Improving Population Health Through an Innovative Collaborative

Key Findings

• The Be There San Diego Data for Quality (DFQ) Group created a safe venue for competing health care organizations throughout San Diego County to share performance data on quality-of-care measures, insights, lessons learned, and challenges faced by each organization.

• Unlike many regional data collaboratives, which focus on improving performance within individual organizations, the DFQ Group also placed strong emphasis on trends in combined data across organizations and collective action toward reducing cardiovascular risks throughout the region's population.

• By fostering collaboration among health care organizations that collectively serve a large portion of the local population and other key community stakeholders, including public health, the DFQ Group has helped form the foundation for a growing number of novel, innovative initiatives in San Diego.

In 2012, leaders from disparate, and sometimes competing, health care organizations joined together to establish a data group aligned around a regional goal of preventing heart attacks and strokes in San Diego. The group—now named the Be There San Diego (BTSD) Data for Quality (DFQ) Group—is a safe venue for medical directors and other quality improvement leaders to share performance data on quality-of-care measures for diabetes, hypertension, and cardiovascular disease, as well as share insights, lessons learned, and challenges faced by each organization in treating these conditions. Similar to other quality improvement collaboratives, the DFQ Group has focused its efforts on improving the quality of services provided by each participating health care organization. Importantly, the DFQ Group has also placed a strong emphasis on analyzing trends in combined quality data to better understand the health of the entire San Diego population. By fostering collaboration among organizations that collectively serve a large portion of the local population and other key community stakeholders, the DFQ Group has helped form the foundation of a unique, multifaceted, multi-stakeholder, regional effort that is gaining national attention and funding for its community­-driven approach. This brief report describes the origins and initial years (2011–2015) of the BTSD DFQ Group.

Origins of the Be There San Diego Data for Quality Group

In 2008, the California Department of Managed Health Care and the University of California, Berkeley, School of Public Health launched the Right Care Initiative (RCI), a novel public­-private partnership that sought to improve cardiovascular and diabetes care and outcomes statewide by promoting uptake of evidence-based practices or, when such practices were lacking, developing and testing new ones.¹ The RCI grew to include California health care industry leaders, providers, consumer groups, employers, the RAND Corporation, Stanford University, the University of California, Los Angeles (UCLA), the University of California, San Diego (UCSD), and other key stakeholders throughout the state. In 2010, leaders from the RCI effort began discussions with clinical and community leaders in San Diego about the possibility of the county serving as a regional demonstration site for the RCI. Based on clinical performance data, the envisioned effort would focus not only on the spread of best practices within health care settings throughout San Diego, but also on extending these best practices beyond the clinic walls, emphasizing cooperative approaches among stakeholders across the region.

By late 2010, the San Diego community agreed to adopt the RCI model and to put theory into practice. As part of this effort, the San Diego University of Best Practices (UBP) was established in early 2011 as a monthly forum for health care leaders to learn about the latest research, guidelines, and best clinical practices related to preventing heart attacks and strokes. The UBP, a component of the BTSD project, provides a specific focus on application into real-world clinical and community practice. The UBP and related activities in San Diego were partly supported by funds from a Grand Opportunities grant from the National Heart, Lung, and Blood Institute (NHLBI).² Additional funds came from local philanthropists Jack and Judith White, whose support helped catalyze the goal of preventing heart attacks and strokes in San Diego.

To support the progress needed to meet this goal, clinical leaders participating in the UBP sought a vehicle to compare quality-of-care measures within and across health care organizations and explore population health trends over time. Given real and perceived restrictions on sharing data, San Diego leaders asked the RAND Corporation to serve as a neutral organization that would receive and analyze the data and help facilitate meetings between the various health care organizations participating in data sharing. Thus, the San Diego RCI Data Group, a component of the broader San Diego RCI, was formed. The Data Group collectively developed a set of agreements that enabled health care organizations to share aggregated performance data on a selected set of intermediate outcome measures for hypertension, diabetes, and cardiovascular disease care. The agreements also included mutual nondisclosure agreements that made it possible for the health care organizations to share more-detailed information in Data Group meetings. The Data Group's activities officially began in early 2012, when agreements were signed and data sharing began. Having an established time and private meeting space exclusively for those health care organizations contributing data helped strengthen the early growth of the Data Group, allowing medical directors to develop a significant degree of trust and collegiality that would prove essential in laying the foundation for enhanced data sharing in the future.

When the NHLBI grant period ended in mid-July 2012, much of the Data Group's participation and activities were undertaken on a voluntary basis, with in-kind donations by participating individuals and organizations. In late 2014, after local clinical leaders took full ownership of the efforts, RCI San Diego was renamed Be There San Diego, with the Data Group rebranded as the BTSD Data Group. With funding in 2015 from the Centers for Disease Control and Prevention (CDC) to the County of San Diego to support analytic infrastructure and technical support, the BTSD Data Group renamed itself the BTSD Data for Quality Group.

