Avoid Readmissions through Collaboration

September 2010 – July 2012

Overview The Avoid Readmissions through Collaboration (ARC) program seeks to bring together hospitals and their partners in the community to prevent readmissions. ARC is a partnership between Cynosure, CQC, and funded by the Gordon and Betty Moore Foundation. Readmission rates in California are not much better than national norms. Our vision is to reduce 30 and 90 day readmission rates by 30% by 2013.

Structure Quarterly on-site Learning Sessions in Oakland started as a forum for hospitals and their partners (e.g. medical groups/IPAs, home health, long term care, and health plans) to understand existing evidence-based models to reduce admissions. The program led participants through a self-assessment leading to an action plan tailored for their own institution. Each Learning Session fostered exchange among program participants and exposure to new ideas through national experts. Action Network participants received more intensive support to reach their goals including coaching and access to planning grants.

Measurement The hospitals in the Action Network submit data monthly, including the following measures:

1. 30-Day All-Cause Readmissions Rate 2. 90-Day All-Cause Readmissions Rate 3. HCAHPS Patient Survey: Did hospital staff talk with you about whether you would have the

help you needed when you left the hospital? 4. Did you get information in writing about what symptoms or health problems to look out for

after you left the hospital?

Optional Measures 30-Day All-Cause Readmissions for a specific clinical condition or subpopulation (e.g. chronic conditions like CHF, COPD, psychiatric, or subgroups like Obstetrics or pediatrics)

Hospitals & partners interested in sharing their data and their improvement work with peers.

Any organization, including hospitals & partners interested in learning more about readmission reduction.

Learning Community 50 Hospitals

Action Network 25 Hospitals

rev 3/12

1

Spread During the collaborative, participants became presenters and teachers. Program participants became champions within their facility for spreading change, and some participating hospitals became champions within their communities to improve care transitions. Some examples include: • Marin General Hospital – The facility is playing a key leadership role in the county’s efforts to

better manage care. Work groups are underway to support patients based on risk of readmission: Lowest risk – ProjectRED; Moderate risk – Care Transitions Intervention (CTI); and Highest risk - on going case management and community support.

• ARC Learning Sessions featured presentations from participants such as San Francisco General Hospital, Lodi Memorial Hospital, Palo Alto Medical Foundation, Anthem Blue Cross, and Sutter Care at Home.

Quotes All the results from the collaborative are not quantifiable. Below are some examples of what we’ve heard. • “We had a lot of people doing a lot of things, but individually. ARC helped us focus on

specific goals.” – El Camino team member • “You can read about all the models, but ARC demonstrated how to make it real – in a

community hospital or teaching hospital. It gives us the courage to do it ourselves”. – Marin General Hospital team member

• “Where else can you get this kind of education? I learn something new every time.” – Washington Hospital team member

Results The 20 hospitals reporting through the Action Network have demonstrated improvement:

2

Action Network: • Alameda County Medical Center • Antelope Valley Hospital • Chinese Hospital • El Camino Hospital • Lodi Memorial Hospital • Marin General Hospital • MemorialCare Health System • Mercy General Hospital • O'Connor Hospital • Queen of the Valley • St. Francis Medical Center • Saint Francis Memorial Hospital • San Francisco General Hospital Medical

Center

• San Joaquin Community Hospital • Ramon Regional Medical Center • Santa Clara Valley Medical Center • St. Rose Hospital • UCSF Medical Center • Ukiah Valley Medical Center • University of California Irvine • VA Medical Center San Francisco • VA Palo Alto Health Care System • ValleyCare Health System • Washington Hospital Healthcare System • Woodland Memorial Hospital

Participants Learning Community: • Alta Bates Medical Center • Community Hospital of the Monterey

Peninsula • Desert Regional Medical Center • Eden Medical Center / San Leandro Hospital • Healdsburg District Hospital • Henry Mayo Newhall Memorial Hospital • Kaiser Permanente Redwood City Medical

