Natural History of Practice Transformation: Development and Initial Testing of an Outcomes-Based Model

Setting

This report examined the first group of practices in the NC IPIP. As described elsewhere,¹⁸ NC IPIP was led by the American Board of Medical Specialties, cosponsored by the certifying boards and specialty societies of internal medicine, family medicine, and pediatrics, and funded by Robert Wood Johnson Foundation in 2005. During the past 5 years, North Carolina Area Health Education Centers (NC AHEC), in partnership with Community Care of North Carolina, the NC Academy of Family Physicians, the NC Pediatric Society, the NC Division of Public Health, and the major insurers in the state, developed a comprehensive statewide model of community-based practice-level support for QI. Key elements of NC IPIP included statewide common quality measures across insurers, a system for collecting and feeding back quality data to practices, regional quarterly collaborative dinner meetings modeled on the Institute for Healthcare Improvement Breakthrough Series model,¹ community-based practice facilitators, labeled QI consultants, modest financial support to defray direct costs of participation, and use of continuing medical education and credit for Maintenance of Certification part IV as incentives.

Practices and Data Systems

The first phase of NC IPIP included 18 primary care practices (family practice, pediatrics, general internal medicine) located in the Mountain and Eastern regions of North Carolina (Table 1). The majority focused on diabetes (13 practices) and used an electronic health record (16 practices). The practices were mixed among private (39%), public (34%), and academic affiliated (27%).

Table 2 describes the clinical and practice measures available. Practices could focus on diabetes or asthma quality clinical measures. These data were stored centrally and made available to the practice via a collaborative website. QI consultants worked with practices to develop customized interventions, with the overall goal of improving measured quality for the whole population of the practice. Emphasis was placed on development of registry functions, standing orders, support for self-management, and fidelity to the program goals, but QI consultants would also work with practices on any practice clinical or business issues the practice identified. In addition to quality data, 2 QI consultants took concurrent notes about the practice goals, leadership, and progress in implementing office systems changes. Finally, QI consultants wrote narrative descriptions of the practice’s progress in improving office systems. These data are the basis of this study.

Model Development

All authors developed and refined the model during multiple meetings over the course of a year. The authors bring extensive experience with practice redesign and transformation various perspectives to the IPIP: one author (A.L.) led the initial NC QI Organization statewide outreach effort to promote electronic health records in private practices (http://www.thecarolinascenter.org/) and was the original QI consultant in all the IPIP practices and trained the subsequent QI consultants; another (M.P.) brought quality collaborative expertise as a pioneer of the original chronic disease collaboratives in the state⁵; a third author (W.P.N.) led the NC IPIP project¹⁸ and 2 successful collaboratives of residency practice redesign in the southeast region,^20,21 and has been extensively involved at the national level with practice redesign with national organizations; and the fourth author (K.E.D.) brought expertise of leading prevention collaboratives in the state, as well as participating in the evaluation of the state health department collaboratives and the IPIP.

During the first phase, the natural history and key influences on practice transformation were identified by drawing from the Chronic Care Model.²² The model was then shared and revised during meetings with national IPIP team members and state leaders who had experience running collaboratives and in monthly discussions with QI consultants working with NC IPIP practices. In the second phase, the authors individually reviewed each practice’s monthly diabetes or asthma quality data, concurrent team assessments from the QI consultant, and the previously written practice story generated by the lead practice QI consultant (A.L.). The authors deliberated findings as a group and tested the model. The group identified 3 emerging patterns and common themes of success and barriers to transformation among the practices and further refined the model.

The Biomedical Institutional Review Board at the University of North Carolina determined that this study was exempt from the Institutional Review Board approval.

Model for the Natural History of Transformation

Our model for the natural history of transformation is depicted in Figure 1. The initiation phase is driven by both internal and external motivators. Internal motivators include improved reimbursement, increased office efficiencies, better patient outcomes, help meeting the requirements of a federally qualified health center, and often simply “to do the right thing” for the patients. External influences come from such areas as institutional leadership, professional organizations (certifying boards), payer programs (pay-for-performance), and local pressures (practice comparisons). The IPIP program can be understood as influencing the initiation phase in a number of ways, including community meetings of peers supporting transformation and provision of continuing medical education or maintenance of certification part IV credit.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Framework for sustainable practice transformation.

