KEY SPEAKER POINTS
- There is a business imperative to conduct research related to institutional goals for health care delivery, James Rohack said, but governance and related operational mechanisms need to be in place to shorten the cycle time from research evidence generation to related health care organization management decisions.
- Grant funding cycle times make it difficult for grant sources to have major roles in research related to a health care delivery organization’s operational goals, Rohack said, and so sustainability should depend primarily on operational funds.
- One way to send the message that health service delivery research is important is to give it the same weight as basic science research in an organization’s reward system, Rohack said.
- “From my standpoint as a leader of an organization,” Mary Brainerd said, “I want you to bring your ‘patientness’ with you to everything that we do, everything we design, every way we think about what we are going to do in research.”
- The time frame for most research is too long, and too much knowledge is not put to use in a productive manner, Brainerd added.
- Transformation of a research enterprise to conduct studies that are larger, of higher quality, faster, less expensive, and more engaged requires transformation of governance, John Steiner said.
- One possible operational definition of network sustainability, Steiner said, relies on the development of shared research assets to facilitate a sequence of research studies in a specific content area or multiple areas and developing a community of researchers and other stakeholders who reuse and develop those assets, both technical and human/cultural.
- Steiner said that the governance of interorganizational research “requires us to develop precision tools on the one hand but also to permit the creativity to use those tools in new ways.” The culture of leadership and decision making in research networks can be characterized as one of leadership without control.
Institutional governance of continuous learning activities that can accelerate progress and sustainability was the focus of a panel of brief presentations by James Rohack, Chief Health Policy Officer for Baylor Scott & White Health, who discussed how his organization aligns research with institutional goals; Mary Brainerd, President and CEO of HealthPartners, who spoke about data sharing in a competitive environment; and John Steiner, Senior Director of the Institute for Health Research at Kaiser Permanente Colorado, who addressed various issues involved in governing interinstitutional research. An open discussion followed the presentations. Session moderator Paul Wallace, the chief medical officer and senior vice president for clinical translation at Optum Labs, started the session by saying that most of the governance that has been created has been for intrainstitutional issues. “We figured out how to check the boxes and get things done within our own shop,” he said, “but what is changing is that we have a national context now, and we have to work across institutions.” He cited PCORnet as the most robust effort at developing interinstitutional governance, and he mentioned other examples, including the NIH Collaboratory and AcademyHealth’s Electronic Data Method’s Forum and its Generating Evidence and Methods (eGEMs) to improve patient outcomes project. He noted that one challenge is to design durable governance structures that are sustainable but not static.
Baylor Scott & White Health, as James Rohack explained, was formed in October 2013 when 117-year-old Scott & White Health Care in central Texas merged with 107-year-old Baylor Healthcare System in the Dallas, Texas, area. The combined organization has a unified approach to governance, which includes education because, Rohack said, education has to be part of the goal of aligning health care with research if “we are going to have the new cadre of people delivering care who understand how culture and quality is part of a driver to improve care all the time.” Rohack described the STEEEP (Safe, Timely, Effective, Efficient, Equitable, and Patient-Centered) Quality Institute that works with system researchers as well as with outside clients to identify opportunities for making health care more safe, timely, effective, efficient, equitable, and patient-centered.
Baylor Scott & White Health is committed to what Rohack called “lean thinking.” Lean thinking, he explained, is not about “more is better” but rather focuses on providing the right care at the right time to the right person. “The mission embeds that personalized care,” he said, “but it also looks at research and improvement as a part of that culture,” and an important piece of realizing that mission is innovation in how to scale things across the entire combined organization. He said that Scott & White had had its own health plan for 35 years, is part of the HMO Research Network, and has been dealing with population health in the capitated payment model, and it now has to push that model through the integrated system (see Figure 7-1).
Having circles of care and innovation and creating strategy maps to align these models is all for the good, Rohack said, but the key is operationalizing these things. That requires, he said, engaging the board of trustees on a regular basis and informing them of progress toward meeting specific metrics for success as well as creating a leadership council composed of the people who are on the ground leading the specific research programs designed to reduce mortality and morbidity, improve patient experience, manage health at a population level, and improve financial operating margins.
Rohack offered four lessons that the northern division of Baylor Scott & White has learned in its efforts to create a learning health care system based on these models. Clearly, he said, there is a business imperative to conduct research related to institutional goals for health care delivery, but governance and related operational mechanisms need to be in place to shorten the cycle time from research evidence generation to related health care organization management decisions. Another lesson was that, although federal funding can be helpful for this type of research, typical federal peer review funding cycle times make it difficult for these sources to play major roles in research related to a health care delivery organization’s operational
FIGURE 7-1 The Baylor Scott & White circle of innovation.
