A Strategy for Identifying and Disseminating Best Practice Innovations in the Care of Patients with Multiple Chronic Conditions or End-of-Life Care Needs

Daphna Gans
Division of Geriatrics, Department of Medicine, David Geffen School of Medicine, University of California–Los Angeles
David A. Ganz
Division of Geriatrics, Department of Medicine, David Geffen School of Medicine, University of California–Los Angeles; VA Greater Los Angeles Healthcare System
Wendy Senelick
Division of Geriatrics, Department of Medicine, David Geffen School of Medicine, University of California–Los Angeles
Heather E. McCreath
Division of Geriatrics, Department of Medicine, David Geffen School of Medicine, University of California–Los Angeles
Jessica Jew
Division of Geriatrics, Department of Medicine, David Geffen School of Medicine, University of California–Los Angeles
Dan Osterweil
Division of Geriatrics, Department of Medicine, David Geffen School of Medicine, University of California–Los Angeles; SCAN Health Plan
Romilla A. Batra
SCAN Health Plan
Zaldy Tan
Division of Geriatrics, Department of Medicine, David Geffen School of Medicine, University of California–Los Angeles
Lee A. Jennings
Division of Geriatrics, Department of Medicine, David Geffen School of Medicine, University of California–Los Angeles
David B. Reuben
Division of Geriatrics, Department of Medicine, David Geffen School of Medicine, University of California–Los Angeles


Patients with multiple chronic conditions and those with end-of-life care needs experience high health care costs and needs for skilled coordination and well-trained staff. Focusing on these populations presents an opportunity to improve the patient experience toward the goal of more patient-centered care and reduced costs. Although innovative programs that provide better care to these patient populations have been developed, these innovations are often localized and not actively disseminated to other settings. This paper describes a quality-improvement project aimed at developing a process to identify best practices implemented in community-based clinical settings, develop a platform to share and disseminate these best practices, and facilitate the adoption of successful practices across other similar settings. The facilitation process involved structured coaching by clinicians and researchers experienced with practice change and quality improvement. The coaching component ensured that implementation teams receive guidance in the planning and adoption process, stay on track with implementation, and have access to timely support in addressing unanticipated barriers.

Key points

  • The SCAN Health Plan, a Medicare Advantage plan in California, and the UCLA Multicampus Program in Geriatric Medicine and Gerontology developed a system for identifying and disseminating best practices for taking care of older patients with multiple chronic conditions and for those facing end-of-life issues.
  • Five best practices were identified through a process that judged practices on how sustainable and scalable they were, along with several other factors.
  • Representatives from 12 medical groups attended a summit-like meeting to learn about the best practices and picked one of them to adopt.
  • Five of the medical groups participating in the summit received guidance from coaches afterward.
  • Fewer than half of the medical practices were able to implement the desired best practices.
  • The authors from SCAN Health Plan and UCLA argue that they have developed a process that can help “overcome the inertia and resource barriers to dissemination” but that additional incentives are needed for adoption of best practices.


Health care organizations continuously strive to achieve the “Triple Aim” of improving patients’ experience of care and their health outcomes while simultaneously lowering the costs of such care (Berwick 2008). Achieving the Triple Aim among older patients is especially challenging because of the number of patients with multiple chronic conditions and those with end-of-life care needs (Berwick 2008). High costs and the need for skilled coordination of care and provider training are associated with these two types of care.

More specifically, patients with multiple chronic conditions (e.g., ischemic heart disease, arthritis, diabetes, chronic kidney disease, dementia, and depression) often require care from multiple specialists and may experience a multitude of care transitions across different care settings. Without appropriate care management and care coordination, the care of this population not only is very costly but also may result in poor patient experience and patient outcomes (Hong 2015).

The documented high cost of attempted life-prolonging treatments in acute care settings is an issue in end-of-life care. Moreover, the care is often inconsistent with the individual’s and family’s goals and preferences and with the individual’s prognosis (Sabatino 2014). Discussions of end-of-life care preferences, standardized documentation of these preferences, accessibility of these documents across care settings, adherence to advance directives, and availability of palliative care and hospice referrals are all consistent with patient-centered care, yet they are not always common in practice (IOM 2014).

