Case Study Organized Health Care Delivery System • June 2009 HealthPartners: Consumer-Focused Mission and Collaborative Approach Support Ambitious Performance Improvement Agenda D ouglas M c C arthy, K imberly M ueller, and I ngrid Tillmann I ssues R esearch , I nc . The mission of The Commonwealth ABSTRACT: HealthPartners is the nation’s largest nonprofit, consumer-governed health Fund is to promote a high performance care organization, providing health and dental care and coverage to more than 1 million health care system. The Fund carries individuals in Minnesota and surrounding states. Key factors driving HealthPartners’ per- out this mandate by supporting formance are a consumer-focused mission; a regional focus, scale, and scope integrat- independent research on health care ing a broad range of services; strategic use of electronic health records to support care issues and making grants to improve redesign; and a culture of continuous improvement. A comprehensive model for improve- health care practice and policy. Support for this research was provided by ment includes setting ambitious targets for health system transformation; measuring The Commonwealth Fund. The views what is important in order to optimize care; agreeing on best care practices and support- presented here are those of the authors ing improvement at the clinic level; aligning incentives with goals; and making results and not necessarily those of The transparent internally and externally. HealthPartners’ experience suggests that a nonprofit Commonwealth Fund or its directors, health plan market oriented to physician group practice—supported by collaborative mea- officers, or staff. surement, improvement, and reporting structures—creates a community environment that helps each participant achieve objectives more effectively. For more information about this study, please contact:      Douglas McCarthy, M.B.A. Issues Research, Inc. OVERVIEW dmccarthy@issuesresearch.com In August 2008, the Commonwealth Fund Commission on a High Performance Health System released a report, Organizing the U.S. Health Care Delivery System for High Performance, that examined problems engendered by fragmenta- tion in the health care system and offered policy recommendations to stimulate greater organization for high performance.1 In formulating its recommendations, To download this publication and the commission identified six attributes of an ideal health care delivery system learn about others as they become available, visit us online at (Exhibit 1). www.commonwealthfund.org and HealthPartners is one of 15 case-study sites that the commission examined register to receive Fund e-Alerts. to illustrate these six attributes in diverse organizational settings. Exhibit 2 sum- Commonwealth Fund pub. 1250 Vol. 12 marizes findings for HealthPartners, focusing primarily on the ambulatory care 2T he  C ommonwealth F und Exhibit 1. Six Attributes of an Ideal Health Care Delivery System • Information Continuity Patients’ clinically relevant information is available to all providers at the point of care and to patients through electronic health record systems. • Care Coordination and Transitions Patient care is coordinated among multiple providers, and transi- tions across care settings are actively managed. • System Accountability There is clear accountability for the total care of patients. (We have grouped this attribute with care coordination since one supports the other.) • Peer Review and Teamwork for High-Value Care Providers (including nurses and other members of care teams) both within and across settings have accountability to each other, review each other’s work, and collaborate to reliably deliver high-quality, high-value care. • Continuous Innovation The system is continuously innovating and learning in order to improve the qual- ity, value, and patients’ experiences of health care delivery. • Easy Access to Appropriate Care Patients have easy access to appropriate care and information at all hours, there are multiple points of entry to the system, and providers are culturally competent and respon- sive to patients’ needs. setting. Information was gathered from HealthPartners’ Today, HealthPartners provides individual, system leaders and from a review of supporting docu- group, and public insurance coverage to more than ments.2 The case-study sites exhibited the six attributes 1 million members of health and dental plans in in different ways and to varying degrees. All offered Minnesota, western Wisconsin, North and South ideas and lessons that may be helpful to other organiza- Dakota, and Iowa (Exhibit 3). Members receive care tions seeking to improve their capabilities for achiev- from a network of some 30,000 providers including ing higher levels of performance.3 both owned and contracted medical groups, specialty clinics, hospitals, and dental practices. Other lines of ORGANIZATIONAL BACKGROUND business include behavioral health, eye care, disease HealthPartners, headquartered in Minnesota’s Twin management, integrated home care and hospice, phar- Cities, is the nation’s largest nonprofit, consumer- macy, wellness, and personalized health promotion governed healthcare organization. Its mission is to for individuals and groups. The organization employs “improve the health of our members, our patients, and almost 10,000 and has annual revenue of $3.1 billion. the community.” The organization was formed through About one-third of HealthPartners’ 640,000 a 1992–1993 merger between Group Health, one of the health plan members receive care from the nation’s oldest staff-model health maintenance organi- HealthPartners Medical Group (HPMG), a multi- zations (HMOs) founded in 1957; MedCenters Health specialty group practice that employs more than 600 Plan, a network-model HMO; and Regions Hospital physicians who practice at 50 HealthPartners clinic (formerly St. Paul-Ramsey Medical Center), a 427-bed locations throughout the Twin Cities and in St. Cloud teaching hospital and level I trauma center. Two 25-bed and Duluth, Minn. (Exhibit 4). HPMG also provides critical-access hospitals have since joined the system: care for patients who have other insurance (includ- Westfields Hospital in New Richmond, Wisconsin, and ing Medicare or Medicaid), who represent about 40 Hudson Hospital and Clinics in Hudson, Wisconsin. percent of the medical group’s 400,000 patients. Each H ealth P artners : C onsumer -F ocused M ission and C ollaborative A pproach 3 Exhibit 2. Case Study Highlights Overview: HealthPartners is a family of nonprofit, consumer-governed, integrated health care organizations including a teaching hospital and two critical-access hospitals; the multispecialty HealthPartners Medical Group (HPMG), with more than 600 physicians practicing in 50 clinics; health and dental plans offering group, individual, and public insurance cover­ ge to more than 1 million individuals through a a network of 30,000 providers in Minnesota, western Wisconsin, North and South Dakota, and Iowa; a research foundation; and a medical-education institute. Attribute Examples from HealthPartners Information Enhanced electronic health record (EHR) system. Patient information is integrated across HPMG clinics with Continuity disease registries, clinical reminders, safety alerts, and evidence-based decision support to guide care processes before, during, and after the patient visit. Online personal health record and health assessment. HPMG patients also can schedule ap­ ointments, refill p prescriptions, share secure e-mail with clinicians, receive preventive care reminders, and view lab results, medications, and immunizations online. Participation in Minnesota Health Information Exchange. Secure interchange of clinical information will facilitate patients’ movement among medical groups and health systems. Care Coordination EHR supports care transitions for HPMG heart-failure patients after hospital discharge. and Transitions; Chronic disease management programs iden­ify eligible health plan members, engage them in self-care, and pro- t System mote medica­ion compliance, appropriate treatment, home monitoring, communication, and follow-up in coordina- t Accountability* tion with primary care physician. For example: Behavioral health management includes early intervention program to identify and refer members at risk of depression or problem drinking, medication management programs to promote treatment adherence, and case management to coordinate services for members at risk of behavioral health crises. Workplace wellness programs foster population health improvement by assessing employees for health risks, offering telephonic coaching and education to support lifestyle changes, and promoting engagement through incentives. Peer Review and Prepared Practice Teams in HPMG primary care clinics use a “Care Model Process” and EHR to standardize care Teamwork for processes, anticipate patient needs, give evidence-based care, and ensure follow-up after visits. High-Value Care Continuous Comprehensive improvement model disseminated through leadership teams, workforce development, and par- Innovation ticipation in collaborations such as the Institute for Clini­ al Systems Improvement help develop common clinical c guidelines and improvement strategies. Elements in­ lude: (1) set ambitious targets for health system transformation, (2) measure what is important in c order to optimize care, (3) agree on best care practices and support improvement at the clinic level, (4) align incentives with goals, and (5) make results transparent. Performance feedback and incen­ives and tiered networks t en­ ourage contracted providers to improve value. c Easy Access to Health plan offers “nurse navigators,” after-hours nurse-advice call line, and open-access options with no referral Appropriate Care required to see a specialist. Advanced-access scheduling is associated with reduced appointment waiting time and increased continuity of care with the same provider in HPMG primary care clinics. Walk-in urgent care and retail convenience clinics seek to integrate with traditional clinics. Well@Work work-site clinics offer acute care and health promotion. Cultural competency initiatives include professional transla­ors, translated materials, educa­ional resources, and t t the collection of demographics at point of care. *System accountability is grouped with care coordination and transitions, since these attributes are closely related. 