emphasis on sponsor/support organization schedules than those of the firm participants. The lack of active use of the internal Web site was mentioned One participant suggested that using this site more actively between meetings to post substantive content could both help add to that dimension of the Collaborative and maintain momentum in between meetings.
E. Firm Perceptions of Resources Involved in Collaboration
"Worthiness" of Participation. All eight of the firms we interviewed in the third round of the Collaborative responded positively to our query about whether their participation in the Collaborative had been worthwhile relative to the amount of time and resources it required. The value, as they elaborated, was viewed as:
- Being able to leverage the Collaborative to attract internal resources (in other words, moving disparities up on the priority list).
- Keeping momentum going and identifying pockets of opportunity.
- Helping meet contacts from other plans and obtaining internal attention.
- Identifying a direction and generating enthusiasm because others were also engaged making it a "movement."
- Using the Collaborative to highlight their ongoing work and help gain internal access to information.
- Bringing plans together and contributing to a team effort.
- Raising awareness of the issue, focusing on firm strategy, and fueling internal efforts.
- Serving as the impetus to start programs within their firms.
These perspectives paralleled in many ways the concerns firms cited as original reasons for participating in the Collaborative.
Firm Resources for Participation. As firms requested during the June 2006 meeting, we asked in the third round of interviews whether they had quantified the resources required for firm participation in the Collaborative. No one had done so comprehensively (and most not at all). One firm explained that it only looks at the issue of resource requirements when doing activities that take on a certain threshold of staff burden, but the Collaborative had so far been manageable, and requests had not yet reached that threshold. The firm explained that the real burden comes when it has to construct measures, hire staff, and so forth. The distinction between the costs of participation and of pursuing work related to addressing disparities appears critical for all firms.
Large-scale interventions and data system redesign are what appear to create large demands on resources for these firms. Firms pursuing these efforts typically reported that they were doing so as part of the costs of doing business, not because of the Collaborative. Put another way, firms would not make large investments solely because of their participation in the Collaborative—the investments had to make sense in terms of strategic and business objectives.
This is not to say that firms found it easy to secure the resources needed for the Collaborative. To undertake geocoding, firms needed resources to create records and—because most wanted to maintain control—skilled staff to analyze the resulting data. Although firms acknowledged that RAND support and economies of scale inherent in the Collaborative helped offset costs, generating sufficient financial and human resources was challenging for some firms and typically involved a significant commitment. One large firm said it had a consulting data analyst working 60 percent time over two years (at a cost of around $50,000) to support the geocoding work, in addition to the costs involved in developing the data and providing staff oversight. We were not clear whether these were resources were new or reallocated from related or different purposes. Another firm, smaller in scale and able to build on existing capacity, said that the geocoding/surname analysis work took about 20 to 40 hours annually (to prepare files for RAND and later analyze the geocoded and surname analyzed data). One firm supported geocoding by internally funding another part of their organization to do it. Another firm devoted a full-time staff member to this effort, although responsibilities also included other analytical support relevant to disparities. Because the firm's programming resources were limited, it had to purchase from a vendor some services that other firms could access internally.
Interventions undertaken specifically for the Collaborative also generated costs. Internal quality improvement initiatives, for example, required efforts to sample, prepare letters, construct measures, and so forth. Although the overall costs for such activities may be small in relation to the scale of these firms, accessing such resources often required negotiating for time and priority among scarce resources in a highly competitive environment. Garnering such resources without benefit of the Collaborative would probably have been more difficult.
The other main cost associated with the Collaborative involved the time of senior firm executives. One large national firm estimated that their lead contact, a senior executive, spent five percent of time on business related to the Collaborative. Another noted that it took time to brief staff and prepare for the weekly meetings. A third firm said that the Collaborative required the time of the executive lead contact plus the briefings of senior management. By and large, firms seemed to regard these as good expenditures, particularly when for meetings that might serve other purposes, such as professional staff development or learning. On the other hand, the fact that executives were "doing it off the sides of our desks," as one lead contact from a national firm observed, made staff particularly sensitive to spending any time that was regarded as "busywork" that might satisfy the requirements of support contractors but not contribute to efforts regarded as valuable to the firm.
F. Issues for Future Consideration
The experience of the Collaborative illuminates important issues to consider in designing future initiatives with large firms sponsoring health plans. (Lessons for Phase II of the Collaborative are discussed separately in the next chapter.)
