Secondary Logo

Journal Logo

Innovation Reports

The Harvard Medical School Academic Innovations Collaborative

Transforming Primary Care Practice and Education

Bitton, Asaf MD, MPH; Ellner, Andrew MD, MSc; Pabo, Erika MD, MBA; Stout, Somava MD, MS; Sugarman, Jonathan R. MD, MPH; Sevin, Cory RN, MSN, NP; Goodell, Kristen MD; Bassett, Jill S. MS; Phillips, Russell S. MD

Author Information
doi: 10.1097/ACM.0000000000000410
  • Free



Academic medical centers (AMCs) traditionally have focused on biomedical research and tertiary care and may be poorly prepared to meet society’s need for higher-quality health care at a sustainable cost. As AMCs often serve medically and psychosocially complex patients, strengthening their primary care delivery and training could serve important societal priorities. Establishing residency practices as high-functioning patient-centered medical homes (PCMHs) focused on quality improvement may help more trainees envision a career in primary care through improved provider work experience and morale,1 helping to ease the shortage of primary care providers. In this Innovation Report, we describe a learning collaborative model for transforming primary care practice and education at six AMCs that included both their hospital- and community-based primary care teaching practices. Our principal aims were to create a learning community focused on continuously improving systems for primary care practice and education; achieve sustainable improvements in the experience of care for patients and trainees; and improve health care value.


Our innovation occurred in the context of state-based health care reform and a renewed focus on primary care at Harvard Medical School (HMS). In 2006, Massachusetts-based health care reform expanded health insurance, which increased demand for primary care. To tackle some of the highest per capita costs in the country, major payers and providers began to shift toward shared risk and savings models. Massachusetts-based AMCs examined how they would redesign for success, and strengthening primary care was a key element of their strategy. In response to both a local and national workforce crisis in primary care, the HMS dean created a new Center for Primary Care (CPC) to improve the health of our communities through a transformation of primary care practice and education.

Leveraging the renewed local emphasis on primary care, in 2012 the CPC launched the Academic Innovations Collaborative (AIC), a partnership with six major HMS clinical affiliates (Beth Israel Deaconess Medical Center; Brigham & Women’s Hospital; Cambridge Health Alliance; Boston Children’s Hospital; Mount Auburn Hospital; and Massachusetts General Hospital), the Institute for Healthcare Improvement (IHI), Qualis Health, and an external evaluation team. We included major teaching practices based at AMCs, and several affiliated community-based teaching sites at each AMC (primarily community health centers in underserved communities).

The AIC focused on four key areas: (1) establishing high-functioning interprofessional teams, (2) proactively managing populations, (3) identifying and providing tailored care to medically and psychosocially complex patients, and (4) promoting patient engagement and empowerment. These content areas correspond to the theory of PCMH, for which there is mounting evidence of improvement in costs and quality of patient care.2

AIC theory of change

Collaborative improvement networks use standardized quality improvement methods to translate evidence into practice, and support teams to test and implement changes in a reliable, sequenced way.3 A learning collaborative is a basic structure of collective transformation and consists of a sequenced, ordered, and layered series of in-person, Web-enabled, and data-oriented cycles of pedagogy aimed at building accountable capacity for team-based testing and transformation.

We structured the AIC as a “formative” learning collaborative, adapting a prespecified model for change to a very specific local educational and practice context. Practice change was sequenced and measured in a regular way, but local contextual variation and measurement needs were respected and leveraged to ensure buy-in and promote new insights. The tasks of transformation were codified into concise aim statements at each practice with linked measures. Practices rigorously reported and measured their progress across a series of well-defined tests of change. Thrice-yearly in-person learning sessions and cross-site visits provided a means for practices to learn from each other.

Our learning model was an amalgamation of the Safety Net Medical Home Initiative change concepts developed by Qualis Health and the Group Health Research Institute4 for sequenced primary care practice transformation, along with IHI’s model for breakthrough improvement. Change was accomplished through a combination of individualized and team coaching; leadership training; senior leadership engagement; and customized, in-person learning sessions. The first-year focus of the AIC primarily centered on the change concepts of leadership engagement; population empanelment (knowing who the practice cares for at all times); quality improvement; and continuous, team-based healing relationships. In the second year, the focus shifted to organized evidence-based care, patient-centered interactions, enhanced access, and care coordination (see Figure 1).

Figure 1:
Patient-Centered Medical Home Assessment (PCMH-A) scores, reported by the transformation teams of each participating practice in the Academic Innovations Collaborative at six-month intervals, based on Safety Net Medical Home Initiative change concepts developed by Qualis Health and the Group Health Research Network. Scores range from 0 to 12; 12 = best; P < .001 for each comparison of baseline versus 18-month score.

A number of collaboratives of residency training practices have been established across the United States to promote transformation toward the medical home model. In Table 1, we compare salient features of these collaboratives with those of the HMS AIC.