SOURCE: Reprinted with permission from James Rohack.
goals, and so sustainability should depend on operational funds, not grants. Finally, successful delivery organizations will have a robust infrastructure in place to support these aligned research efforts as a core component of day-to-day operations.
One issue that Baylor Scott & White has addressed is the need for a workforce that understands in a deep way the role for and the value of quality-improvement research. Its approach in its central division has been to embed its research projects as a core function and engage its medical residents in the design, planning, and execution stages. “Every resident receives didactic training on how to conduct such projects and how to integrate the learning that results into their day-to-day practice,” Rohack said. To emphasize the importance of this training, Baylor Scott & White gives health service delivery research the same weight as basic science research in its reward system. “This is an important message,” Rohack said.
As an example, he briefly discussed a traumatic brain injury research project that connects basic scientists exploring the blood brain barrier with clinical scientists exploring biomarkers for traumatic brain injury, clinicians studying how the symptoms manifest, and quality investigators exploring how data can be used to improve quality of care and quality of life. Together, these teams are aligned in a way that will allow them to
address questions regarding how to best deliver care to patients who have traumatic brain injuries.
One lesson learned from the central division’s experiences, Rohack said, is that partnerships are hard because they rely on the personalities of those in charge. He said that the High Value Healthcare Collaborative, of which Baylor Scott & White is a member, has been successful because of the personalities of the leaders and the researchers, who are willing to share with others. Rohack added that as part of the recent merger of the Baylor Scott & White systems, they have encountered a number of challenges associated with bridging academic and clinical service organizations. He highlighted issues with faculty titles and tenure—tenured versus nontenured, professor versus clinical professor—as a major stumbling block. He cited a lack of understanding across partner organizations, for example, when the academic center does not run a clinical enterprise but expects the clinical enterprise to teach the medical students and residents. Who controls the indirect dollars from outside grants can be a difficult issue to resolve, he said, particularly in instances where the money flows through the academic center, which takes its cut for overhead and leaves the clinical enterprise to use more of its own funds to support the work; “Trust is the key here,” he said. Rohack concluded his remarks by highlighting what he considered an underappreciated crisis, saying that it is important in all of these efforts to remember the incredible pressure that health care providers are under these days. “There is a crisis in caregiving that nobody seems to be talking about—caregiver satisfaction,” he said.
The origins of HealthPartners as a not-for-profit, consumer-governed organization that grew out of the credit union cooperative movement have a great deal to do with how governance is organized at this health system, Mary Brainerd told the workshop. HealthPartners was formed when credit union leaders looked at why their members were in debt and found that the main reason was health care. The result was the first prepaid health plan coupled with a consumer-led delivery system. “I think no set of roots could be more closely aligned with the Triple Aim than people wanting affordable health care and wanting it for themselves and their families and the quality and delivery system approach that worked for them,” Brainerd said.
Today, she explained, HealthPartners resembles the Geisinger system in the sense that only about 35 percent of its members use the HealthPartners delivery system and only about 35 percent of its delivery system patients are HealthPartners members. Brainerd said that HealthPartners has about 1 million patients, 7 hospitals, and 22,000 employees. She added that HealthPartners is somewhat unusual in that it conducts dental research to
inform its 60 dentists at 22 locations, making it one of the few health care organizations working on dental issues.
As part of its mission to improve health and well-being in partnership with its members, patients, and community, HealthPartners has come to realize that it “needs to do a better job of caring for and treating whole people, not just medical issues but mind, body, and spirit,” Brainerd said. “We know there is a whole world of opportunity for us to be more effective.” The organization has come to stress partnerships with its patients and members. “From my standpoint as a leader of an organization,” Brainerd said, “I want you to bring your patientness with you to everything that we do, everything we design, every way we think about what we are going to do in research.” For example, she said, one of her goals is to eradicate words such as “comply” and “adhere” and to eliminate thought processes that involve “doing” something to patients; instead, the mind-set should be to think about what caregivers can do together with the people served. “This redefinition of partnership should be and is at the core of the work of PCORI,” she said. In that type of partnership the only measures that count are the ones that matter to patients: Do you feel better? Can you do what you need to do every day? Can you accomplish the goals you have for your life?
Partnerships with outside organizations are also important, Brainerd said. HealthPartners, for example, has partnerships with the HMO Research Network and the Institute for Clinical System Improvement. One important feature of the latter partnership is that it requires sharing its data publicly and transparently. HealthPartners has also been the sponsor and creator of the Minnesota Community Measurement Community, which delivers performance metrics directly to consumers in Minnesota. One outcome of being part of the Measurement Community was that HealthPartners’ orthopedic department, which scored below average on quality metrics, began an effort to improve its standings on future assessments. “Sooner or later,” Brainerd said, “results are going to be known everywhere, so we thought it was an advantage to get out in front of them and focus on improving performance, not challenging the appropriateness of measurement.”
