The imperative to change from an organizational culture and structure built around subspecialists and lucrative procedures to one that optimizes patient experience and health outcomes is one of the greatest challenges confronting academic medicine today. To drive change in this direction, the Harvard Medical School (HMS) Center for Primary Care led an ambitious 4-year initiative, called the Academic Innovations Collaborative, to transform primary care delivery in the clinics where most Harvard trainees learn how to practice primary care. It did so with expert assistance from Qualis Health and the Institute for Healthcare Improvement.
The learning intervention, which ran from 2012 to 2016, used a team approach within residency training practices and community health centers affiliated with the medical school. It included in-person learning sessions, monthly webinars, practice coaching, and leadership engagement. An external evaluation team from the conducted a rigorous evaluation of the intervention’s impact on cost, quality, and experience.
The initial results are promising. The evaluation team reported that in the first 18 months, by appropriate ambulatory care at the collaborative’s teaching practices decreased by more than 10%, compared to control practices. There were also statistically and clinically significant decreases in outpatient visits, emergency room use, and total hospitalizations, as well as improvements in staff perceptions of their team’s performance. Importantly, data from the indicate patient experience improvements in several areas, including willingness to recommend the practice to others.
High-functioning primary care is at the core of strong health care systems. In this article, we offer five strategies that were pivotal in successfully managing this complex change process, in the hope that they will be valuable to others planning similar efforts.
As background, the HMS Center for Primary Care was founded in 2011, with an initial gift of $30 million, to play a catalytic role in elevating and strengthening primary care — which previously had a modest presence at Harvard Medical School. The center invested nearly $10 million, with matching funds from six major Harvard-affiliated hospital systems, to transform the organizational culture and structure at clinics where Harvard students and residents learn primary care skills toward a team-based, patient-centered model. A $3.2 million grant from the CRICO Risk Management Foundation extended the collaborative until 2016.
Five Keys to Leading Transformational Change in Primary Care
- Evidence-based change concepts and tools
- Fostering strong relationships within and across practices
- Simple system for reflection and feedback
- Structured time for team discussion and planning
- Regular and meaningful engagement of leaders
In developing a strategy for managing this change process, we frequently consulted with experts and literature on quality improvement, change management, and organizational behavior. Our approach was particularly aligned with the framework that Chip and Dan Heath articulated in their bestselling 2010 book Switch: How to Change Things When Change Is Hard. They observe that leaders need to appeal to both the rational and emotional sides of their constituents as they manage their group through complex change. The five principles below reflect the Switch framework in the context of large, hospital-based primary care practices and community health centers.
1. Evidence-Based Change Concepts and Tools
Health care providers, particularly physicians, rely on scientific evidence to guide action. From the start of the innovations collaborative, we were careful to cite literature on gaps or deficiencies in the quality and safety of the current state of primary care delivery and the evidence supporting the new approaches we were introducing.
The Safety Net Medical Home Initiative change concepts were key to the overall change process toward a team-based care model. The change concepts, part of a five-state initiative to help safety-net providers become high-performing patient-centered medical homes, offer a framework for primary care practice transformation, with a heavy emphasis on leadership engagement as the starting point. (The initiative was funded by The Commonwealth Fund, Qualis Health, and the MacColl Center for Health Care Innovation.)
Another important tool introduced during the collaborative’s last 2 years was the driver diagram. Driver diagrams help translate a high-level improvement goal into sequentially smaller and smaller “chunks” for action. They are typically contained on one page, illustrate how an improvement goal might be met, and provide a menu of intervention options for teams to test. In developing driver diagrams for the collaborative, we conducted extensive literature reviews and expert interviews.
In October 2014, we introduced one for reducing missed and delayed diagnoses for colorectal cancer during one of our tri-annual learning sessions, careful to highlight the role of scientific evidence and expertise in constructing it. Teams had time to review and adapt the tool, as needed, for their practices. The driver diagram became a road map for improvement efforts, and many practices tested and implemented changes drawn from or inspired by it.
2. Fostering Strong Relationships Within and Across Practices
While participants in the Academic Innovations Collaborative wanted to improve quality and safety for their patients, they were also motivated by the opportunity to connect with like-minded professionals within and across teams and practices. The intervention offered an antidote to the isolation and monotony of primary care with its tri-annual, in-person learning sessions.