Successful innovative programs that provide better care to these patient populations with high needs and high costs are often localized and are not actively disseminated to other settings. In academic settings, there are incentives to publish descriptions of innovations that improve care but few incentives to disseminate the innovations into practice. Conversely, in community practices, there is little incentive to publish but often considerable incentive to disseminate within a health care system. Neither setting offers incentives to disseminate innovations to other health systems.

To address the gap between community practices and academic researchers, the SCAN Health Plan has partnered with the UCLA Multi­campus Program in Geriatric Medicine and Gerontology. Together, they have developed a quality improvement project to identify local innovations aimed at improving the care of older adults with multiple chronic conditions and those with end-of-life care needs, and to facilitate the diffusion of those practices across medical groups that partner with SCAN Health Plan. More specifically, the SCAN–UCLA project developed a process to identify the leading innovations in practice among SCAN-partnered medical groups, created a platform to share and disseminate these best practices, and facilitated the adoption of successful practices across other medical groups in the SCAN Health Plan network.


The SCAN Health Plan is a California-­based, not-for-profit Medicare Advantage plan that serves approximately 170,000 members. SCAN Health Plan began a provider integration initiative in which leaders of medical groups meet to discuss and address performance concerns. The idea of sharing innovations that were then implemented in community-based practices stemmed from these meetings and also led to this project and partnering with the UCLA Multicampus Program in Geriatric Medicine and Gerontology, a national leader in academic geriatrics.

The medical groups involved in this project were located in both northern and southern California. They are organized in a variety of ways and include independent practice associations (IPAs), staff models (i.e., physicians are employees of the medical group), and mixed models with some practices operating independently and others employing their providers. The medical groups ranged in level of expertise and experience in quality improvement and in inclination toward innovation; while some groups can be considered innovators and are willing to lead the field in trying new innovations, others range from early adopters to cautious late adopters that require progressively higher rigor of evidence before trying out a new innovation to laggards, groups that resist change (Rogers 1995).


To identify existing best practices addressing the two patient populations—those with multiple chronic conditions and those with end-of-life care needs in community settings—we used existing data on quality measures and surveyed medical groups.

Quality measures

To identify medical groups with high quality of care, we used scores for two consecutive years of the Centers for Medicare and Medicaid Services (CMS) Five-Star Quality Rating System for Medicare Advantage Plans. The star ratings are derived from four sources of data: (1) CMS administrative data on plan quality and member satisfaction; (2) the Consumer Assessment of Healthcare Providers and Systems (CAHPS) survey; (3) Healthcare Effectiveness Data and Information Set (HEDIS) measures; and (4) the Health Outcomes Survey (HOS). The Star ratings are scored 1, 2, 3, 4, and 5, representing poor, below average, average, above average, and excellent performance, respectively. These ratings are tied to financial incentives in the form of bonus payments (Jacobson 2011). We identified groups for which quality measures were uniformly high or quality measures had improved significantly, possibly indicating implementation of an effective innovation.

Survey and interviews

To identify specific best practices for patients with multiple chronic conditions and end-of-life care needs among the high-achieving groups, we collected data from key informants using focused surveys and interviews. Data collected included background information on the structure of the organization (medical group, IPA, management services organization), types of innovations used in treating the two focused types of populations, how the medical group delivered these innovations to patients, and measures used to assess the innovation outcomes. Ten of the 16 groups contacted provided data (Figure 1).

Figure 1
Total number of medical groups participating in the program

Selection of best practices

Once the 10 potential best practices were identified, the SCAN and UCLA team convened a judging panel composed of UCLA faculty coaches, SCAN coaches, and SCAN steering committee members. That panel selected the five most promising best practices. To evaluate the best practices fairly, a scoring matrix was developed based on several criteria with a rating system that scored unsatisfactory as 0, satisfactory as 1, and excellent as 2. The scoring matrix included six components, including the degree to which the innovation addressed an established need, its effects on achieving the Triple Aim, how well the best practice could be adapted to other settings (scalability), and how sustainable the best practice would be over the long term (Table 1) (Berwick 2008).