4T he  C ommonwealth F und Exhibit 3. HealthPartners Network Area Source: HealthPartners. clinic, and the medical group as a whole, is led by a HealthPartners Simulation Center for Patient Safety at physician-administrator pair. Metropolitan State University, which provides “real- The HealthPartners Research Foundation istic hands-on experiential learning opportunities” for conducts clinical, health-services, and basic sci- health care professionals and medical and nursing stu- ence research in the public domain, with a focus on dents from Minnesota and neighboring states. improving health care and health through partnerships Minnesota, and the Twin Cities in particular, has with care delivery organizations. The HealthPartners been a leader in developing innovative approaches to Institute for Medical Education sponsors 16 medi- health care financing and delivery, with a continuing cal residency programs and 240 continuing medical orientation toward physician group practice. Public and education programs. The institute jointly sponsors the private employers are collectively active in value-based Exhibit 4. HealthPartners Medical Group Clinic Locations Source: HealthPartners. H ealth P artners : C onsumer -F ocused M ission and C ollaborative A pproach 5 purchasing initiatives that develop shared strategies to own system to add information that is not available in promote quality and cost-containment goals.4 Several ambulatory care records, such as hospital admissions collaborative organizations bring stakeholders together and ER visits. to develop common clinical guidelines, improvement All health plan members can create an online strategies, measurement metrics, and performance personal health record (PHR) to keep a medical his- reporting and incentive programs (see Appendix A). By tory, track health goals, take an online health assess- law, HMOs are nonprofit organizations in Minnesota. ment, and view their medical claims. Patients of the Three large health plans—HealthPartners, Medica, and HealthPartners Medical Group can access additional Blue Cross Blue Shield—dominate the market.5 online capabilities to schedule doctor appointments, request prescription refills, send secure e-mail com- INFORMATION CONTINUITY munications to their care team (“e-visits”), receive All HealthPartners Medical Group clinicians have e-mail reminders for preventive or chronic care, and access to electronic health records (EHRs) for their view their laboratory test results, medication lists, and patients. The EHR was implemented in stages begin- immunization records. In adopting this technology, ning with pilot sites in the 1990s. In 2001, the medical HealthPartners aimed to promote a more collaborative group implemented online medication ordering and relationship between patients and caregivers while also simple documentation using a basic Web-based EHR. giving patients greater control of information to better By 2003, the group determined that it needed a more manage their own health.7 robust EHR providing four key capabilities: chart HealthPartners is participating in a public– review, physician-order entry (including medications, private partnership called the Minnesota Health laboratory tests, and images), documentation, and best- Information Exchange to enable the secure exchange of practice alerts and reminders. HealthPartners selected clinical information such as medical histories, labora- and enhanced a third-party software system (EpicCare tory orders, and test results between providers and from Epic Systems Corp.) to meet these requirements. payers as patients move among medical groups and Installation was completed in primary care clinics by health systems. 2005, Regions Hospital by 2006, and specialty and behavioral health clinics by 2008. CARE COORDINATION AND TRANSITIONS: HealthPartners has customized the EHR to TOWARD GREATER ACCOUNTABILITY FOR include advanced capabilities such as disease registries, TOTAL CARE OF THE PATIENT clinical reminders, safety alerts, and decision sup- Improving care transitions. The HealthPartners port for evidence-based guidelines. Panels of medical Medical Group and Regions Hospital are working experts developed clinical content in core topic areas together to improve care transitions for patients with that was embedded in the EHR to support the delivery heart failure, according to Beth Averbeck, M.D., asso- of preventive and chronic care services before, during, ciate medical director for primary care. For example, and after the patient visit. In contrast to stand-alone primary care physicians receive an electronic alert disease registries, the EHR integrates patient informa- when one of their heart failure patients is admitted to tion across health conditions so that clinicians can have Regions hospital. When the patient is discharged, the a unified view of a patient’s history.6 hospital’s care managers notify the medical group’s The health plan supplies chronic disease registry heart failure clinic and telephone the patient at home data to its contracted medical groups so that physi- to ensure that he or she has a follow-up appointment cians can track and identify their patients who are in and is taking the proper medications. The patient’s pri- need of evidence-based chronic care services. Medical mary care physician and a cardiac specialist in the heart groups that have an EHR can import the data into their failure clinic then comanage the patient with a jointly 6T he  C ommonwealth F und agreed-upon follow-up schedule, using the EHR to • 6 percent reduction in all-cause admissions for facilitate communication and patient reminders. members with asthma To promote improved care transitions across • 5 percent reduction in all-cause admissions for its network, the health plan recently began reporting members with diabetes on hospital readmissions for heart failure patients in each of its cardiology care groups. As part of its per- • 13 percent reduction in admissions for heart formance incentive program for contracted providers attack, heart bypass surgery (CABG), and (described below), the plan has set a goal of reducing chest pain (angina) for members with coronary readmissions within 30 and 90 days of an initial hospi- artery disease talization to 5 percent and 15 percent of these patients, • 6 percent reduction in all-cause admissions for respectively, from current planwide rates of 7.9 percent members with chronic heart failure. and 17.3 percent during 2005–2007.8 Improving behavioral health. Behavioral health Managing chronic disease. HealthPartners has management programs illustrate how HealthPartners engaged in a series of innovative and collaborative is seeking to develop a proactive approach to care disease management activities since the early 1990s, management that supports the relationship between focused initially on diabetes. The authors of a previous patients and their physicians (or other providers) but Commonwealth Fund report noted that the integrated does not rely exclusively on a patient visit to identify nature of HealthPartners Medical Group (formerly the and address health problems. These programs are staff-model HMO) likely reduced the costs and increased part of the organization’s broader strategy to promote the success of developing disease management pro- health by removing barriers so that health plan mem- grams in comparison to efforts by looser networks of bers can more easily access mental health or chemical independent physicians. They estimated that the eco- health evaluation and treatment services when needed, nomic value of improved quality of life (from reduced according to Karen Lloyd, senior director of behavioral disease complications) would be $31,000 for a diabetic health strategy and operations. For example, a behav- patient who participated in the program for 10 years.9 ioral health direct-access network allows members to The health plan now offers a suite of disease see any outpatient behavioral health professional with- management programs under the name CareSpan that out prior approval or authorization. can be purchased by employer groups for their health In an early intervention program, licensed plan members with conditions such as asthma, diabe- behavioral health professionals (social workers or psy- tes, heart disease, heart failure, and chronic obstructive chologists) contact health plan members whose health pulmonary disease. CareSpan uses disease registries, assessment indicates a risk for depression or problem health assessments (described below), and referrals drinking—two modifiable risk factors that can affect a from physicians to identify patients who would benefit person’s productivity and ability to manage a chronic from early intervention, disease management, and case disease. During the outreach call, the behavior health management programs. Participants receive personal- professional conducts additional screening to ascertain ized education and support from nurses or other pro- the nature of the individual’s concerns or symptoms. If fessionals such as dieticians for self-care, medication the individual appears to have an undiagnosed, clini- compliance, home monitoring, and follow-up as needed cally treatable condition, the professional provides in coordination with their physician and clinic. The education and guidance to motivate him or her to see a plan reported the following audited results for partici- behavioral health professional for a full evaluation. Those pants in these programs from 2003–2004 to 2005–2006: with subclinical conditions are offered guidance and provided educational resources on how to reduce their risk for developing depression or alcohol dependency. H ealth P artners : C onsumer -F ocused M ission and C ollaborative A pproach 7 A behavioral health disease management pro- The plan’s analysis of program effectiveness gram focuses on health plan members with depression comparing the study group (whether engaged in the who are beginning antidepressant medication. The pro- program or not) to a historical comparison group (with gram sends these members monthly educational mate- costs trended forward) found that ambulatory behav- rial and reminders to refill their prescriptions for six ioral health visits were 35 percent higher among the months. The member’s physician receives a letter if the study group, medication costs per member per month patient fails to refill his or her medication in a timely were 11 percent lower, inpatient behavioral health days manner. Anecdotal feedback suggests that physicians per 1,000 members were 4 percent lower, and costs find this service useful for prompting follow-up with per member per month were 18 percent lower in the patients. The health plan credits this program with a 17 latest annual measurement period.11 The overall return percent improvement in rates of six-month medication on investment was estimated at $4 saved in medical adherence.10 The plan has expanded the program to costs for every $1 spent on program administration. promote medication adherence and improved self-care Recently, the plan has found that residential chemical among members with bipolar disorder or schizophre- health days have increased as inpatient mental health nia, two conditions that put an individual at high risk days have decreased. Anecdotal information suggests for poor health outcomes. In addition to sending refill that many members at highest risk for hospitalization reminders, the program offers brief telephone coun- have an undiagnosed or untreated chemical health con- seling and referral for those who are not adhering to dition coexisting with a mental health condition. treatment. This program puts a special emphasis on maintaining physical health, as research indicates that Promoting healthy lifestyles. The health plan encour- patients with severe mental illness and taking atypical ages each adult member to complete an online health antipsychotic medications lose an average of 25 years assessment (integrated with his or her personal health of lifespan. record) designed to identify those at risk of developing Several years ago the health plan implemented a chronic illnesses, such as diabetes or heart disease, who telephonic case management program after discovering would benefit from prevention.12 Participants receive that 5 percent of its members with behavioral health– immediate online feedback via a personal report fea- related diagnoses accounted for 50 percent of expen- turing a modifiable risk score (including the change ditures. The program uses a predictive algorithm to in score since a previous assessment) and an action identify members who are at risk of behavioral health plan for making lifestyle changes. Results are used to crises and hospitalizations. A behavioral health case invite the member to participate in disease manage- manager invites these members to participate (by let- ment programs for which they may be eligible. While ter and then by phone) and provides participants with health plan–initiated communications strategies help to self-care education, health coaching, decision support, raise members’ awareness of this service, they have not and care coordination services. Case managers can resulted in high participation rates, nor are physicians access the EHRs of patients seeing physicians in the always prepared to use such information in clinical HealthPartners Medical Group to facilitate care plan- practice. ning and communication with the care team. In 2007, HealthPartners has found that the most effective the engagement rate was about 38 percent and partici- strategy for engaging individuals in healthy lifestyles pant satisfaction was 94 percent. Similar case manage- is to implement the online health assessment together ment services are offered to all health plan members with employer-sponsored programs for improving with illnesses that put them at risk for poor outcomes population health. The health assessment “is a power- and high costs. ful tool to create ‘teachable moments’ for people that can help mobilize them [into] taking active steps to 8T he  C ommonwealth F und health improvement,” said Nico Pronk, Ph.D., vice opportunities and systems issues to be addressed president and health science officer at HeathPartners’ for improvement. Physicians are invited to join JourneyWell program for employers. Realizing this quality improvement teams and to receive training in potential requires an integrated approach to connecting improvement methods based on their clinical interests. employees with programs, he said. The goal is to develop informal leaders who will spread To meet this need, HealthPartners works with knowledge and mentor their peers, said Averbeck. employers locally and nationally to develop workplace Primary care clinics within the HealthPartners health programs that offer incentives (such as reduced Medical Group have adopted a “Care Model Process” copayments and deductibles) for employees to engage (adapted from Wagner’s Chronic Care Model14) that in annual health assessments and follow-up programs. defines “a standard set of workflows for delivering These programs include curriculum-based telephonic evidence-based care that provides a consistent clini- counseling and educational courses to support indi- cal experience for patients and a consistent process viduals in making lifestyle changes such as smoking for care teams.”15 Each clinic’s staff is organized into cessation or weight loss, online programs to promote “prepared practice teams” composed of a physician, increased physical activity levels, and referral to dis- a rooming nurse, a receptionist, and others such as ease management programs and to workplace-specific a pharmacist or dietician when needed for particular resources such as employee assistance programs. patients. The goal is to create a “continuous healing One large Twin Cities employer, BAE Systems, relationship” between caregivers and patients by mak- experienced the following results after participating in ing the best use of collective team skills, enhancing such a program for three years: communication, and ensuring that care is well-coor- dinated and responsive to patient needs. These teams • high levels of reported employee satisfaction typically huddle each morning to review their schedule with the program and objectives for the day. Through standardization of processes and • sustained participation rates of 89 percent or clearer specification of roles, the care team focuses on higher annually among the company’s 1,300 reliably performing core patient interactions within employees and their spouses a defined patient visit cycle—scheduling, pre-visit, • 6 percent improvement in employees’ modifi- check-in, visit, and post-visit—to anticipate patient able risk scores and health behaviors needs, remind patients of health issues, and provide • 3.3 percent annual reduction in medical claims follow-up after the visit. For example, pre-visit plan- costs (about half of which was attributed to ning may include identifying preventive care services lower-than-expected hospital admissions), equal that will need to be provided at the visit and contacting to about $59 per employee per year and yielding the patient to schedule laboratory tests so that results a 2:1-to-3:1 return on investment are available for review during the visit. At the patient visit, the team uses the EHR to address the patient’s • improved workforce productivity valued at more health maintenance and/or chronic care needs, refill than $1 million.13 prescriptions if needed, and schedule future appoint- ments. Patients receive an “after-visit summary” of PEER REVIEW AND TEAMWORK their care plan to promote treatment adherence and FOR HIGH-VALUE CARE receive outstanding lab results by their preferred Physicians in the HealthPartners Medical Group engage method of notification (letter, phone, or e-mail). in a formal peer review process at the departmental Implementation of the Care Model Process, level. Cases are referred