CME = continuing medical education; IT = information technology; MOC = maintenance of certification; QI = quality improvement.

The intermediate phase included implementing initial system changes, such as a disease registry (stand-alone or within an electronic health record), templates to guide care, disease-based protocols for the entire treatment team (ie, standing orders), and an effective patient self-management support program. As shown in the model, these system changes help a practice work toward a paradigm change that includes organizational redesign throughout the practice, development of a culture that supports using data to observe the patient population, and a team-based care approach with an emphasis on the patient’s self-management of their disease. The model suggests that familiarity with data systems and early successes lead practices to belief and trust in the process and their data, which sustains the practices to the advanced phase and allows for a true practice transformation that sustains the improvements gained in both process and clinical outcomes. In this framework, NC IPIP and other similar programs can support practice transformation by framing organizational learning through data review and bringing learning about best practices from other sites.

Examples

By examining each of the practices in the NC IPIP program through the use of this model of sustainable practice transformation (Figure 1), participating practices separated into 3 emerging categories: practices experiencing robust improvement over a sustained period of time, practices with moderate change, and practices with minimal change over a period of time. Table 3 describes our practices categorized as engaged, activated, and transformed. Figure 2 gives a pictorial representation of each of these practice types over time.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Representation of level of quality over time for transformed practices, activated practices, and engaged practices over 2 years.

Transformed practices separated into 2 subgroups, both with substantial broad-based improvement, but one group reached a threshold (subgroup A) and the other continuing to evolve (subgroup B). A key environmental influence for transformed practices was highly engaged organizational leadership. For example, the leadership in one practice showed their support of QI by implementing a peer-review program throughout the organization, where clinicians reviewed and discussed each other’s charts as a group as part of the QI initiative. Examples of internal motivatorsin robust practices included the organizational need to attain federally qualified health center status, which required a QI program. Key practice support elements included active use of the QI consultant on a regular basis for all QI activities. Several transformed practices took advantage of the credit offered for continuing medical education and maintenance of certification. Large multidisciplinary practice teams attended the peer networking meetings and shared regularly at these meetings. The change paradigm for transformed practices included beginning to use their data to drive decision making and change the culture of the organizations. Several robust practices shared physician-level data at all meetings and displayed outcomes data in a public area for patient viewing. For example, practices used data to decide on the specific office systems for redesign and then used the data to monitor their reliability; one practice even planned and measured an intervention to decrease the number of geese crowding the parking lot. Transformed practices also began to include QI activities in job descriptions and compensation plans for staff, thus institutionalizing the practice transformation.

Activated practices included practices that improved in at least one measure with a slower rate of improvement. These practices had a slower trajectory, but eventually reached goals (Figure 2). As with transformed practices, an important environmental influence was leadership supportive of QI work. Internal motivators included physicians on staff who were QI experts. Practice support included the use of the QI consultant in some practice meetings. Compared with transformed practices, however, only the lead physician in several of the activated practices attended all peer networking meetings and shared perspectives. In regard to paradigm change, clinical data were incorporated and shared in large QI meetings, and there was some systems redesign, but the impact on or interaction with the staff was minimal. In all activated practices, there were key systems that limited the abilities to perform. For example, implementation of an electronic health record was a common obstacle; compared with transformed practices, activated practices were less able to problem solve issues in this area. Variability in staff engagement was also more common, in part as a result of physician and staff turnover. Another kind of challenge was involvement in too many practice redesign projects, which limited performance in one.

Engaged practices either did not improve in any clinical measure over the first year or improved but did not sustain improvement (Figure 2); as a result, there was no paradigm change during the first year. Despite an initial interest and commitment to the program, the practices were “warming up” and not ready to fully incorporate a sustainable QI program. Causes included staff turnover, other practice priorities, or fiscal stresses. An engaged practice could have positive motivations. For example, in one practice the leadership cited their conviction that QI was the right thing to do for their patients, their professional society’s involvement in the program, and their priority for achieving maintenance of certification part IV requirements. This practice was limited, however, in their capacity to make practice changes (eg, the leadership’s vision was not shared throughout practice). There were too many distractions in key areas, ranging from business-related issues (keeping the lights on) to the lack of trust in the team care concept (physicians less willing to engage staff in tasks, staff less willing to do things outside a current job description). In regard to practice support, only the lead physician engaged with the QI consultant and usually on an infrequent basis. At this point, the practice was seeking more consultation for practice management and for electronic health records than for QI.