Center • Little Company of Mary - San Pedro • Little Company of Mary - Torrance • Mee Memorial Hospital • Providence – Holy Cross • Providence – St. Joseph

• Providence – Tarzana • Rancho Los Amigos National Rehabilitation

Center • Riverside County Regional Medical Center • Saddleback Memorial Medical Center • Salinas Valley Memorial Healthcare • Santa Rosa Memorial Hospital • Sharp HealthCare • Sierra Nevada Memorial Hospital • Sierra View District Hospital • St. Joseph Orange • St. Jude Medical Center • Sutter Delta • Sutter Gould Medical Foundation • UC Davis Medical Center

3

Chronic Care Collaborative February 2011 – February 2012

Overview The Chronic Care Collaborative set the goal to assist physician groups and independent practice associations to measurably improve measures of diabetes and heart disease for patients living in areas of California where data shows the poorest outcomes. Participating organizations, caring for 250,000 patients, set a goal to close the gap between current performance on publicly reported measures and the 90th percentile by 20% within one year.

Structure Collaborative participants from five physician groups met at four on-site meetings to implement multiple changes based on the Chronic Care Model. In between on-site meetings, CQC held webinars on related topics, conducted regular, frequent coaching calls, and two on-site visits to each provider group.

Measurement Each team reported data quarterly on three core measures and selected up to four optional measures for their patients. Core measures collected during the collaborative were: 1. HbA1c Testing – Percent of diabetics with one test in the last twelve months 2. HbA1c Control – Percent of diabetics with HbA1c > 9.0 3. LDL Testing - Percent of diabetics with one test in the last twelve months

Optional measures included: 1. HbA1c Control – Percent of diabetics with HbA1c < 8.0 2. LDL Control in Diabetes – Percent of diabetics with LDL-c < 100 3. Nephropathy Monitoring – Percent of diabetics with evidence of, or monitoring for,

nephropathy 4. Blood Pressure Control – Percent of diabetics with BP < 140/80

Participants • Bakersfield Family Medical Center (BFMC) • Desert Oasis Healthcare (DOHC) • EPIC Management, Beaver Medical Group (EPIC) • NAMM California/Mercy Physicians Medical Group (NAMM) • United Family Care (UFC)

rev 3/12

4

Interventions Highlights of the interventions were: • Groups adapted information systems and refined reports to physician practices to make patient

information more usable at the point of care and to give physicians feedback on practice performance.

• Practices made changes to use registries and lists to regularly recall patients with gaps in care or labs outside of recommended ranges.

• Groups provided incentives and public recognition for improved performance and shared best practices across the network.

• Groups provided clinical support through health educators and clinical pharmacists and provided helpful tools to practitioners to educate patients at the point of care.

Results Summary data for all groups show that six of the seven measures improved by at least 20%, all except HbA1c < 8.0. Individually, every group improved at least one measure by at least 20% in twelve months and one group improved six of seven.

5

Spread During the collaborative, participants became presenters and teachers and acquired communication skills they found valuable in presenting data to practices. Participating physicians who became champions within their groups spread changes and shared better and best practices to achieve high performance. Groups were able to pilot test and spread changes to all or most of each network by the end of 2012. They will continue to measure and monitor performance and have learned skills that can be applied to other improvement initiatives.

Lessons Learned • Active participation in the collaborative provided motivation to pilot test, measure, and

spread changes.

• A functioning registry is fundamental to improvement as it enables participants to measure performance regularly, assess the effectiveness of interventions, and to identify patients needing services.

• A key to success is for team members to meet regularly both at the group and practice level and to communicate progress and results frequently.

• Competing priorities, such as the implementation of electronic health records, prevented spread from occurring as quickly as hoped. Groups recognized this reality and included active management of competing priorities into their work plan.

• Some solo/small practices may not always be disadvantaged. They can implement changes quickly and delegate tasks to other practice staff.