Members of the Be There San Diego Data for Quality Group

The BTSD DFQ Group includes clinical leaders representing eight health care organizations in the county of San Diego:

• Arch Health Partners

• Council of Community Clinics (network of federally qualified health centers)

• Kaiser Permanente

• MultiCultural Primary Care Medical Group

• Scripps Clinic Medical Group

• Scripps Coastal Medical Center

• Sharp Rees-Stealy

• University of California San Diego Medical Group.

It also includes subject-matter experts in data analysis and clinical quality improvement and a core administrative staff. Several of the participating health care organizations have been recognized for their performance and innovation, routinely winning both national awards and awards from the California Department of Managed Health Care at annual RCI meetings.

The clinical leaders participating in the DFQ Group represent diverse health care organizations in San Diego, including organizations within integrated delivery systems, multispecialty medical groups, community health centers, an academic medical center, and a primary care independent practice association, which presented some challenges early on that needed to be overcome. In addition, many of the participating health care organizations are direct competitors in the San Diego marketplace, which led to initial skepticism regarding the value of sharing data and clinical practices. Furthermore, health care organizations differed by patient populations and organizational characteristics. For example, several of the groups are large, integrated, multispecialty medical groups with well-developed electronic health record and population health management analytic tools, while others are small or moderate-sized groups with relatively limited quality data tracking tools and systems. Even where the size of the overall population managed is similar, the organizations vary in their respective patient mixes and in the types of insurance they accept. In particular, while the majority of the organizations provide care primarily to commercially insured and Medicare patients, one large organization composed of several community health centers primarily serves Medi-Cal and uninsured patients.

Despite these challenges, clinical leaders' commitment to a shared vision of working together to address challenges related to heart attacks and strokes at the population health level moved the group forward. These motivated clinical leaders were instrumental in establishing the DFQ Group, and their continued support of the group's mission, willingness to share data, and meaningful conversations allowed the DFQ Group to thrive. These leaders illustrate that finding the right clinical champion from each participating health care organization—someone who is open to collaboration and supportive of improving regional population health—is essential to continuing the DFQ Group's collaborative environment.

In addition to clinical leaders from health care organizations, membership in the DFQ Group includes subject-matter experts in data analysis and clinical quality improvement and a core administrative staff. RAND took the lead in managing the shared data and examining community health trends, serving as the neutral entity and data broker to accept and process the data shared by each health care organization and facilitating group discussions and activities. With the group focused on identifying actionable steps, members with expertise in clinical quality improvement and population health helped to guide conversations around goal setting and changing care processes to improve population health. Two physicians with expertise in data analysis and clinical quality improvement chair the DFQ Group with support from the chair of the BTSD Executive Committee, who serves as a neutral leader and promotes discussion and action toward achieving the goal of preventing heart attacks and strokes in San Diego. Finally, the DFQ Group benefits from an administrative staff that organizes monthly meetings, recruits participants, and works to achieve the goals of BTSD. With impassioned clinical leaders, analytic and quality improvement expertise, and a collaborative spirit, the DFQ Group illustrates how diverse health care organizations can come together to promote regional population health.

Collecting Key Data on San Diego Residents

By collecting meaningful, actionable data from the major health care organizations serving San Diego, the DFQ Group has sought to drive positive changes in population health. Participating health care organizations shared aggregate data on several intermediate outcome measures related to care of patients with hypertension, diabetes mellitus, and cardiovascular disease (CVD) (see Table 1). Many of these measures were adapted from Healthcare Effectiveness Data and Information Set (HEDIS) measures and were already routinely being reported by most health care organizations to the Integrated Healthcare Association (IHA), California's value-based pay-for-performance program, and to the Health Resources and Services Administration's Uniform Data System. Several of these measures were also tracked by the statewide RCI with the goal of having health care organizations achieve a national 90th-percentile ranking for these measures. While some stewards of quality measures have revised their recommendations for measures over time (for example, both HEDIS and IHA no longer endorse measuring low-density lipoprotein [LDL] control for patients with diabetes), the DFG Group has continued to collect all of the below measures while focusing its analysis and discussions on measures of blood pressure (BP) control for patients with hypertension and diabetes.

As of 2014, all health care organizations were reporting rates of BP control for hypertension and diabetes, including the number of patients with each condition and the number of patients with each condition under control. Reporting of CVD measures was less common because of the lower prevalence of this condition. Figures 1 and 2 illustrate the journey of the DFQ Group as the number and size of health care organizations sharing data grew steadily since 2012. As illustrated by Figure 1, the number of patients with hypertension included in the shared data was more than 2.5 times greater in 2014 than in 2010, as the number of health care organizations sharing data increased from two to eight. Similarly, the number of patients with diabetes included in the shared data more than tripled since 2010 (Figure 2). In 2014, the shared data included over 176,000 patients with hypertension and 86,621 patients with diabetes. Compared to California survey data, which provides an estimate of self-reported prevalence of these conditions, the shared data represent 26 percent and 53 percent of individuals in San Diego with hypertension and diabetes in 2014, respectively. However, as the number and diversity of health care organizations participating in the DFQ Group grow and their reporting capabilities evolve to track patients of varying insurance types and statuses, these data are becoming more representative of the overall county.