Table 1:
A Comparison of U.S. Primary Care Academic Residency Learning Collaboratives

Leadership engagement, funding support, and organizational structure

Before launching the AIC, the CPC engaged key stakeholders at all six AMCs to create the vision for the collaborative and ensure priority alignment. Acknowledging that each AMC was transitioning from volume-based payer contracts towards shared risk and global capitation requiring upfront investment in primary care, the CPC contributed up to $1 million to each AMC over two years towards practice transformation in a one-to-one match of AMC resources.

The resulting two-year collaborative included 19 primary care practices serving nearly 260,000 patients. In total, approximately $5.4 million was committed by the AMCs and directly matched in the same amount by the CPC, with the CPC committing another $3 million to support technical assistance and an external evaluation. The level of support for transformation, staff, and evaluation efforts corresponded to approximately $3.20 per patient per month, roughly equivalent to many payer-led demonstration projects supportive of primary care transformation.5 A substantial proportion of the investment awarded by the CPC to AMCs protected time for faculty and staff to engage in the work of transformation while the resources committed by AMCs supported new staff for primary care teams.

AIC structure

Prework. Before the launch of the collaborative in July 2012, the CPC supported practice-level prework. Each clinic formed a “transformation team” with an executive leader, a physician champion, an education leader, a day-to-day leader, and representative practice staff. We suggested that each team include trainees and patients as core members. Each transformation team was asked to develop an aims statement to guide their work—for example:

We seek to elevate the quality of care delivered to all of our patients, and to improve patient, staff, and resident experiences. Within two years we will have fully functional patient care teams, in which 95% of patients will be able to identify a member of their care team, 85% of visits will occur within the care team, and staff members will experience increased job satisfaction.

Learning sessions. Three times per year, the transformation teams attended a required in-person 1.5-day learning session that included didactics on practice transformation, team sharing of best practices, and team time to plan upcoming work. The learning sessions offered an opportunity for individuals at different institutions to form connections for future collaboration.

Action periods. During action periods, transformation team members developed and tested changes aimed at improving the clinic’s practice in the sequenced change areas. Most teams met once per week and joined a monthly AIC phone call to share challenges and lessons learned. In addition, the CPC hosted monthly meetings for the day-to-day leaders to support their work, and sponsored regular site visits.

Practice coaching. During the first year, six sites requested and received intensive practice coaching from advisors affiliated with Qualis Health and IHI that included site visits and regular check-ins that matched their needs.

Leadership academy. The CPC designed a Leadership Academy to help emerging practice leaders manage and lead change. Any member of the interprofessional transformation team could participate, ideally with at least two others from their clinic. The first year of the Leadership Academy included nearly 30 leaders and met for a half day every one to two months. The Leadership Academy offered a curriculum on facilitative leadership, leading and managing teams and change, and creating infrastructure for sustainable improvement. Sessions typically included a review of core concepts, a business-school-style case discussion, skill development, and peer mentoring.

Educator and trainee integration. The 19 sites were selected in large part because they were major training sites for medical students and residents across the HMS system. We encouraged each site to include medical students and residents on their transformation teams, and to have those trainees lead improvement projects. Each site was also asked to integrate trainees into their new models of team-based care.

At the end of the first year, the CPC provided additional support to educators and trainees at each site to support continuous systems improvement and practice transformation. On the basis of interviews with residency program directors and relevant literature, the CPC developed a customizable three-year resident curriculum. The curricular goals specified that residents (1) understand the centrality of primary care within the health care system; (2) demonstrate competence in population management, team-based care, and care coordination; (3) demonstrate the ability to lead quality improvement projects; and (4) occupy key roles on AIC teams. Each of these goals was broken down into specific, measurable learning objectives mapped to residency curricular milestones. Because AIC practices and resident schedules were variable, we provided a menu but left the specific curricular activity choice to the sites. Table 2 shows the key features of the resident curriculum that could be slightly modified for students.

Table 2:
Customizable Resident Curriculum Developed by Harvard Medical School’s Center for Primary Care (CPC) for the Academic Innovations Collaborative (AIC), 2012

Measurement. We focused on creating the capacity to track and use performance data. We required each team to have a balanced measurement strategy, choosing from a three-tiered and flexible measurement strategy with a beginner, middle, and expert level. The only required measure was the percentage of patients empaneled to their clinician or care team. Measurement domains included practice operation, clinical process and outcomes, and patient experience. Teams were expected to use their measures to test changes and to report on their measures monthly.

External evaluation. An external evaluation team at Harvard School of Public Health (HSPH) used a mixed-methods approach to measure how well the AIC is achieving its aims and exploring variation in progress across sites. Early findings from the external evaluation were shared with the transformation teams in a rapid-cycle way to improve the intervention.


In its first year, the AIC created an engaged, active learning community. All sites made fundamental changes in how they operate, towards working in teams to more effectively address population health needs. On the basis of repeated use of the PCMH Assessment (PCMH-A, a validated self-report tool that accurately tracks progression toward team-based, patient-centered care),6 we found a consistent and substantial improvement in all eight of the change concepts (Figure 1). We expanded our Leadership Academy to accommodate increasing interest, and practices added patients and trainees to their transformation teams.