Table 1
Criteria for selecting best practices
Selection criteria Unsatisfactory
(0 points)
(1 point)
(2 points)
Need identified
How effectively does this best practice address an identified need within the medical group?
  1. Incomplete QI plan
  2. No or little data
  1. QI plan developed
  2. Some data collected
  1. QI plan developed, implemented, and staff trained
  2. Regular data collection
Quality of care is measurable (patient satisfaction)
Does the best practice measurably improve the quality of care received by the patient?
  1. No evidence of improvement
  2. No or little data
  1. Some evidence of improvement
  2. Some data collected
  1. Demonstrates meaningful impact
  2. Regular data collection using survey tools
Cost savings per patient is measurable
Does the best practice measurably lower the per capita cost of treating the patient?
  1. No evidence of cost savings
  2. No or little data
  1. Some evidence of cost savings
  2. Some data collected
  1. Demonstrates reduction in cost per patient
  2. Regular data collection
Demonstrates improvement in health
Does the best practice measurably improve target population health?
  1. No or low evidence
  2. No or little data
  1. Some evidence
  2. Some data collected
  1. Demonstrates meaningful impact
  2. Regular data collection
Easily adaptable to many settings
Is this best practice unique to the setting where it was developed, or can the principles be adopted in other care settings?
  1. No, too specific to one care setting
  1. Yes, but will require hiring or significant staff training
  1. Yes, easily adapted in other settings
  2. Use existing staff with ease
Is this intervention sustainable in the medical group?
  1. Demonstration project/pilot only
  1. Funding year-to-year
  1. Budget neutral or ongoing financial support
  2. Comprehensive staff training to sustain change

Of the five best practices, two innovations addressed care of patients with multiple chronic conditions: (1) enhanced care coordination for high-risk patients with multiple chronic conditions, and (2) a virtual interdisciplinary care team for patients with multiple chronic conditions.

The remaining three focused on caring for patients with end-of-life care needs, including (1) a nursing approach to introducing palliative care in a hospital setting, (2) outpatient palliative care using multidisciplinary home visits, and (3) advance care planning (ACP) for high-risk patients.

Disseminating best practices

A one-day summit was used as the platform for sharing the selected best practices. Ten of the 14 medical groups approached regarding best practices sent representatives to the meeting (Figure 1). Two additional groups were invited by SCAN, so a total of 12 groups were represented at the event and 75 individuals attended.

The objective of the summit was for each of the participating medical groups to leave the meeting with a decision on a specific innovation to adopt and a specific implementation plan. More specifically, the objectives included:

  • Identifying a specific evidence-based best practice to adopt in the area of care for patients with multiple chronic conditions and/or end-of-life care
  • Developing an action plan for adoption of the selected best practice with specific goals and timelines to achieve the goals
  • Outlining the potential barriers to achieving the action plan and identifying strategies to address and overcome these barriers
  • Identifying appropriate measures and data-collection plans for tracking progress and documenting outcomes

To achieve these objectives, the agenda for the summit began by introducing the concepts of best practices and quality improvement, followed by presentations of the best practices by each speaker and dedicated time for the medical groups to meet with quality improvement coaches from SCAN and UCLA to develop their implementation plan.

For each of the five best practices selected for presentation, a representative from the medical group initially presented a brief description of the innovation to the entire audience. Following these general presentations, the attendees selected three best practices to learn about in greater depth. A subsequent 15-minute in-depth presentation of each best practice innovation was conducted in a round-robin format; each innovation was presented three times, allowing attendees to attend three different presentations and ask questions of the presenters regarding implementation. After hearing the three presentations, participants from each attending organization picked one innovation to adopt.