for review based on patient along with other interventions, was associated with or staff concerns, with a focus on identifying learning improvements in the quality of care received by H ealth P artners : C onsumer -F ocused M ission and C ollaborative A pproach 9 primary care patients, while also laying a foundation and invited to schedule a visit or for making future improvements in care.16 other needed services. • Incorporating the PHQ-9 patient health • Pre-visit planning activities increased from 8 questionnaire, an assessment tool for depression, percent of patients in 2005 to 70 percent in 2006 into the primary care visit cycle (completed by (with improvement continuing to more than the patient and documented in the EHR by the 90 percent of patients today), and accuracy of rooming nurse) resulted in a doubling of patients health maintenance records rose from 56 percent who use it, from 32 percent of primary care to 95 percent and has remained near that level clinic patients with newly diagnosed depression since that time (Exhibit 5). in 2004 to 65 percent in 2007. The tool provides • Patients receiving optimal diabetes care— a structured way for physicians to communicate measured as a composite, or “bundle,” of five with patients about their symptoms and to make treatment goals including control of blood treatment adjustments as needed.18 glucose, blood pressure, and cholesterol levels; • Patient satisfaction (percentage reporting a prob- aspirin use; and non-use of tobacco—increased lem) has improved 24 percent since 2006 as the from 4 percent of diabetic patients in 2004 to intervention has shifted focus to improving the 15 percent in 2006 and 25 percent in the fourth patient experience. Areas of attention included quarter of 2008 (Exhibit 6). This increase builds improving communication with patients about on more than a decade of work to improve the expected waiting time, training staff to consis- quality of diabetes care.17 Recent improvements tently demonstrate respect, and making sure were facilitated by the use of a monthly that the patient’s main reason for the visit has “exceptions report” that identifies diabetic been addressed. patients who are not up-to-date on planned- care visits, have missed follow-up care, or are The medical group developed the Care Model not achieving treatment goals. These patients are Process starting in 2002 through its participation in the contacted by telephone or electronic reminder Pursuing Perfection initiative, funded by the Robert Wood Johnson Foundation and led by the Institute for Exhibit 5. HealthPartners Medical Group Care Model Process: Summary of Implementation Results Baseline (Apr.–Jul. 2005)* After implementation (Feb. 2006) 100 95 86 80 75 70 60 56 50 40 35 20 8 0 Visit scheduling Pre-visit planning Health maintenance Opportunity management Note: Visit scheduling = percent of primary care visits scheduled where patient was offered needed health maintenance screening; Pre-visit planning = percent of primary care visits that pre-visit planned; Health maintenance = percent of primary care patients where the electronic medical record accurately reflects the patient’s needs; Opportunity management = percent of patients will all health maintenance services discussed, offered, ordered, scheduled, and/or provided at the primary care visit. Source: M. McGrail and B. Waterman, “HealthPartners Medical Group: Care Model Process,” Group Practice Journal, Nov.–Dec. 2006 55(10):9–20. 10T he  C ommonwealth F und Exhibit 6. HealthPartners Medical Group: Achieving Optimal Diabetes Care Percent of diabetic patients achieving all five treatment goals* 30 25 25 20 15 15 14 14 10 5 4 0 2004 2005 2006 2007 2008* *Optimal diabetes care means the percentage of patients aged 18 to 75 with diabetes (Type 1 or 2) who had hemoglobin A1c <7%, LDL cholesterol <100 mg/dl, blood pressure <130/80 mmHg; daily aspirin use (patients ages 41–75), and documented tobacco-free. **Preliminary fourth quarter data provided by HealthPartners. Source: Minnesota Community Measurement (www.mnhealthcare.org) and HealthPartners (2008 data). Healthcare Improvement. Frontline staff from three ent internally and externally. Each of these components pilot sites mapped workflows to optimize the patient is described below.20 visit process during a two-day rapid-design process. The model was refined and disseminated to all primary Setting ambitious targets. The organization sets its care sites through an internal learning collaborative. priorities through a strategic plan and a balanced score- Researchers who studied the change at an early stage card with four components: people (the organization’s reported that care teams found it challenging to trans- workforce), health outcomes, consumer and patient late general principles from Wagner’s Chronic Care experience, and financial stewardship (Exhibit 7). The Model into clinical practice, leading to some trial and health component includes Health Goals 2010 (see error as they sought to define a workable approach.19 Appendix B), the organization’s blueprint for achiev- Ongoing redesign is based on information gathered ing the Institute of Medicine’s (IOM’s) six criteria for a from audits and measures of effectiveness, with a cur- successfully transformed health care system: care that rent focus on improving outreach between visits (as is patient-centered, safe, timely, effective, efficient, and described above for diabetes). equitable. Setting ambitious goals implies that the organi- CONTINUOUS INNOVATION zation is committed to creating “the capacity to try and HealthPartners has developed a comprehensive model make them a reality,” said Mary Brainerd, HealthPartners’ for improvement that is disseminated through leader- CEO. This means “not just setting a goal and hoping ship councils that oversee improvement work, through for the best, but a strong commitment of resources to workforce skills development, and through participa- make it happen.” To ensure that these aspirations will tion in learning collaborations. The interrelated compo- be translated into action, the board of directors estab- nents of this model include (1) setting ambitious targets lished the Health Transformation Committee, which for health system transformation, (2) measuring what is sets goals and oversees the organization’s efforts to important (rather than what is simply easy) for optimiz- redesign systems in pursuit of the IOM aims. ing patient care, (3) agreeing on best care practices and One of the plan’s health goals, for example, is supporting improvement at the clinic level, (4) aligning to achieve 100 percent improvement in a composite incentives with goals, and (5) making results transpar- of lifestyle measures for adults including tobacco and alcohol use, physical activity, healthy weight, and H ealth P artners : C onsumer -F ocused M ission and C ollaborative A pproach 11 nutrition. To reach tobacco use prevention goals, the percent of Minnesota adults from 1999 health plan offers incentives and supports collaborative to 2007.22 efforts to help medical groups adopt tobacco control • Parent-reported secondhand smoke exposure interventions recommended by the Centers for Disease among children of health plan members declined Control and Prevention: asking patients about tobacco from 23 percent in 1998 to 5 percent in 2008. use (by making tobacco use a “vital sign” in the medi- Measuring what is important. Optimizing care for cal record), advising tobacco users to quit, and assist- chronic conditions can improve patient outcomes ing them with a plan to quit (such as by prescribing while also reducing costs. For example, HealthPartners medication and referring them to telephone counsel- found that diabetic patients whose risk factors for ing). The plan has seen the following improvements disease complications were not well controlled expe- among its member population (Exhibit 8).21 rienced $60,000 in average medical costs per year, as compared to $5,000 for those whose risk factors were • Patients who were assessed by their clinicians controlled. However, the current practice of measur- for tobacco use increased from 71 percent of ing individual care processes separately can obscure health plan members in 1998 to 96 percent in the need to address all of the risk factors affecting a 2007. Almost two-thirds (65%) of tobacco users patient’s health outcomes. reported that they were offered assistance in In response, HealthPartners in 1996 began quitting in 2008, as compared to fewer than half developing composite measures (“bundles”) of optimal (47%) who said so in a 2001 health plan survey. care to set a high bar that would encourage clinicians to meet all evidence-based care practices. Bundles • Self-reported tobacco use declined by almost currently address diabetes, coronary artery disease, half among adult health plan members, from depression, preventive care, and lifestyle. By 2006, 25 percent in 1998 to 13 percent in 2006—a more than one in five health plan members with diabe- rate that was sustained through 2008. This was tes, hypertension, and heart disease met all cardiovas- twice the improvement seen in tobacco use cular risk targets and over half met four of five targets, statewide, which fell from 27 percent to 21 contributing to 4,000 fewer deaths from heart disease. Exhibit 7. HealthPartners Strategic Objectives Dimensions Key Strategic Objectives Success Indicator People Live the HealthPartners values Employee well-being Diversity