Each model State worked with three or four partner States to encourage spread of their extension approach. These efforts are described in this section.
New Mexico
The New Mexico IMPaCT project partnered with Kansas, Kentucky, and Oregon. New Mexico did not seek to replicate its own structure in the partner States, but instead created a learning community to share information about how to build partnerships that could support transformation in the context of each State’s unique set of needs and resources. Shared learning from the New Mexico program continues to build, and additional States, including Michigan, Utah, and Georgia, have been invited to work with New Mexico’s HERO program to learn from its efforts.
The New Mexico IMPaCT team worked closely with the Department of Family Medicine at the University of Kansas to launch a Kansas version of the New Mexico program, exploring a partnership with Kansas State University. Kansas State University, in turn, hosted a statewide meeting presenting and discussing this model, linking health extension with agricultural Cooperative Extension. New Mexico’s HERO program consulted with Kentucky about opportunities for developing a statewide improvement infrastructure in that State.
The creation of the learning community emerged as a result of in-person meetings among all four States involved in the project in Albuquerque, New Mexico and Portland, Oregon, in addition to ad hoc meetings at conferences and national meetings where all four States were present. In all States, a relationship was formed between the academic health center and the land grant university’s Cooperative Extension Service in the development of a statewide health extension service, and presentations were made at professional conferences of both medical and extension disciplines.
The result of this learning community was the development of an online health extension toolkit (www.healthextensiontoolkit.org), which was later broadened to include all 18 States involved in PCEP nationally (as described below in cross-grantee efforts).11 This toolkit encompasses the multiple models in which health extension has emerged across the country, and was built to fit multiple perspectives—from academic health centers, to land grant universities, to public health infrastructure, to primary care practices.
North Carolina
North Carolina partnered with Idaho, Maryland, Montana, and West Virginia. North Carolina selected the four partner States through a competitive application that was designed to assess applicants’ comprehensive capacity for improvement, including a commitment to advancing policy and practice changes; multisector State-level collaborations; and involvement of an entity that could function as a primary care extension service. The selected partner States received individual and group technical assistance from North Carolina faculty and experts and peer-to-peer learning with other partner States. These spread efforts were coordinated by the National Academy for State Health Policy.
While North Carolina IMPaCT focused on amplifying existing QI support through the Regional Leadership Collaborative, efforts in partner States focused on helping teams understand how the North Carolina support structure could be applied to the existing networks within their State. Each State convened a team of about five people, consisting of State government representatives and providers. There were two in-person meetings, several Webinars, and frequent interactions throughout the life of the grant. Each State adapted the North Carolina model to meet their State’s specific needs, as described below, but all gravitated to the care management resources and insights, likely because of the potential for cost reductions.
Idaho
Idaho took a statewide approach to the redesign of primary care support. The core team included members from the State Medicaid agency, the Idaho Hospital Association, the Idaho Medical Association, and the Idaho Primary Care Association. These partners designed a statewide improvement model building on existing initiatives, including the Idaho Medical Home Collaborative, a collaboration of primary care physicians, health care organizations, and payers who make recommendations to the governor on statewide PCMH efforts. Extensive outreach to multiple stakeholders led to application for a State Innovation Model design grant.
Maryland
Maryland built its program on the Maryland Learning Collaborative, which provided practice coaching to help practices achieve PCMH recognition from the NCQA, with funding from the Multi-Payer Program for PCMHs. The University of Maryland used the North Carolina AHEC model of practice facilitation and coaching to reach out to primary care practices. Maryland hired and trained two practice coaches and each supported approximately 15 practices. Maryland also extended North Carolina’s care management work; each participating practice identified a person to lead care management tasks with a job description based on CCNC guidance. Maryland assisted practices with electronic health record implementation and with connecting to the State health information exchange. With assistance from a combination of practice coaches and expert consultants, each of the 52 participating practices received NCQA Level 2 or 3 PCMH recognition, and all implemented an electronic health record.
Montana
Montana built upon existing State efforts to help practices achieve NCQA PCMH recognition, as well as AHEC efforts, such as community health worker training. Montana introduced and advocated for legislation to promote PCMH through provisions to allow multiple payers to share the costs of primary care transformation, State oversight, and rulemaking authority for standards for the insurance commissioner. While use of the political process had the advantage of fostering alignment at high levels, this approach was slower than expected. Montana also extended the data infrastructure available to support improvement through partnership with a Quality Improvement Organization.
West Virginia
West Virginia convened many stakeholders to produce a shared vision and a white paper titled “Building the Infrastructure for a Healthy and Prosperous West Virginia.” The West Virginia Health Care Innovation Collaborative originated from this work. The Collaborative is a public/private partnership working to achieve better health care quality, lower health care costs, and better health outcomes for West Virginians through healthy lifestyles.
Oklahoma
Oklahoma partnered with Arkansas, Missouri, and Colorado. Oklahoma’s spread efforts included regularly scheduled phone, email, and in person visits between the IMPaCT PI and State representatives. Oklahoma’s PI was central in convening the national IMPaCT conference and encouraged spread across States, adapting to very different contexts.