Process and content coaching

Preparation for implementation of the chosen best practice was facilitated through process and content coaching. Trained researchers and clinicians with experience in implementing innovations and quality-improvement projects served as process coaches. Before the summit, the process coaches were given detailed information regarding the best practices and were asked to familiarize themselves with the innovations.

At the summit, each implementation team from an adopting medical group was assigned a specific process coach from SCAN or UCLA, who provided the team with an organized method of implementation using a structured action plan template (available upon request from the authors and on the Managed Care website as online-only supplementary material to this article, Appendix 1). This process coach was tasked with working with that team on implementation during the summit and for six months following the summit. A coaching manual (available upon request and as online supplementary material, Appendix 2) developed for the process coaches outlined coaches’ responsibilities prior to and during the summit and during the six-month coaching process following the summit.

During the summit, the participants also had access to content coaches. The presenters of the best practices were asked to serve as content coaches and contribute their expertise on the implementation of their respective best practices. The content coaches were available to any of the implementation teams to answer questions they had regarding the specific best practice while they were working on the detailed implementation plan. While the content coaches were mostly available to the implementation teams during the summit, some implementation teams reached out to the content coaches during the implementation process seeking more specific information.

During the summit, the process coaches spent two hours guiding implementation teams in the development of a detailed action plan to adopt their innovation. Action plans were designed to be dynamic documents to be adapted and updated as changes were made during the post-summit six-month coaching process.

The action plans included the following components:

  • A clearly articulated statement of the planned innovation, based on the implementation team’s selected best practice or components of a best practice
  • A statement of the overarching goal of the innovation and its intended outcomes
  • A description of the target population or patients affected by this innovation
  • Three Specific, Measurable, Achievable, Realistic/Results Oriented, and Time-Dated (SMART) Goals (CDC 2009) related to innovation adoption or implementation
  • An assessment plan including quality improvement methods to study the project
  • A measurement plan including quality-improvement tools to measure the project’s outcomes and achievement of SMART goals
  • A plan for achieving commitment from the organization’s leadership including identification of key stakeholders in the organization and a concrete plan of achieving their support
  • Resources needed to accomplish the plan
  • Anticipated barriers and obstacles to implementation
  • Strategies to overcome barriers
  • A time frame with specific tasks that need to take place in the six-month implementation period

Eleven of the 12 groups that attended the summit completed an action plan for a new best practice. One group chose to focus its efforts on further improvement of an innovation it had already developed. Two groups declined coaching because they had sufficient resources and in-house expertise in quality improvement, and the remaining four were not able to pursue the innovations for a variety of reasons.


The UCLA coaching model involved an intense postsummit coaching component, conducted through regularly scheduled 30-minute conference calls, usually twice a month, during the six months of the implementation process. The calls included key personnel from the implementation team, the assigned process coach, and the UCLA program manager.

Before each call, the implementation team completed a best practices adoption dashboard tool (available upon request and as online supplementary material, Appendix 3), which included prompts regarding progress on the SMART goals outlined in the action plan, barriers to completing the goals, and questions or requests for the coach. The tool served as a quick self-assessment for the implementation team, as a reminder of the planned call, and as a communication method between the coach and team to allow efficient work during the conference call. Eleven teams completed an action plan and five participated in coaching calls (Figure 1).

The post-summit process coaching supported the implementation teams in several ways. First, it provided a mechanism for keeping the teams on track and facilitating a timely and structured implementation. Second, the process coaches provided the teams with information regarding implementation details as needed. Finally, drawing on their experience in quality improvement projects, the process coaches provided in-time assistance in overcoming unanticipated barriers during implementation. In addition to the scheduled calls with process coaches, implementation teams had phone conversations and email exchanges with content coaches from the groups that had developed the innovations, as well as with other groups that chose the same innovation to adopt.