Health Be the best at improving health Healthier patients and members Health Goals 2010 performance Experience Deliver an experience that consumers want Increased patient, member, and employer and deserve at an affordable cost satisfaction Stewardship Deliver improved value, growth, and finan- Growth cial results Improved margin Reduced cost trends Documented community benefit 12T he  C ommonwealth F und Exhibit 8. HealthPartners Health Plan: Tobacco Use and Exposure Rates Percent of members Adult Tobacco Use Prevalence Rate 30 26 25 Second Hand Smoke Exposure of Children 25 21 22 23 23 20 18 20 19 15 16 15 15 13 14 13 14 13 10 11 9 5 6 6 5 0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 Note: Adult Prevalence Rate represents member responses to the question, “During the past year, have you used tobacco products such as cigarettes, cigars, pipes, snuff, or chewing tobacco?” Second Hand Smoke Exposure represents member responses to the question, “During the past year, have any of your children been exposed to second-hand smoke at home or day care?” Source: HealthPartners. Among approximately 20,000 members with diabetes, Several clinics within the HealthPartners for example, this improvement means 100 fewer heart Medical Group and contracted medical groups are attacks, 740 fewer eye complications, and 140 fewer among a growing number statewide participating in amputations annually compared to 1995, according to another collaborative ICSI initiative called DIAMOND the health plan’s calculations. (Depression Improvement Across Minnesota: Offering a New Direction), which is applying an evidence-based Agreeing on best care practices and supporting model known as IMPACT to improve the identification improvement at the clinic level. HealthPartners partici- and treatment of depression in primary care practices.23 pates in and financially supports Minnesota’s Institute ICSI identified common practice redesign (Exhibit 9) for Clinical Systems Improvement (ICSI), which brings and payment reform elements to implement the model together health plans and medical groups to develop in a systematic, staged fashion among medical groups evidence-based clinical guidelines and sponsors collab- that demonstrate a readiness for change. The medical orative improvement activities (see Appendix A). For groups have negotiated with health plans to receive example, the plan has been able to reduce unnecessary a periodic fee to cover the cost of these enhanced imaging studies for lower back pain—saving an esti- services based on evidence that they will ultimately mated $6.6 million in 2007—in part because guidelines reduce costs while improving patient outcomes. based on American College of Radiology recommen- Early results of the DIAMOND Initiative are dations were adopted collaboratively through a process promising: Patients in the participating clinics are facilitated by ICSI. HealthPartners promoted commu- more regularly being assessed with the PHQ-9 and nitywide adoption by sharing decision-support algo- are achieving substantially higher rates of treatment rithms for medical groups to embed in their own EHRs response and symptom remission than are primary care and processes. Allowing medical groups to implement patients with depression statewide.24 the guidelines internally, rather than being subject to Other innovations to improve the quality and onerous preauthorization requirements, helped over- efficiency of care saved the HealthPartners Medical come their resistance to change, according to George Group an estimated $74 million in 2007 and almost Isham, M.D., medical director and chief health officer. $100 million in 2008. For instance, an initiative to increase the use of generic pharmaceuticals involved H ealth P artners : C onsumer -F ocused M ission and C ollaborative A pproach 13 analyzing data to identify opportunities for inter- HealthPartners first began using payment incen- vention, systematizing generic drug conversions by tives in 1996 to stimulate improvement among its con- embedding standing orders in the EHR, giving clini- tracted providers. In 2007, the health plan paid more cians feedback on their prescribing patterns, and com- than $21 million in incentives (representing about 2.2 municating progress.25 As a result of these efforts, percent of total reimbursement) to contracted medical generic prescribing rose to 72 percent in 2007 from groups and hospitals for meeting quality and patient- 45 percent in 2002. With an average difference in cost experience targets and contractually negotiated goals between branded and generic drugs of almost $150 per such as the use of health information technologies (see prescription, each percentage point increase in the rate Appendix C). Some medical groups may redistribute of generic usage translates to $1 million in savings. incentives to individual physicians while others use the performance payment to fund improvements in their Aligning incentives with goals. Management incen- quality infrastructure. tives are linked to the organization’s improvement HealthPartners was the first health plan to refuse goals. Within the HealthPartners Medical Group, to pay hospitals (and to prohibit them from billing primary care physician compensation is based 87 per- its members) for so-called “never events,” which are cent on productivity to assure timely access to care in rare medical errors such as surgery on a wrong body an efficient manner, 3 percent on quality and service part that should never happen to a patient. The health metrics, and 10 percent on participation in improve- plan adopted this policy in 2005 following passage of ment activities. Changing from salary- to productivity- a Minnesota law requiring hospitals to disclose such based pay (while also implementing advanced-access events. Medicare has since adopted a similar policy. scheduling, described below) was associated with a 38 About 150,000 health plan members are percent increase in primary care physician productivity enrolled in value-based tiered networks that encourage and a 20 percent decrease in cost per relative value unit them to select efficient providers by varying copay- of work from 1998 to 2002.26 Implementing the Care ment and coinsurance levels based on more than 70 Model Process and other interventions was associated measures of the cost and quality of care provided. To with a further 14 percent increase in physician produc- promote treatment adherence among individuals with tivity along with increased patient satisfaction between chronic illnesses, the health plan also offers a value- 2004 and 2005, and physician productivity has contin- based drug plan with reduced copayment or coinsur- ued to increase since that time.27 Exhibit 9. The DIAMOND Initiative: Key Components of Depression Care 1. Standard and reliable use of a validated screening tool—the PHQ-9 patient health questionnaire—for assessment and ongoing management of depression. 2. Systematic patient follow-up tracking and monitoring (based on repeat PHQ-9 measurements and use of a patient registry). 3. Use of evidence-based guidelines and a stepped-care approach for treatment modification or intensification. 4. Relapse prevention plan for patients ready to move out of the care management program. 5. Addition of a care manager to staff to educate, coordinate, and troubleshoot services for patients with depression. (HealthPartners trained medical assistants to fill this role under the supervision of the consulting psychiatrist.) 6. Psychiatric consultation and caseload review. Source: Institute for Clinical Systems Improvement, the DIAMOND Initiative, http://www.icsi.org/diamond_white_paper_/diamond_white_paper_28676.html. 14T he  C ommonwealth F und ance levels for certain drug categories used to treat tion. To help improve rates of breast cancer screening chronic conditions. among underserved populations, two HealthPartners Medical Group clinics began offering same-day mam- Making results transparent. HealthPartners mea- mograms to women who were due or overdue for sures comparative clinical indicators of performance screening at the time of a clinic visit. Results were for its network providers and for the health plan as a promising, and the innovation is being spread to other whole. It has reported these results to providers and locations with on-site mammography service. the public for more than 10 years.28 In addition, the HealthPartners Medical Group has joined more than Reducing Appointment Waiting Time. In 2000, the 120 other physician group practices in the region in HealthPartners Medical Group instituted advanced- publicly reporting clinical performance as part of the access scheduling in 17 primary care clinics to promote Minnesota Community Measurement initiative, a non- the availability of standardized, same-day appoint- profit collaboration between the state medical associa- ments with a patient’s regular physician. Today, all tion and participating medical groups, consumers, busi- primary care clinics offer same-day access and almost nesses, and health plans (see Appendix A). For 2007, 30 percent of primary care visits are same-day appoint- HealthPartners Medical Group received three stars ments. Researchers who studied the change reported (the highest rating representing above-average perfor- that the most important influences on successful imple- mance) in nine of the 11 clinical categories reported by mentation were strong leadership and accountability, Minnesota Community