• Development of a strong lead with excellent project management skills is critical to success both at the group and practice level.

• Use of evidenced based quality improvement models such as the Model for Change and PDSA cycles led to improved outcomes.

• Data accuracy is critically important to provider outreach efforts.

• Setting realistic goals based on the availability of organizational resources was critical to success.

• Persistence and follow through seemed to be more critical to success than great ideas.

• Well designed, thoughtfully implemented small financial incentives and non-financial incentives were very helpful in changing behavior.

• Engaging senior leadership was important to realize organizational change.

6

Complex Care Management Action Community

January – November 2011

Participants Eight organizations participated in the Action Community:

• Bristol Park Medical Group • CalOptima • Central California Alliance for Health • HealthCare Partners

• High Desert Medical Group • Humboldt Del Norte IPA/Foundation • Prime Care • St. Joseph Heritage Medical Group

Overview CQC’s Complex Care Management Action Community, funded by the California HealthCare Foundation, brought together eight leading California organizations that are working to redesign and improve care for their “complex” patients, those with multiple chronic conditions, limited functional status, and psychosocial needs, that account for a disproportionate share of health care costs and utilization. While much is known about how to care for patients with single chronic conditions, such as diabetes or heart disease, less is known about how to manage high-cost patients with multiple and complex chronic conditions. The purpose of the Action Community was two-fold: 1) To facilitate peer learning among the organizations so that they could learn about, share, and test improvements in care that lead to better health outcomes at lower cost; and 2) To develop a toolkit based on the learning from the Action Community that CQC will then spread to other organizations in California. The eight organizations participating in the Complex Care Management Action Community implemented changes to improve the care they provide to their sickest patients. Common changes across the eight groups included: Although the program concluded at the end of 2011, the lessons learned from these organizations will live on through a toolkit that is currently being developed to summarize the key changes to test in complex care management and the major resources in the field. CQC will broadly disseminate this tool once it is available.

• Strengthening approaches to risk stratification of patients

• Developing more robust measurement strategies

• Improving care transitions • Tightening care coordination through

improved communication • Redesigning care team models

Quotes “It’s easy when you’re confronted with so many problems to move around. The nature of the job is that everyone needs something done now; it’s easy to be frantic and less intentional. To have someone tutor you through a vision helps…the Action Community helped to keep us moving forward with a plan, and to be less reactive, more proactive and intentional.” - Maria De Lima, MD, High Desert Medical Group

1 rev 3/12

Inland Quality Collaborative April 2007– February 2012

Overview The Inland Quality Collaborative was launched after publicly reported measures of clinical care and patient experience showed lowest performance in Riverside and San Bernardino Counties – the Inland Empire. Physician groups, health plans serving the region, and the California Association of Physician Groups (CAPG) came together to invest in learning how to accelerate improvement in patient care for the 1.5 million commercial patients living in the region.

Structure • Created a local Learning Network: Quarterly

on-site sessions for physician groups, health plans, and community agencies in the region to foster peer-to-peer exchange of ideas and approaches to improve patient care.

• Sponsored one-day training sessions on aspects of quality improvement.

• Sponsored dinner sessions among health care executives on the case for care improvement.

• Regional conferences showcasing efforts to improve patient care, such as supporting patients leaving the hospital, chronic care registries, and patient outreach.

Measurement Progress of the teams participating in the program were tracked using IHA P4P measures.