Colorado
Colorado already has a powerful QI network with strong organization and experienced support personnel, particularly in the Denver area. The State also already has a strong commitment to establishing an extension center, and key stakeholder organizations meet regularly to engage all of the major players in primary care, health policy, and practice transformation efforts in the State. Colorado plans to build on its practice facilitation resources and deploy them through a network of collaborating organizations. It would like to adopt the CHIO model, and is currently working to find the appropriate geographical units for organizing the regions. (It may choose to use Medicaid districts, since some of its counties are sparsely populated.) Additionally, an appropriate division of roles among key stakeholders also needs to be determined. Colorado used its work with the Oklahoma and New Mexico primary care extension projects to lay the groundwork for pursuing funding through several sources and is building the extension center concept into its State Innovation Model grant application.
Arkansas
Arkansas currently has two QI initiatives active at the State level that are stimulating the development of PCMHs in primary care—the Comprehensive Primary Care Initiative, sponsored by the Centers for Medicare & Medicaid Services, and the statewide PCMH initiative. The University of Arkansas Medical School regional centers, part of the AHEC system, are also spreading redesign efforts, beginning with their many primary care residency programs and outreach to other practices. These regional centers developed a toolkit for PCMH implementation and piloted it with three practices outside of the University of Arkansas system, spreading the use of best practices such as disease registries, care coordination, information technology support, and chronic disease workflows. The CHIO model is also of interest to stakeholders in Arkansas, but has not yet been implemented. Arkansas has a strong QI environment in terms of reimbursement incentives, support mechanisms, and tools in place, and will therefore be ready to implement a primary care extension service once funding becomes available.
Missouri
The IMPaCT mission in Missouri was complicated by fragmentation in the State’s QI initiatives. There were several active stakeholders, such as the AHECs in St. Louis, Columbia, and elsewhere, as well as a statewide PCMH initiative. So far, however, there has been little communication and coordination between these efforts. The IMPaCT team in Missouri made an early decision to focus their efforts on Pettis County, where they have an established relationship with the University of Missouri, Columbia. The team was able to connect primary care practice managers to community resources, so that each practice now has a resource guide. They have also worked to help advance conversations about PCMH and population health management with interested practices. There is growing communication among local hospitals, Community Health Centers, and public health entities in Missouri about primary care transformation, and the University has joined with several rural hospitals in a collaborative effort to improve care and support smaller health systems. Although the hospital and Community Health Centers are already using electronic health records, they are not yet able to provide adequate data for population health management, and only one private primary care practice had one in place. At the State level, the legislature approved funding for a telemedicine project (Project ECHO) to provide consultation to frontier practices. These efforts aim to break down the existing silos that prevent communication and collaboration on QI between groups.
Pennsylvania
Pennsylvania partnered with New Jersey, New York, Pennsylvania, and Vermont. Pennsylvania’s activities included an environmental scan in each State to learn what funding and support was already in place to advance primary care transformation. Members of PA SPREAD’s National Advisory Group helped make connections and provided key information on each of the States. PA SPEAD’s General Contractor Model proved effective in disseminating transformation activities to the partner States.
Stakeholder meetings were held with the partner States and leaders in primary care transformation efforts from each State met regularly to share their experiences. Because each State was at a different point on the continuum of practice transformation, PA SPREAD acted primarily as a conduit of information sharing among them. New Jersey was at the beginning of the process of implementing practice transformation efforts, while New York was well along the path of transformation, and Vermont was an excellent example of accomplished practice transformation. The partner States learned as much or more from sharing their experiences with each other as they did from the lessons learned from Pennsylvania as the model State.
A second important component of the spread activities in the partner States involved bringing together individuals from within each State who were previously unknown to each other. Meetings allowed time for State-level representatives to talk about internal State issues. Just as this type of relationship building had been critical in the success of spread activities within Pennsylvania, New York was able to further its transformation efforts by connecting work being done in the Adirondacks, Hudson Valley, and Beacon communities and bringing in United Health Group as a partner. These efforts culminated in a State Innovation Model grant submission in New York.
Cross-Grantee Efforts
In addition to activities between model States and partner States, collaborative efforts emerged between the four IMPaCT grantees. The PIs and coordinators talked by phone and emailed regularly throughout the project period, often including AHRQ’s project officer in these discussions. The teams wrote a white paper on the value of developing a national PCEP, including funding options. Model and partner State teams organized and led a national IMPaCT conference in Oklahoma in February 2013. The PIs and coordinators also organized a national PCEP meeting in Washington, DC in February 2014, funded by a grant from the Commonwealth Fund. A white paper that discussed the future of the PCEP was produced from this meeting. As previously discussed, New Mexico and its partners created an online toolkit documenting their insights about health extension. With a grant from The Commonwealth Fund, the toolkit was expanded to produce a national health/primary care extension toolkit incorporating information about all four IMPaCT grantees’ programs.11 The four grantees participated in a Webinar to launch and disseminate the toolkit, held in September 2013. Some IMPaCT awardees participated in an additional meeting funded by The Commonwealth Fund in February 2014.