Measurement. both locally and centrally; a clear vision and well- defined plan of action (developed with the assistance EASY ACCESS TO APPROPRIATE CARE of outside consultants); and training, teamwork, and HealthPartners offers centralized scheduling, urgent support through collaborative learning sessions to help care clinics, and open-access health plan options that clinics overcome obstacles to change.29 The researchers do not require a referral to see a specialist. Health plan reported the following results: members have access by phone to assistance in several forms including “nurse navigators” for questions about • Overall, advanced-access scheduling led to a coverage, networks, and services; the Personalized 76 percent reduction in average waiting time Assistance Line, for questions related to behavioral at the 17 clinics, from 17.8 days in 1999 to 4.2 health issues; the nurse-staffed CareLine, for after- days in 2001 (Exhibit 10). (Waiting time was hours advice on treatment options; and the BabyLine, measured to the third-next-available appoint- staffed by trained ob-gyn nurses for questions related ment to minimize variations due to canceled to pregnancy and postmaternity care. appointments.) Patient satisfaction rose dur- The organization is testing several innovative ing this time, from 36 percent to 55 percent of models of primary care delivery to improve access, patients reporting being “very satisfied” with preserve or improve quality, and expand service offer- quality and service. ings. For example, convenience clinics are being devel- oped as a response to so-called “minute clinics” in • Among patients with diabetes, heart fail- retail stores, focusing on delivering quality of care that ure, and/or depression, advanced access was is equal or superior to traditional primary care delivery, accompanied by a 5 percent to 9 percent and on integration with traditional clinics. “Well@ decrease in urgent-care visits, a higher propor- Work” is a primary care program offered at the work- tion of physician visits being made to primary place that combines acute care services, health risk care physicians, and increased continuity of assessment, health promotion, and behavior modifica- care with the same physician.30 Better continu- H ealth P artners : C onsumer -F ocused M ission and C ollaborative A pproach 15 Exhibit 10. HealthPartners Medical Group: Effect of Advanced Access Scheduling on Appointment Waiting Time Average waiting time to third-next-available appointment, in days 20 17.8 15 10 9.6 5 4.2 3.9 2004 0 1999 2000 2001 2002 Source: L. I. Solberg, M. C. Hroscikoski, J. M. Sperl-Hillen et al., “Key Issues in Transforming Health Care Organizations for Quality: The Case of Advanced Access,” Joint Commission Journal on Quality and Safety, January 2004 30(1):15–24. ity of care was associated with improved qual- populations to help bridge language and cultural ity of care for diabetic patients.31 barriers. (Minnesota is first in the nation in refu- gees as a percentage of immigrants, with the Improving Cultural Competency. HealthPartners is largest Hmong, Somali, and Oromo populations engaged in a multifaceted initiative to improve its abil- in the U.S.) ity to deliver equitable care in a linguistically and cul- • Developing the “Language Assistance Plan” turally competent manner for patients of varying racial to systematize best practices for interpreter and ethnic backgrounds.32 At the care delivery level, services. The plan includes a user’s guide that the organization is seeking to instill equity as a prin- describes how and when to access services, a ciple to be achieved through a consistent care process provider manual that establishes quality and per- across its facilities, while customizing services to meet formance expectations for interpretation service individual needs. Programmatic components of the ini- providers, an annual survey to gauge staff satis- tiative include: faction with different interpreting methods, and reimbursement information related to interpreta- • Establishing a consistent process for asking tion services.33 patients to voluntarily provide demographic information, including race and country of ori- • Sponsoring leadership symposiums, community gin (collected at the point of care) as well as lan- forums, and other forms of outreach to cultural guage spoken and need for interpreter services groups in its communities to build trust, gain (collected during appointment scheduling), to insight into health care access needs, and solicit help guide and improve care delivery. advice on how to improve communication and care delivery. • Broadening the diversity of its workforce (among whom several languages are spoken) Results to date include the near-elimination and providing training, resources, and tools such of ethnic/racial disparities in a composite measure of as professional interpreters, translated materi- adult preventive care among Asian, Hispanic/Latino, als, and educational resources about immigrant and Black/African American patients as compared to 16T he  C ommonwealth F und Exhibit 11. HealthPartners’ Cultural Competency Initiative: Preventive Services Composite Measure* Percentage of patients seen during the quarter who received all preventive screening appropriate to each patient’s age and gender, third quarter 2007 100 80 70 67 67 67 60 60 40 20 0 American Indian Asian Black or Hispanic White or Alaska Native African American or Latino *Composite includes cholesterol, colon cancer screening, mammography, Chlamydia screening, and pap smear. Source: B. Averbeck and N. McClure, “Toward Equity, Addressing Disparities in Care and Experience.” Presented at the American Medical Group Association Annual Conference, Orlando, Fla. March 6–8, 2008. white patients (Native Americans continue to experi- Oct. 2007 to Oct. 2008, according to the Minnesota ence a lower rate), and the elimination of disparities Department of Health.35 The organization’s track between white and non-white patients in the provision record of improvement suggests that it will continue to of heart-attack and pneumonia care in the hospital innovate so as to achieve higher levels of performance. (Exhibit 11). In this instance, the hospital has joined a collaboration sponsored by the Minnesota Hospital Association to RECOGNITION OF PERFORMANCE support efforts to reduce patient falls. In addition to the results of the specific interventions described above, HealthPartners has achieved notable INSIGHTS AND LESSONS LEARNED results on selected externally reported performance Key factors driving HealthPartners’ performance, indicators and has received recognition for its perfor- according to its CEO, Mary Brainerd, and chief mance from several national benchmarking or award health officer, George Isham, M.D., are the organiza- programs (Exhibit 12). In terms of efficiency, data tion’s nonprofit, consumer-focused mission as well from the Dartmouth Atlas of Health Care, which exam- as the leadership and accountability engendered by a ined care at the end of life for Medicare patients with consumer-elected board. “We have a very clear line of chronic illness, indicate that those who received the sight to who our customer is. We’re not confused at all majority of their inpatient care at Regions Hospital had about who we need to solve health care problems for: similar overall Medicare spending per person but fewer It’s for the end consumer,” Brainerd said. hospital days (68 percent) and physician visits (61 per- Brainerd also said she sees “huge opportunities” cent) compared to the U.S. average.34 for an integrated system like HealthPartners “to sup- The identification of areas of excellence does port our members and patients much more effectively, not mean that HealthPartners has achieved perfection, addressing their health needs not only when they’re however. Like the other organizations in this case- in the exam room in the traditional means, but [also] study series, HealthPartners has room for continuing supporting them through programs at the work site, improvement in several areas of care. For example, through linking care delivery and disease manage- Regions Hospital reported six patient falls resulting ment and health improvement capabilities we have in serious disability during a one-year period from developed across the organization.” Organizational H ealth P artners : C onsumer -F ocused M ission and C ollaborative A pproach 17 traits that promote integrated health care delivery at group found that first-generation EHR products had to HealthPartners include a regional focus, scale, and be adapted to include more advanced functions such scope integrating a broad range of services, the stra- as disease registries and decision support to enable the tegic use of electronic health records, and skills for full scope of quality improvement and changes in clini- measuring quality and improving care that have been cal practice. Because automating traditional ways of honed over many years. working will not enable breakthroughs in performance, Although the EHR has been an important tool HealthPartners follows a design principle that desired supporting change, reaping its potential has been an clinical workflow should drive the EHR workflow, and evolutionary process. “Not all the improvements that not vice versa. HealthPartners has realized are attributable to its EHR, Brainerd is quick to admit that the organiza- and not all the improvements that the EHR may facili- tion “has a huge distance to go” to realize its goals for tate have yet been achieved,” Isham said. The medical transforming health care delivery. Motivating change Exhibit 12. Selected Externally Reported Results and Recognition* Inpatient Care Quality36 Heart attack treatment (8 measures): Regions Hospital ranked in the top decile of (CMS Hospital Compare U.S. hospitals evaluated. Jan.