Several important themes emerged. First, ensuring progress and building collaboration between the 19 practice sites required balancing central structure and accountability with local flexibility. Sites wanted to be held accountable to goals they set themselves. Second, we needed to engage organizational leaders and train frontline leaders. Change within practices must be aligned with overall strategy of their larger organizations, and, at the practice level, leadership competencies must be augmented and learned if not present. Third, change is not linear. Progress at the practice level may be reliant on, or impaired by, key organizational or personnel changes; the use of practice coaches can help practices who are stuck break through barriers. Fourth, practice and educational change optimally occur simultaneously. Trainees and patients can be agents of change, and their presence provides added motivation for academic faculty to create and support change.

Our findings are limited as they are preliminary and largely qualitative, but are consistent with our results of the validated PCMH-A. Our work is based on a collaborative of affiliates of a single medical school, but it includes six different AMCs serving a large number of patients. Our work required external investment, but is within the range of support provided within nationally representative payer-funded medical home initiatives. Our work is occurring within a local context of substantial pressure on quality and cost, but it is likely that these same pressures will spread nationally.

Next Steps

We have begun a local journey of academic practice transformation to improve how primary care is delivered at AMCs and how trainees are exposed to and involved in new models of primary care. To spread our work and share learning, we are beginning to collaborate with similar collaboratives across the country and will serve as a resource to newly emerging efforts. To expand support, we must rigorously document objective outcomes, and engage payers and providers in supporting such efforts. The health care system has never faced greater challenges, and AMCs—which have long led the way in health care—must position themselves to lead our society forward in its quest for better population health at a cost that will be sustainable.

Acknowledgments: The Academic Innovations Collaborative (AIC) represents the hard work of many individuals. In particular, the authors would like to thank the dean of Harvard Medical School, Jeffrey Scott Flier, MD, for his vision and support of the Center for Primary Care. The authors would like to acknowledge the following groups of people for their critical contributions to the AIC: (1) the leadership of the six teaching hospitals within the AIC, without whose vision and leadership this would not be possible; (2) the principal innovators and coinnovators at each of the six entities who make this vision a reality on a daily basis (Kim D. Ariyabuddhiphongs, MD, Harvey W. Bidwell, MD, Matthew Carmody, MD, Marya J. Cohen, MD, MPH, Joanne M. Cox, MD, Marjorie A. Curran, MD, Judy Fleishman, PhD, Kelly D. Ford, MD, Blair W. Fosburgh, MD, Joseph P. Frolkis, MD, PhD, Barbara Gottlieb, MD, James J. Heffernan, MD, Thomas Kieffer, Colleen A. Monaghan, MD, Charles A. Morris, MD, Roslyn Murov, MD, Barbara Ogur, MD, Holly J. Oh, MD, Richard J. Pels, MD, Linda R. Powers, MD, Yamini Saravanan, MD, Gregory Sawin, MD, MPH, Arshiya C. Seth, MD, Rachel L. Stark, MD, MPH, Julie L. Tishler, MD, Lori W. Tishler, MD, Randy Wertheimer, MD, and Laura Zucker, MD); (3) the teams at the Institute for Healthcare Improvement, Qualis, and the Center for Primary Care, whose tireless support is enabling profound change across the AIC practices (Afiesha Balgobin, Trudy Bearden, PA-C, Nicole van Borkulo, Donna Daniel, PhD, Juliana DiLuca, Caitlin Haskell, Tia Khon, Aka Kovacikova, Regina Neal, Megan Prock, Rebecca Steinfield, Lindsay Swain Hunt, and Trissa Torres, MD); and (4) members of the authors’ evaluation team (Meredith Rosenthal, PhD, Alyna Chien, MD, MS, Sara Singer, PhD, Toni Peters, PhD, and Julia Martin). Finally, the authors would like to thank Ursula Koch, MSc, MS, Ashlin Mountjoy, Elizabeth Griffiths, and Casey Maloney for assisting with preparation of this manuscript.


1. Lewis SE, Nocon RS, Tang H, et al. Patient-centered medical home characteristics and staff morale in safety net clinics. Arch Intern Med. 2012;172:23–31
2. Peikes D, Zutshi A, Genevro JL, Parchman ML, Meyers DS. Early evaluations of the medical home: Building on a promising start. Am J Manag Care. 2012;18:105–116
3. Clancy CM, Margolis PA, Miller M. Collaborative networks for both improvement and research. Pediatrics. 2013;131(suppl 4):S210–S214
4. Wagner EH, Coleman K, Reid RJ, Phillips K, Abrams MK, Sugarman JR. The changes involved in patient-centered medical home transformation. Prim Care. 2012;39:241–259
5. Bitton A, Martin C, Landon BE. A nationwide survey of patient centered medical home demonstration projects. J Gen Intern Med. 2010;25:584–592
6. Daniel DM, Wagner EH, Coleman K, et al. Assessing progress toward becoming a patient-centered medical home: An assessment tool for practice transformation. Health Serv Res. 2013;48(6 pt 1):1879–1897
© 2014 by the Association of American Medical Colleges