The process coaches met monthly on a conference call to review progress and to leverage their combined experience. The coaches’ call provided an opportunity for the coaches to support each other through the process and provide suggestions to one another on how to help their respective teams. A coach feedback tool (available upon request and as online supplementary material, Appendix 4) was developed for coaches to report on team progress, assessments of barriers, and ratings of medical group engagement and communication. Coaches completed the feedback tool monthly and the responses were used to guide the discussions about successes and progress, communication problems, structural challenges, and changes in selected practice innovations and level of engagement in coaching.


This paper documents a process designed to identify and disseminate successful innovations in the care of older adults with multiple chronic conditions or end-of-life care needs. One of the barriers to innovation dissemination is the competing time demands placed on medical groups’ leadership and the need to focus on pressing concerns such as performance measures and other operational and clinical concerns. Recognizing this barrier, SCAN Health Plan developed its provider integration initiative, in which leaders of the various groups meet periodically to discuss common concerns.

Building upon this collaborative culture, SCAN teamed up with UCLA to further encourage the groups to share and learn from one another with the ultimate goal of improving the quality of care provided for the two clinical populations of interest.

Disseminating successful innovations requires an awareness of the intervention (achieved through presentations at the summit) and following principles of decision adoption (e.g., relative advantage, compatibility, complexity, trialability, and observability) (Rogers 1995). The criteria we developed to judge each of the innovations (Table 1) helped select promising interventions that fit these principles.

A common barrier to dissemination is implementing an innovation in a different setting, which often requires adaptation and possibly additional resources. Moreover, despite best intentions, without guidance and support, it is sometimes difficult to carry out even a very detailed action plan because of competing demands on time, shifting organizational priorities, unexpected barriers, lack of resources, and a variety of factors. To overcome these barriers, we developed a structured system of support and guidance to assist teams with the adoption of these best practices by experienced coaches both during a summit-like meeting to introduce and explain best practices and during the six months following the adoption of the best practice. In our project, 5 of the 11 participating implementation teams chose to receive coaching support following the summit.

Dissemination of complex innovations is challenging. Even if the outcomes under “as usual” conditions are suboptimal, changing systems of care to make improvements may be more difficult than keeping the status quo. Few health care systems have the resources and “know how” to identify, adapt, and implement innovations that have worked in other systems. We have described a process that can help overcome the inertia and resource barriers to dissemination. However, it must be recognized that even with the substantial guidance provided by this system, fewer than half of the medical practices were able to implement desired innovations in their practices. Additional incentives will be needed. Once those incentives are in place, the process of selecting innovations and coaching through the implementation stages may be very valuable.


Berwick DM, Nolan TW, Whittington J. The triple aim: care, health, and cost. Health Aff (Millwood). 2008;27(3):759–769.

CDC (Centers for Disease Control and Prevention). Writing SMART objectives. Atlanta, GA: US Department of Health & Human Services. January 2009. Evaluation briefs no. 3b. http://www.cdc.gov/healthyyouth/evaluation/pdf/brief3b.pdf. Accessed June 13, 2016.

Hong CS, Hwang AS, Ferris TG. Finding a match: how successful complex care management programs identify patients [issue brief]. California Health Care Foundation. March 2015. http://www.chcf.org/publications/2015/03/finding-match-complex-care.Accessed June 13, 2016.

IOM (Institute of Medicine). Dying in America: improving quality and honoring individual preferences near the end of life [report brief]. September 2014. http://iom.nationalacademies.org/~/media/Files/Report%20Files/2014/EOL/Report%20Brief.pdf. Accessed June 13, 2016.

Jacobson G, Neuman T, Damico A, Huang J. Medicare Advantage plan star ratings and bonus payments in 2012 [data brief]. Henry Kaiser Family Foundation. November 2011. https://kaiserfamilyfoundation.files.wordpress.com/2013/01/8257.pdf. Accessed June 13, 2016.

Rogers EM. Diffusion of Innovations. 4th ed. New York, NY: Free Press; 1995.

Sabatino CP. Advance care planning tools that educate, engage, and empower. Public Policy Aging Rep (Gerontological Soc Am). 2014;24(3):107–111.