1. HbA1c Testing – Percent of diabetes patients with one test in last 12 months 2. HbA1c Control – Percent of diabetes patients with HbA1c < 9.0 3. LDL Testing – Percent of diabetes patients with one test in last 12 months 4. LDL Control in CAD – Percent of CAD patients with LDL-c < 100

Results When the P4P results for the physician groups actively participating in the Inland Quality Collaborative are compared for the groups from the same region who did not, the results over time are dramatic. The improvement over time for the active groups far surpasses those that did not, and in some cases exceeds statewide averages. Internal data reported by health plans show that improvement in the Inland Empire outpaces all but one region in the state.

rev 3/12

8

Quotes All the results from the collaborative are not quantifiable. In fact, many include personal stories of people who found renewed energy for their work due to the camaraderie of the collaborative meetings. Below are some examples of what we have heard. • “CQC helped us short-circuit the learning process. We had confidence that what we learned at

the CQC meetings would work.” • “Inland Quality Collaborative really gave us the energy and the optimism to make positive

changes within our group.” • “When I come here I realize we are all trying to solve the same problems; I don’t feel so

alone!” • “We send our staff to CQC programs to receive the training we just can’t offer internally.”

9

Lessons Learned • Competitors can collaborate – At the first meeting, local competitors were hesitant to openly

discuss approaches to improve patient care. After the third meeting, competing physician organizations shared freely; meetings became the place to exchange information on what worked, and didn’t work, to improve patient care. In fact, meetings became a place to share and celebrate each others’ successes.

• Small groups can improve too – Most physician groups in the Inland Empire tend to be smaller than those elsewhere in California. The same kind of interventions worked to improve care, just delivered differently. For instance, a large group can produce custom reports for physician practices and produce patient outreach letters. Smaller groups can bring patient lists to the practice and work collaboratively to complete the patient outreach.

• Improvement can be accelerated by regularly structured peer-to-peer networking – Physician groups regularly participating in CQC programs out-performed others working under the same constraints in the same region.

• Don’t make assumptions about causes of lower performance – When first reviewing the lower performance in Inland Empire, many believed that organizations were not interested in improving. In fact, three quarters of the physician groups participated in improvement activities, with half of the groups devoting significant time and investing significant money in the technology and people required to improve patient care.

Participants

• Beaver Medical Group • Pinnacle MG

• Choice Medical Group • PrimeCare/NAMM

• Desert Oasis Health Care • Regal MG

• Heritage Victor Valley • Riverside Medical Clinic

• High Desert MG • Riverside Physician Network

• High Desert Primary Care • San Bernardino Medical Group

• Kaiser - Riverside • United Family Care

• Loma Linda University Health Care

10

Meteor Program July 2011 – January 2012

Overview The Meteor Program took on the challenge of assisting physician groups to improve performance in Medicare Star measures for 2011 working within only the final six months of the year. Medicare Star measures all make clinical sense, and because Medicare Health Plans compensate physician groups on percent of plan revenue, a pay-for-performance mechanism is built into the business model. Additionally, because measurement is at the plan level, every group’s performance impacts the final score; the ideal improvement strategy is a collaborative effort.

Structure The program targeted physician groups with existing registries that wanted to strengthen approaches to patient outreach and move quickly to close care gaps within six months. Two in-person meetings were combined with coaching calls with CQC staff between meetings.

The curriculum focused on the role of centralized group functions in: • Coordinating bidirectional data exchange with plans to generate gap or exception reports • Addressing the specific challenges around improving scores on the priority Star measures, as

defined by CQC sponsor health plans • Physician outreach and engagement • Member outreach to close gaps through phone calls or reminder letters • Group-sponsored senior wellness clinics

• Breast Cancer Screening • Osteoporosis Management for Women

with a Fracture

• Treatment of Rheumatoid Arthritis • Spirometry for COPD

Measurement Health plans defined four priority metrics based on clinical importance, small denominators, and opportunity for performance improvement in a short timeframe. Priority measures included the following, however physician groups selected measures based on organizational strategic goals:

Participants Twelve physician groups enrolled, several of which are umbrella systems, collectively caring for about 125,000 Medicare enrollees served by over 2,600 network primary care physicians:

• Alta Bates Medical Group (ABMG) • Brown and Toland Physicians (BTMG) • Bristol Park Medical Group/Memorial Care

Medical Foundation • Choice Medical Group • Coast Healthcare Management • Good Samaritan Medical Practice Association • High Desert Primary Care Medical Group

• Independence Medical Group • NAMM/Prime Care Riverside • Riverside Physician Network • Sutter Medical Network • Synermed/Multicultural Primary Care

Medical Group

1 rev 3/12

Results Data was not aggregated across the participating groups, the following is a sampling of reports from individual groups: • Coast Healthcare Management increased osteoporosis management at each of its three IPAs

(Alamitos, Lakewood, St. Mary’s) and improved from a one-Star to a three-Star rating.