–Dec. 2007) Heart failure treatment (4 measures): Regions Hospital ranked in the top quartile of U.S. hospitals evaluated. Pneumonia treatment (7 measures): Regions Hospital ranked in the top quartile of U.S. hospitals evaluated. Overall patient rating of care (HCAHPS): Westfields Hospital ranked in the top decile of U.S. hospitals reporting. Ambulatory Care Quality Clinical quality (33 measures): HealthPartners ranked in the top quartile of com- (NCQA Quality Compass 2008) mercial health plans nationally or regionally on 23 measures, 13 of which were in the top decile. Patient experience (10 measures): HealthPartners ranked in the top quartile of commercial health plans nationally or regionally on two measures. National Recognition and Verispan Top 100 Integrated Health Networks (2005–2007). Ratings Leapfrog Group: Regions Hospital designated one of 13 “Highest Value Hospitals” for efficiency in treating heart disease and pneumonia (2008). National Committee for Quality Assurance: Health Plan Excellent Accreditation; Quality Plus Distinction in Care Management, Health Improvement, and Member Connections; Diabetes Physician Recognition Program (HealthPartners); Innovations in Multicultural Health Care Award. US News & World Report Best Health Plans: HealthPartners ranked among the top 25 Medicare plans in 2005 and among the top 50 commercial plans in 2006–2008. JD Power and Associates National Health Insurance Plan Study: Among commercial health plans evaluated nationally, HealthPartners ranked in the top decile in 2008 (104 plans) and the top quartile in 2009 (128 plans). In the Minnesota/Wisconsin region, HealthPartners ranked first among six plans evaluated in 2008 and second among eight plans in 2009. National Business Coalition on Health eValue8: HealthPartners HMO and/or PPO was the Benchmark Health Plan in six areas in 2007 and in seven areas in 2008. National Quality Forum: National Quality Healthcare Award (2007). American Medical Group Association: Acclaim Award (2006) to the HealthPartners Medical Group for its primary care clinic workflow standardization care model process. *See the Series Overview, Findings, and Methods for analytic methodology and explanation of performance recognition. CMS = Centers for Medicare and Medicaid Services; HCAHPS = Hospital Consumer Assessment of Healthcare Providers and Systems; NCQA = National Committee for Quality Assurance (Quality Compass 2008 represents the 2007 measurement year); HMO = health maintenance organization; PPO = preferred provider organization. 18T he  C ommonwealth F und in the workforce requires an “absolute willingness to owned or contracted settings—and engaging provid- reject the status quo and take the risk of pointing out ers in a common measurement and reporting scheme, the flaws in the current system so that across the orga- HealthPartners encourages physicians to improve by nization you have people who are willing to…let go appealing to their professional reputation. “I think of the traditional ways of doing things,” Brainerd said. people take pride in the fact that they’re actively, col- She notes that the workforce is often motivated by a lectively improving diabetes care in their medical “show-me rather than tell-me” approach: Engaging in group here in Minnesota,” said Isham. These efforts a role-playing activity, or seeing a video of a patient are supported by common metrics for performance relating her experience, can be more effective than incentives that health plans and employers have agreed a lecture or memo in helping staff to understand the upon through their participation in the Minnesota human impact of poor quality and to internalize the Community Measurement public reporting initiative goals for improvement. and the Minnesota Bridges to Excellence pay-for-per- formance program (see Appendix A). This market alignment increases the power of Ultimately, health care reform should seek not only incentives while also reducing the burden of measure- to “defragment” health care delivery so that it is less ment. “The reason that it works in Minnesota is that chaotic, but also to develop the infrastructure and people are committed to that framework and they performance framework that health care organizations get something out of it. They get decreased hassle will need to achieve their potential for providing in terms of different measurement frameworks, they optimal care. get more alignment, they get more power for their HealthPartners Chief Medical Officer own incentive programs because they’re pooled with George Isham, M.D. everybody else’s,” Isham said. Despite these strides, the Minneapolis-St. Paul Business Journal recently HealthPartners’ experience suggests that a non- reported that some Minnesota doctors still complain profit health plan market oriented to physician group about the administrative burden created by subtle dif- practice and supported by collaborative measurement ferences in eligibility for incentives.37 and improvement organizations creates a community While participating in this collaborative envi- environment that helps each participant achieve its ronment, HealthPartners has continued to innovate in objectives more effectively. Isham noted that col- developing approaches that are important to achiev- laborating with other health plans and medical groups ing a higher-performing health system. For example, through the Institute for Clinical Systems Improvement the organization is increasingly focused on improving develops common “know-how” and critical mass for health, not just health care, through strategies such as making changes in clinical practice that physicians measurement and intervention on lifestyle risk factors. might otherwise resist or lack the ability to bring about It is also supporting practice redesign both in its own on their own. Giving physicians a forum to develop clinics and in contracted medical groups so that clinical guidelines and improvement strategies that physicians can build internal capacity for managing are recognized throughout the community enables chronic diseases more effectively. These efforts hold HealthPartners to find common ground with those phy- the promise of providing an evolutionary path toward sicians more easily as they pursue common goals, said broader implementation of the primary care “medical Michael Trangle, M.D., associate medical director for home” model. the behavioral health division. HealthPartners’ shift to an open health plan net- By setting ambitious objectives across its work (in which individuals have a choice among con- member population—whether they receive care in tracted medical groups and HealthPartners’ own clin- H ealth P artners : C onsumer -F ocused M ission and C ollaborative A pproach 19 ics) has had both positive and negative consequences, bursement model, which doesn’t reward care coor- according to Brainerd and other system leaders. The dination or cost-efficient practice. The organization change was necessary for market survival and has has adapted to this market dynamic by leading the motivated HealthPartners to innovate and improve as development and use of performance information and it competes for members and patients in a marketplace incentives to help overcome the limitations of fee-for- that values choice at both the health plan and physician service payment. In the future, Brainerd would like to group level. “We want to be the one they choose,” said see a further shift toward episode-based payment to Beth Averbeck, M.D., associate medical director for promote greater accountability for the total care of the quality and primary care. A mixed-model network also patient. Ultimately, health care reform should seek not allows HealthPartners to involve physicians in its inter- only to “defragment” health care delivery so that it is nal medical group in testing innovations before rolling less chaotic, Isham said, but also to develop the infra- them out to contracted groups. structure and performance framework that health care On the other hand, market adaptation has shifted organizations will need to achieve their potential for the organization’s orientation away from its roots in providing optimal care. prepaid practice and toward the fee-for-service reim- For a complete list of case studies in this series, along with an introduction and description of methods, see Organizing for Higher Performance: Case Studies of Organized Health Care Delivery Systems— Series Overview, Findings, and Methods, is available online at www.commonwealthfund.org. 20T he  C ommonwealth F und N otes 8 The plan reports readmission rates representing the percentage of patients who were readmitted 1 T. Shih, K. Davis, S. Schoenbaum et al., Organiz- with fluid and electrolyte imbalance, pneumonia or ing the U.S. Health Care Delivery System for High cardiac-output disorders diagnosed within 90 days, Performance (New York: The Commonwealth Fund 60 days, and 30 days of an initial heart failure hospi- Commission on a High Performance Health System, tal discharge over a two-year period. D. Wehrle and Aug. 2008). S. Bussey, HealthPartners 2008 Clinical Indicators Report (Bloomington, Minn.: HealthPartners, 2008). 