These 1- or 2-day events drew between 200 and 300 multidisciplinary team participants. We designed them to be high energy and interactive, and to give attendees a chance to share best practices, challenges, and successes with other practice participants. The agenda maximized opportunities to connect participants around topics of mutual interest, from an “Epic Learning Lunch” to Medical Director breakfasts. Participants reported feeling energized; one participant remarked, “We drew inspiration from other sites and from experts . . . getting ‘fired up’ and bringing the enthusiasm back [to the practice].”
In addition, the collaborative’s planning team visited individual practice sites each summer. In the final year, our team sought to further promote cross-institutional learning and relationships by inviting other practices to join the site visits.
3. Simple System for Reflection and Feedback
We required teams participating in the intervention to complete monthly progress reports and other ad hoc assignments. These reports were 6- to 10-page slide decks that included teams’ aim statement, actions taken, questions/challenges for which leadership help was needed, run charts (graphs that display data in a time sequence), and at least one recent plan-do-study-act cycle. The reports compelled teams not only to catalogue their accomplishments and track monthly data, but also to reflect upon learnings and challenges. Faculty members reviewed them and highlighted individual team successes on monthly calls, an act that promoted positive peer pressure.
There were other occasional assignments, such as the request for teams to identify their process for tracking rectal bleeding (spoiler: most teams didn’t have a process for this). While these requirements were added work for busy teams, many practice leaders used the reports in their communications with senior leaders and cited these reports and assignments as the needed nudge to propel their improvement work and stay focused.
4. Structured Time for Team Discussion and Planning
Among the challenges that fee-for-service financing has introduced into health care is a reward system that makes it difficult and expensive for practices to set aside time for organizational improvement. The paucity of protected improvement time can create a vicious cycle in which health systems and practices rarely address inefficiencies and problems due to the “tyranny of the urgent,” leading to progressively less time for important but non-urgent activities, such as team meetings and improvement work.
Many of the collaborative practices initially had no regular, structured time for clinical and managerial teams to meet. However, the teams that did make time (for example, by not scheduling patients at the beginning of the day) found it to be extremely worthwhile. These sessions can be used to discuss improvement efforts, emerging challenges or changes, or other issues that require attention and action.
Each of our tri-annual learning sessions included ample time outside the clinic for teams to meet, plan, and deepen interpersonal relationships, and this was consistently rated among the top activities in end-of-event evaluations. One participant reported, “Learning sessions provided us with the opportunity to meet for [an] extended period with team members without distractions of day-to-day emergencies.”
5. Regular and Meaningful Engagement of Leaders
While leadership engagement — the first of eight Safety Net Medical Home Initiatives change concepts — was a key design element from the start, the way that leaders were engaged evolved during the collaborative. A steering committee with representation from each institution met three times per year and at the learning sessions. There was also a “Leadership Academy” that provided coaching and content for the frontline leaders at practices.
However, one of the most powerful engagement activities took place at the end of each learning session. During this 60-minute exercise, senior leaders from the institutions — often including C-suite staff — met with their practice teams, who presented their aims and plans for the next 3 months. This was also an opportunity for teams to discuss various challenges and concerns with their leaders and to jointly explore possible solutions.
There were many successes over the 4 years, but there were also challenges that slowed improvement efforts. A number of the organizations struggled through painful transitions to new electronic health record systems, but they also benefited from the wisdom shared by other institutions that had already adopted the same system. Turnover of staff, clinicians, and leaders was another challenge. In some cases, the collaborative gave participants the experience and ability to apply for bigger roles elsewhere. In other cases, the new culture of improvement created in many practices drove out those for whom it was not a good fit. While difficult, this turnover allowed practices to hire candidates who were excited by improvement and change, and to build teams with the skills to succeed in the ever-changing climate of primary care.
Change is always difficult, but it is particularly challenging in primary care practices that have been running on the hamster wheel of fee-for-service medicine for decades. The five principles described above, like the Switch framework, sound deceptively simple and straightforward. Yet, for most practices, it would have been difficult — if not impossible — to achieve these fundamental changes without the powerful momentum of the collaborative framework. While the transformation journey for these practices is far from over, many practices continue to adopt and adapt these five principles. These elements will help primary care teams to thrive on the road ahead.