• ABMG reached a five-Star rating for Breast and Colorectal Cancer screenings.

• Both ABMG and BTMG are on the brink of five-Star ratings for Monitoring Persistent Medications and Diabetes measures.

• High Desert Primary Care improved Star ratings over the course of six months by implementing birthday card reminders and comprehensive visits at a dedicated wellness center:

Interventions Participating groups adopted a variety of interventions to improve scores over the last six months of 2011. The following are examples of the most common or most effective strategies adopted. • Centralized outreach letters and phone calls to patients with PCP support

o Personalized language to make the case to patients o Bilingual nursing staff

• Standing lab orders and mailed lab slips to patients • Send gap reports to PCP practices monthly • Educate physicians and office staff about Medicare Star and business case

o Measure-specific provider Continuing Medical Education, such as disease-modifying antirheumatic drugs (DMARDS) to treat rheumatoid arthritis

• Develop sole-source contract with imaging provider for mammograms • Contract for and provide transportation for patients to imaging appointments • Provide incentives to office staff • Conduct wellness clinics with in-house bone density dexa scan, lab draw, and mammograms • Improve Risk Adjustment Factor (RAF) scoring, Hierarchical Condition Categories (HCC) coding,

and Annual Health Assessments (AHA) by o Adopting a standard AHA form, o Setting and communicating standards for PCP practices, o Delivering pre-populated forms to PCPs for completion o Implementing pay cuts for not meeting standards after incentives failed.

High Desert Primary Care Star Rating Measure July 2011 December 2011 Blood pressure control 1.7 3.4 Glaucoma screening 1.3 4.6 Diabetes eye exams 1 4.8

• Choice Medical Group improved multiple Star measures through outreach and education of PCPs and a senior health and wellness center.

• Prime Care Riverside conducted patient outreach through letters and phone calls, delivered gap reports to PCPs, and conducted PCP and office staff education to improve multiple measures.

12

Lessons Learned With the Meteor Program, CQC demonstrated that individual groups can improve quality performance within just six months with a focused, centralized effort by the group staff. Achieving five-Star performance will require integration of Medicare Star measures into groups’ registries. In addition, data reconciliation with health plans uncovered a number of opportunities to improve outreach to patients and capture services already provided. Both physicians and members are challenged to fully understand new Star requirements and the implications of poor performance. Homebound and frail members present additional challenges that can be overcome by interventions such as arranging and providing transportation, engaging community resources, and providing in-home care.

Quotes

“Being plugged in to the health plans and to the status of Star measures and updates was big for our organization. We’re so busy with so many priorities, Meteor has been a great way for us to focus on MA Stars … and doing it in an environment where we can network with others in the industry helped us know we were on the right track.” - Sutter Medical Network

“This is all new and we’re a small group. The Meteor Program gives us context, it gives us resources. We literally take some of the materials word-for-word that Meteor has given us and we teach that to our providers, so it’s really an extension of us. It’s what we would like to be if we had the resources ourselves. If you’re not in Meteor, you’re waiting around … you’re going to be left behind.” - Choice Medical Group

Results Cont.

13

California Quality Collaborative

would like to acknowledge the following organizations for supporting the program:

Anthem Blue Cross Blue Shield of California

California Association of Physician Groups Cigna

Health Net Pacific Business Group on Health

SCAN Health Plan United Healthcare

14