Comparative Models of Practice Transformation

Our model is similar in some respects to that proposed by Cohen et al.²⁴ Like that group, we emphasize the role of motivation of key stakeholders, resources for change, external motivators, and the interaction of these influences over time. We emphasize, however, the critical importance of early experiences. As described by Edmondson and others,^{16,17,25–27} organizations must learn how to improve care, and a key mechanism is early success. Many have commented on the importance of leadership and adaptive reserve.^13,14,28 Our model suggests that early success leads to what might be thought of self-efficacy, in that the practice can change in response to data and urgent concerns—this attribute is a key ingredient of adaptive reserve. Practices are learning organizations, and they must learn how to learn and learn how to manage change. Physician leadership is critical in this realm.

The external environment is also critical. As emphasized by the future of family medicine report,²⁹ the environment for primary care is harsh. Over the course of the first 2 years of our study, 3 of our 18 practices reorganized or changed ownership, suggesting the strength of the headwinds facing primary care in our current environment. The community environment can also promote positive change. Our study takes place in the context of a large-scale project to transform practices, in which substantial efforts have been made to provide external support for practice transformation. In addition to QI consultants, practices were given consistent quality measures across insurers, data support, and frequent feedback; furthermore, regional collaboratives provided peer support along with specialty society and certifying board support. Most important are the data and feedback, which played a critical and simplifying role. Relentless focus on data allows a practice to learn and organizes the culture of the practices. Our experiences suggest that, as we seek to transform health care, ongoing community level support and infrastructure are essential.

Our results raise a number of important questions about the natural history of practice transformation. A first issue is our understanding of practice transformation. For some writers, practice transformation seems almost like a religious conversion experience: once it happens, the practice continues to improve along all dimensions continuously and forever. Our best performing practices fell into 2 groups. Indeed, some do seem to be improving quality continuously across all fronts. Others, however, reach a threshold and then move on to other diseases or to other aspects of care or office systems. Although continual improvement across all dimensions is an attractive ideal, primary care must manage many chronic diseases, as well as other dimensions of care, including access, prevention, and patient experience. Reaching a threshold, such as the National Committee for Quality Assurance threshold for a specific disease and measure, and then moving on to other problems may be a more realistic goal for most practices. What we emphasize here, however, is that how we conceptualize transformation has major implications for how we promote and sustain practice transformation through public policy and community interventions.

Policy Questions in Practice Transformation

Public policy makers often imply that practice redesign and systems change are plug and play, a tangible endpoint that can be scheduled precisely. From our data, however, it is clear that practice redesign takes a long time: these practices are 2 years out and counting. Contemporary practices face limited resources and many substantial challenges to change. Other studies of practice redesign have had similar results.³⁰ We also do not know which of our 2 groups—the activated or the transformed—will achieve the most improvement over the long term (eg, 10 years). Will the tortoise or the hare win the race? Finally, although the warming-up practices with limited or no quality improvement may represent those without the leadership, office systems, or incentives to change, they may in fact also be doing important work that is very necessary but has not yet resulted in outcomes improvement.

Role of Patients and Patient Experience

Our experience so far is that patients’ perspectives do not weigh in appreciably as a motivator or external influence on the initiation or the growth of practice transformation. One reason is because our project initially focused on chronic disease, which often does not address patient satisfaction as directly as other practice transformation interventions, such as advanced access. This focus may change with the further evolution of patient-centered medical homes, the more widespread use of measurement of patient satisfaction and patient experience, and the development of patient advisory boards.

Practice transformation is not an overnight sensation; rather, it is a medium-to long-term process that will vary by practice capacity, leadership, community context, and support. Our focus moving forward is now on improving each critical phase: initiation, early success, and then maintenance, even as the dimensions of practice transformation broaden to include patient experience, transitions of care, and cost of care.