1. Who Participates?
The health care industry is concentrated in a fairly small number of firms that play a major role sponsoring health plans nationally or regionally. These firms are diverse in structure (centralized versus decentralized), investment in quality improvement, linkages with provider systems based on ownership or history, geographic coverage, and other dimensions. Although the Collaborative sought to build on large national firms, the distinctions among these national firms did not appear well-recognized by the sponsors and support organizations at the beginning of the Collaborative, and it appears that decisions on who else would participate were strategic to individual support organizations that issued the invitations—particularly, RAND. As we noted in the interim report, sponsors said they were attracted to national plans because they touched many lives, but sponsors also appeared unprepared for the implications of such scale. Particularly absent in the considerations was a framework for the bridge between what large national firms provide (large-scale efforts and the potential to shift incentives across complex structures that ultimately link to providers) and what clinicians do (deliver care to individual patients in specific locations.)
Given the small number of major firms and their diversity, there is likely no optimum set of participants for any health plan collaborative. We cannot say that the firms in the Collaborative could have been better selected. However, it could be useful to make decisions on participation more strategically in light of the objectives of the Collaborative. For example, if the goal is to influence national policy or change policies across the industry, participants should include the largest and most influential firms. On the other hand, if the goal is to intervene in care delivery, one must recognize the way health plans relate to providers, the diversity in the industry, and the distinctions between centralized and decentralized firms so that the right firms and people therein are included. Established distinctions within the industry, particularly between commercial and Medicaid dominant firms, warrant consideration to encourage a mix of participants who can learn from one another. It also would be valuable to consider whether to include firms that are heterogeneous or homogeneous with respect to their level of sophistication, work on disparities, and quality improvement.
2. What Model for Collaboration?
There are a variety of ways to structure a group collaboration. Ultimately, the form chosen should support the overall goals. In this case, neither the goals nor the structure seemed to have been given sufficient consideration at the outset. The decision to have RWJF sponsor CHCS to complement RAND's work for AHRQ with the Collaborative was a significant one that probably had more influence over the Collaborative than has been recognized. Specifically, through this decision, and the selection of support organizations (both CHCS and its subcontractor IHI), the Collaborative was structured according to a model of traditional quality improvement work with smaller, less complex organizations (typically providers or small health plans with strong links to provider groups). There were to be three meetings, with intervening work to measure, share data, intervene on a specific condition and measure success at the patient level. Firms would work internally and on the same things. When participating firms urged more flexibility of goals, this structure was modified but maintained.
Other structures may be more appropriate, depending on the goals. For example, if the goal is to inspire firms to make addressing disparities a greater priority and to leverage firm scale to remove environmental barriers to doing so, a workgroup model using politically savvy expert facilitators with deep knowledge of how firms work—for example, a former CEO who is respected by firms with a good grasp of public policy concerns—and support of consulting content experts might be appropriate. This certainly is a model used by many industry groups. Another option is the "Learning Community" or "Learning Laboratory" mentioned by some participants where the expectation probably is less achieving certain documented accomplishments but sharing and expanding knowledge generally on ways to address disparities.
3. How to Effectively Leverage the Private Sector?
Working with large private sector organizations that function in highly competitive markets is different from working with grantees beholden to the sponsor and financially motivated to cooperate. Sponsors seeking to engage large private sector organizations in group efforts should understand and respond to the reasons—business, political, personal, and other—that drive a firm to participate, the constraints that are likely to limit their response, and the processes required to link the external work within the Collaborative to the firm's infrastructure and decisionmaking process.
Goals and requirements need to account for each of these factors and also for the diversity in ways firms are likely to handle the same function. For example, the experience of the Collaborative reinforces the importance of appreciating bureaucratic processes that apply to making decisions for the firm or approving information for external communication, collecting information that may be the responsibility of more than one division within the firm, and taking action that may require commitments of multiple firm units. Similarly, the Collaborative experience highlights the importance of anticipating firm concerns related to the market—for example, anti-trust issues or what firms view as their unique "edge"—and such political considerations as constraints imposed by the incentives of the reimbursement system, specific regulations that may limit firm options for collecting data on race/ethnicity, and others.
4. How to Encourage Sustainability?
In considering how best to encourage sustainability, an important issue to consider—which is not unique to the private sector—involves maintaining continuity of participation and memory on projects that span long periods of time. There are several aspects of this. First, because turnover in firm personnel can be anticipated, it may be valuable to consider ways in which short-term successes can be institutionalized in organizations and thus less vulnerable to change with personnel. Second, because firm participants are likely to have many demands on their time, effective techniques for maintaining their interest without imposing unnecessary distractions also could be valuable. The Web could be used to reinforce decisions and share content and progress between Collaborative meetings. And third, it probably is worth considering upfront to whether to select participating firms in ways that limit barriers from instability. In the health care industry, instability through mergers, acquisitions, and leadership turnovers is to be expected. However, at certain points an organization may be just too distracted to be able to effectively participate in a collaboration.
The other side of sustainability involves doing as much advance thinking as possible about how to sustain work that has been supported by external support organizations after those resources are no longer available to participating firms. In the future, AHRQ may want to consider building more formal requirements for technology transfer into its external support contracts so that public support for technology development can be leveraged to serve a broader audience.