She noted that when she looks at where her organization is spending its money, she preferentially funds those projects that lead to performance enhancement. “Why?” she asked. “Because the time frame is too long for most research, and too much of the knowledge that we have already created isn’t being put to use. We need performance change on time horizons that are much more rapid than those created by a traditional research structure. That’s not to say that we have decided that research is not important. It is just not as important to the performance gains that I need to see as we look to deliver on the triple aim.”
Brainerd said she is now working to better align the organization’s performance improvement initiatives with its research initiatives. She said she hopes that HealthPartners’ current work with Partners for Better Health Goals 2020 to create long-range stretch goals for meeting the triple aim will help her organization’s educational, research, and operational focus. She is also emphasizing creating systems that make it easy to do the right thing. “I am concerned that when we get research results, we don’t have an effective way to implement them quickly to become reliably delivered and sustained,” she said. One approach that Minnesota is taking is to work through Minnesota Community Measurement and the Institute for Clinical Systems Improvement (ICSI) to drive change in all of its members by helping them implement improvements identified by individual members. It is a highly collaborative approach to improve results everywhere across the region.
HealthPartners, Brainerd went on to describe, has developed a ClickReduction program as part of its commitment to its physicians to reduce the amount of time it takes to do the things effectively using EHRs and to make the right thing easier to do. She noted that from her perspective there is not a problem of knowledge hoarding, as was discussed previously, but rather a failure to implement the knowledge that already exists about what works in a way that can realize real improvement in quality and outcomes.
In her closing remarks, Brainerd said that she is optimistic that early efforts to learn how to partner with patients will yield important improvements. A key to this effort will be to learn how to communicate better with patients and not leave them confused by the often conflicting information that so often appears in the popular media. “We have, through all of the knowledge that we have created, confused our patients and consumers to the extent that even care that would be beneficial is sometimes not being sought,” she said. “I think the opportunities are there to connect in new and different ways with our patients, to remember they come to us with their own knowledge and beliefs.” Patient councils can be one avenue for connecting, and Brainerd noted that a patient council that was created to help redesign a physical space for mental illness care not only accomplished that but also ended up catalyzing a new care model for patients with mental illness. She also expressed concern that there is not enough research on the intersection of health care and behavior change. For example, it would be valuable to learn more about how to connect better with patients to encourage healthy behaviors or how to increase the likelihood of effective medication use because currently medications are used effectively only about 50 percent of the time.
Speaking from his perspective as a scientist running an 11-site diabetes network with EHR data from more than 1.3 million individuals with diabetes as well as the chair of both the Kaiser Permanente National Research Council and the HMO Research Network governing board, John Steiner said he tries to make sense of the relationship between sustainability and governance. His first message in that regard was that “transformation of a research enterprise, which we all agree needs to happen, really requires a transformation of governance.”
Research studies, he said, need to be larger, which means they must include more sites and subjects. Research needs to be of higher quality, which means it needs to use trustworthy, high-quality data and better analytic methods while achieving or maintaining regulatory and fiscal compliance. Research needs to be faster, in terms of initiating studies, organizing the contractual relationship between collaborators, and getting studies approved by IRBs. Research need to be less expensive, which means relying not only on data collected primarily in the course of large, randomized trials but also on data collected during the course of routine care. Research needs to be more engaged, which means integrating patient and organizational/clinical perspectives as well as the members of interinstitutional research teams. The connection among all of these needs is that they all require skillful governance, Steiner said.
Steiner offered one possible operational definition of network sustainability that consisted of two parts. First, sustainability involves the development of shared research assets in order to facilitate a sequence of research studies in a specific content area or multiple areas. Second, sustainability has to do with developing a community of researchers and other stakeholders who reuse and develop those assets, both technical and human/cultural. Governance of shared institutional technical assets requires a great deal of attention to detail. The governance of a community requires broad principles and a light touch. Summarizing these ideas, Steiner said, “Governance of interorganizational research requires us to develop precision tools on the one hand, but also to permit the creativity to use those tools in new ways.”
Concerning research assets—that is, the precision tools—the organization and governance of a network needs to be well defined, and the relationships among institutions need to be clear. “This is not limited to what we often focus on in governance,” Steiner said, “which is governance of the data. That aspect has to do with discussions about the data model you use, how you ask questions of that data and of the sites that hold it, as well as a whole long list of issues around assuring the quality and validity of that data.” Although it is important to have mechanisms in place to ensure
that data are trustworthy and that there is a data governance structure, the other elements of sustainability—including shared knowledge about tools and research methods, administrative efficiency, physical infrastructure such as biobanks, and predictable infrastructure funding—have to be done effectively for this kind of research to work.