Changes in the size and composition of the combined data set over time have made longitudinal comparisons of quality-of-care measures challenging. However, data sharing has been stable and has included a large number of organizations since 2013, making more-recent comparisons of measures possible. From 2013 to 2014, the weighted average rate of BP control for patients with hypertension improved from 72 percent to 74 percent and the weighted average rate of BP control for patients with diabetes remained unchanged at 71 percent.

Finally, we note that health care organizations' contributions shared within the privacy of the DFQ Group have gone beyond the sharing of aggregate data. For example, during 2012 and 2013 the health care organizations rotated responsibility for presentations about their respective organizations' quality-improvement strategies and trends in performance on different quality measures, routinely presenting detailed results and examples at the clinic or provider level. During these early years, the DFQ Group was able to see a wide range of ways that health care organizations used their data to guide quality-improvement efforts.

Early Successes of the Data for Quality Group

In contrast to regional data collaboratives whose members often use such forums solely as a means to improve their respective quality performance, the BTSD DFQ Group's aim from the outset was to create an ongoing learning laboratory that seeks to understand and address challenges related to heart attacks and strokes at the population health level. A key achievement of the DFQ Group has been the development and adoption of a standardized regional approach to treating patients with hypertension and those at risk of heart attack or stroke. Specifically, participating health care organizations have

• agreed to use a simplified guideline algorithm for treating patients with hypertension

• promoted use of the American College of Cardiology and American Heart Association atherosclerotic CVD risk estimator

• disseminated an evidence-based medication protocol for prescribing a bundle of medications to prevent heart attacks for at-risk populations.

In doing so, the DFQ Group has demonstrated its utility as an action-oriented group focused on changing care processes to improve population health.

The DFQ Group, which built trust and fostered a collaborative spirit among participating health care organizations, facilitated BTSD winning a Health Care Innovation Award from the Center for Medicare and Medicaid Innovation (CMMI) in 2014. Several health care organizations participating in the DFQ Group are engaged in BTSD's CMMI award, which involves a patient-level health coaching intervention and prescription of an evidence-based bundle of medications to prevent heart attacks and strokes. Additionally, a CDC grant to the County of San Diego provided funding to the DFQ Group in 2015 to enhance data-sharing activities and analyses in order to guide ongoing discussions regarding quality-improvement work and help to set community-wide targets for hypertension control.

Next Steps for the Be There San Diego Data for Quality Group

In 2015, health care organizations participating in the BTSD DFQ Group began sharing more information about their patient populations to guide ongoing discussions regarding quality-improvement work and help to set community-wide goals for controlling hypertension and diabetes. Specifically, health care organizations began reporting previously reported intermediate outcome measures by age group, sex, payer, and, for select measures, ZIP code. These details will allow the DFQ Group to examine variation in the prevalence of hypertension, diabetes, and cardiovascular disease between men and women and between different age groups across San Diego. Additionally, collecting quality measures at the level of patients' ZIP codes will enable the DFQ Group to explore geographic trends, identify “hot spots” of poor health outcomes, and incorporate salient community characteristics, such as poverty rates and the distribution of historically underserved racial and ethnic groups. This will allow stakeholders to better understand how social determinants affect the health of San Diegans and to identify strategies for how clinicians and the community can work together to prevent heart attacks and strokes for all San Diegans.

While many DFQ Group discussions and actions have focused on patients with hypertension, due to the high prevalence of hypertension in San Diego and across the county, the DFQ Group is also invested in reducing the burden of diabetes. Utilizing the rich data collected in 2015 and beyond, the DFQ Group will explore trends with the goal of controlling BP, HbA1c, and LDL for patients with diabetes by age group, sex, and payer, and discuss successful strategies for treating patients with diabetes and those at risk of diabetes.

Furthermore, with the overall goal of improving regional population health, the DFQ Group is actively using public health data sources to measure health outcomes in addition to traditional intermediate outcome measures reported by the participating health care organizations. By examining community trends in hospitalizations, emergency department visits, and deaths due to heart attack and stroke, the DFQ Group can examine whether improvements in control of BP and HbA1c have led to overall improvements in health outcomes. Use of these and other state and national data sources will also enable the group to explore how health outcomes in San Diego compare to other regions. Moving forward, the DFQ Group will continue its effort to identify proven strategies for promoting health and collectively implement these strategies to reduce gaps in quality of care and increase health and wellness for all of San Diego County.