Next, there needs to be governance that helps foster a shared sense of mission, vision, and values and that protects human subjects, who are also stakeholders. Governance structures need to create a strong relationship with the delivery system, which is another stakeholder. Finally, Steiner said, governance should encourage the development of a culture of leadership and collaboration based on fair and transparent decision making. Each of these general items, he said, generates a long list of specific governance issues that require consideration if one is to create sustainable networks.
One of the lessons about interinstitutional research collaborations that Steiner said he has learned is that single-investigator-driven projects rarely add substantially to infrastructure. “To achieve stable governance of a research network, you need to be able to adapt to different principal investigators, different lead institutions, different scientific priorities, and a whole range of varying pressures and incentives,” Steiner said. “Governance structures need to be flexible enough to adapt to the legitimate needs of those projects while also being inclusive enough to gather the learnings from each of these independent scientific networks and studies and to collate them into some organized whole so that you develop common models of analyzing data and the like.” The second lesson is that even with substantial infrastructure investments, research networks are unlikely to become independent from institutional support.
Turning to the subject of leadership, Steiner said that the culture of leadership and decision making in research networks can be characterized as one of leadership without control. “These organizations are decentralized, and it is impossible to enforce top-down mandates,” he said. Instead, the emphasis has to be on decision making based on trusting the capacity of researchers to self-organize and innovate. Drawing on chemistry for an analogy, Steiner said the organizations and investigators in a network are linked by hydrogen bonds rather than covalent bonds.
As an example of how such an interinstitutional research network can form, Steiner discussed how the seven regional and independent research departments in Kaiser Permanente came together under the umbrella of a strategic research plan. This plan, Steiner explained, led to the creation of the Kaiser Center for Effectiveness and Safety Research in 2009, which in turn made the conversations between researchers and organizational leaders across the Kaiser system more systematic. This center has invested heavily in building a data coordinating center, which has enhanced Kaiser’s data model and data quality. Kaiser is now having conversations about the use of
metrics to judge how its seven IRBs are functioning and to start the process of coordinating IRB activity throughout the Kaiser system.
One of the best decisions that Kaiser made, Steiner said, was to develop a national research administrators council whose members tackle the thorny problems associated with contracting, pre-award work, and post-award work, which all bog down research in an environment that needs to be more nimble in order to execute on task orders and contracts within a 12- to 24-month time horizon. These actions have facilitated several new initiatives at Kaiser, including the creation of a Kaiser Permanente national biobank whose goal is to link rich clinical information with biological samples.
In conclusion, Steiner reiterated what other speakers had said over the course of the day: Make the right thing to do the easy thing to do. He also quoted a Taoist precept that says, “Ruling a big country is like cooking a small fish. Too much handling will spoil it.”
To start the discussion, moderator Wallace asked the panelists if they had any thoughts about whether there were features of a network that could be specified so as to not duplicate efforts and that would provide the capacity to be adaptive with other features. Brainerd thought that one good area for collaborative work would be measurement, where work under way could be coordinated so as to not produce too many of the competing measurements that often stand in the way of progress. Steiner cautioned against overgoverning, and, along those lines, Rohack wondered if the meaningful use criteria are a distraction. “Are they taking people away from what we want to do, which is to incentivize people’s intrinsic motivation to do this kind of research, with mastery, purpose, and autonomy built into it?”
Harold Luft of the Palo Alto Medical Foundation asked how governance structures can help increase data transparency among collaborators without putting institutions at risk when data are made public. Brainerd said that in Minnesota, there was a history of shared measurement development among health systems in the state that forced organizations to be transparent with one another and to give up control of what was being measured. In addition, the collaborating organizations committed to a year of collecting data on a measure before publishing any specific measure. Wallace added that the opportunity exists to develop governance structures that encourage data sharing within the constraints of boundaries that have yet to be established. It was noted by several participants that sharing and publishing data are risky.
During the course of the discussion, Steiner reiterated the importance of governance structures that enable research to occur without interrupt-
ing clinical workflows. “The less the disruption, the more likely it is to get top-level buy-in,” he said. Sometimes, however, it is the workflow itself that needs to be studied, Steinder added, and this is where codesigning interventions with the people who have to carry them out and the people who have to live with the consequences is of critical importance. Susan Huang noted that in the REDUCE MRSA trial, every thread of every protocol that was developed considered whether a particular action would be feasible in common practice. Involving both patients and those on the front lines of clinical practice has to be a critical part of any discussions and planning activities, and that must be part of any governance structure, she said.
This page intentionally left blank.