2 Information on HealthPartners was synthesized from telephone interviews and e-mail correspon- 9 N. D. Beaulieu, D. M. Cutler, K. E. Ho et al., The dence with the individuals named in the Acknowl- Business Case for Diabetes Disease Management at edgments; from a presentation by George Isham, Two Managed Care Organizations: A Case Study of M.D., to the Commission on a High Performance HealthPartners and Independent Health Association Health Care System, Minneapolis, July 2007; a pre- (New York: The Commonwealth Fund, April 2003). sentation by Donna Zimmerman to the Partnership The authors estimated that, for a patient enrolled for Quality Care Summit, Washington, D.C., March in HealthPartners’ diabetes disease management 2008; and from other presentations and published program for 10 years, the savings in avoided medi- literature (cited below), information on the organi- cal costs would exceed the operating cost of the zation’s Web site, and HealthPartners’ application program by $75 per patient. The economic value for the National Quality Forum’s National Quality of improved quality of life assumed a 1 percent Healthcare Award. improvement in hemoglobin A1c level. The authors concluded that “the magnitude of the difference 3 A summary of findings from all case studies in the between costs and patient benefits is so great that series can be found in D. McCarthy and K. Mueller, we believe, at the social level, the outcomes of these Organizing for Higher Performance: Case Studies comprehensive programs will always be worth the of Organized Delivery Systems—Series Overview, investment needed.” Findings, and Methods (New York: The Common- wealth Fund, 2009). 10 According to data from the National Committee for Quality Assurance’s Quality Compass 2008, Health- 4 S. Silow-Carroll and T. Alteras, Value-Driven Partners ranked in the top 10 percent of commercial Health Care Purchasing: Case Study of Minnesota’s health plans on a measure of antidepressant medica- Smart Buy Alliance (New York: The Commonwealth tion continuation (percentage continuing on an anti- Fund, August 2007). depressant for at least six months), achieving a rate 5 According to the Minnesota Department of Health, of 57.8 percent in 2007 as compared to a national HealthPartners’ 2007 market share was 25 percent average of 46.1 percent. of the fully insured private market by premiums and 11 The plan is investigating the degree to which the 40 percent of HMO enrollment in the state (Min- reduction in medication costs reflects increased nesota Health Care Markets Chartbook, http://www. substitution of generic for brand medications. health.state.mn.us/divs/hpsc/hep/chartbook/index. html). 12 HealthPartners, The “Your Health Potential” Health Assessment. For an example of one predictive algo- 6 R. L. Reece, “EMRs Help Transform Processes of rithm, see: T. L. Pearson, N. P. Pronk, A. W. Tan et Care for HealthPartners’ Physicians: Interview with al., “Identifying Individuals at Risk for the Devel- Kevin Palattao,” Practice Options, July 2004:8–11. opment of Type 2 Diabetes Mellitus,” American 7 K. J. Palattao, “Connecting with Patients: Health- Journal of Managed Care, 2003 9(1):57–66. Partners’ eStrategy Is Good Business,” Group Practice Journal, Oct. 2005 54(9):9–16. H ealth P artners : C onsumer -F ocused M ission and C ollaborative A pproach 21 13 N. Pronk and M. Thygeson, “From Managing Dis- 17 A review of factors that contributed to improved ease to Managing Health,” Group Practice Journal, diabetes care in the HealthPartners Medical Group Oct. 2006: 9–12; N. M. Thygeson, J. Gallagher, K. from 1995 to 2005 identified “drug intensification, Cross et al., “Employee Health at BAE Systems: leadership commitment to diabetes improvement, An Employer-Health Plan Partnership Approach,” greater continuity of primary care, participation in ACSM’s Worksite Health Handbook, Second Edi- local and national diabetes care improvement initia- tion. A Guide to Building Healthy and Productive tives, and allocation of multidisciplinary resources Companies, N. P. Pronk, ed. (Champaign, Ill.: Hu- at the clinic level to improve diabetes care.” J. M. man Kinetics, 2009; Chapter 36). Sperl-Hillen and P. J. O’Connor, “Factors Driving Diabetes Care Improvement in a Large Medical 14 E. H. Wagner, B. T. Austin, and M. Von Korff, Group: Ten Years of Progress,” American Journal of “Organizing Care for Patients with Chronic Illness,” Managed Care, 2005 11:S177­ S185. – Milbank Quarterly, 1996 74(4):511–44. 18 K. Kroenke, R. L. Spitzer, and J. B. Williams, “The 15 Information on the Care Model Process was ob- PHQ-9: Validity of a Brief Depression Severity tained in part from M. McGrail and B. Waterman, Measure,” Journal of General Internal Medicine, “HealthPartners Medical Group: Care Model 2001 16(9):606–13. Process,” Group Practice Journal, Nov./Dec. 2006 55(10):9–20; Anonymous, “Pursuing Perfection: 19 M. C. Hroscikoski, L. I. Solberg, J. M. Sperl-Hillen Report from HealthPartners on Prepared Practice et al., “Challenges of Change: A Qualitative Study Teams,” Improvement Stories (Boston: Institute for of Chronic Care Model Implementation,” Annals of Healthcare Improvement, undated); B. Averbeck Family Medicine, 2006 4(4):317–26. and B. Waterman, “Embedding Reliability in Am- bulatory Care: The Care Model Process,” presented 20 Information in this section was based in part on a at the Institute for Clinical Systems Improvement’s presentation by George Isham, M.D., to the Com- Colloquium on Redesign for Results Quantum mission on a High Performance Health Care Sys- Leaps in Healthcare, St. Louis Park, Minn., May tem, Minneapolis, July 2007. 2007, http://www.icsi.org/colloquium_-_2007/aver- 21 D. Wehrle and S. Bussey, HealthPartners 2008 back.html. Clinical Indicators Report Technical Supplement 16 Researchers who studied the redesign at an early (Bloomington, Minn.: HealthPartners, 2008). stage found a 24 percent improvement in chronic 22 ClearWay Minnesota, Blue Cross and Blue Shield care model implementation overall from 2002 to of Minnesota, and Minnesota Department of Health, 2004 (changes were variable across sites) along Creating a Healthier Minnesota: Progress in with concurrent improvements in quality of care for Reducing Tobacco Use (Minneapolis: Minnesota diabetes, heart disease, and depression. Changes Center for Health Statistics, Sept. 2008). in some care model components correlated with improvement in diabetes control. L. I. Solberg, A. 23 Institute for Clinical Systems Improvement, L. Crain, J. M. Sperl-Hillen et al., “Care Quality Groundbreaking Approach for Improving Depres- and Implementation of the Chronic Care Model: A sion Care Introduced at 10 Minnesota Clinics Quantitative Study,” Annals of Family Medicine, (Minneapolis: ICSI, May 2008). Information about 2006 4(4):310–16. the IMPACT model of depression care, and the evidence base supporting it, can be found at http://impact-uw.org/. 24 Personal communication with Michael Trangle, M.D., associate medical director, HealthPartners Behavioral Health Division, Jan. 2009. Comparable measures of assessment, treatment response, and remission are being collected by the Minnesota Community Measurement Initiative. 22T he  C ommonwealth F und 25 R. A. Williams and J. Flaaten, “Maximizing Phar- 32 B. Averbeck and N. McClure, “Toward Equity: Ad- maceutical Affordability: Systematically Improv- dressing Disparities in Care and Experience,” pre- ing Generic Utilization,” presented at the Medical sented at the American Medical Group Association Group Association Annual Conference, Orlando, 2008 National Conference, Orlando, Fla., March March 7, 2008. 2008. HealthPartners, Strategies to Identify and Reduce Health Disparities (Bloomington, Minn.: 26 S. Lewandowski, P. J. O’Connor, L. I. Solberg et al., HealthPartners, 2007), http://www.healthpartners. “Increasing Primary Care Physician Productivity: A com/files/40901.pdf. Case Study,” American Journal of Managed Care, 2006 12:573–76. Productivity-based pay was insti- 33 National Health Plan Collaborative, “HealthPart- tuted at about the same time as primary care clinics ners: Formalizing Organizational Best Practices for adopted advanced-access scheduling. Language Services Through the Development of a Language Assistance Plan” (Princeton, N.J.: Robert 27 Averbeck and Waterman, “Embedding Reliability in Wood Johnson Foundation, 2008). Ambulatory Care: The Care Model Process.” 34 Dartmouth Atlas Project, http://www.dartmouthatlas.org. 28 R. Bohmer and N. D. Beaulieu, HealthPartners The analysis focused on Medicare patients with one (Cambridge, Mass.: Harvard Business School, of nine chronic conditions who died between 2001 Nov. 1999). and 2005, controlling for differences in patients’ 29 L. I. Solberg, M. C. Hroscikoski, J. M. Sperl-Hillen age, sex, race, and primary chronic diagnosis. et al., “Key Issues in Transforming Health Care 35 Minnesota Department of Health, Adverse Health Organizations for Quality: The Case of Advanced Events in Minnesota: Fifth Annual Public Report Access,” Joint Commission Journal on Quality (St. Paul: Minnesota Department of Health, Jan. and Safety, Jan. 2004 30(1):15–24. At the time as 2009). it instituted advanced-access scheduling, the medi- cal group also introduced a call center for book- 36 Rankings for CMS Hospital Compare clinical ing patient appointments and converted physician topics (heart attack, heart failure, and pneumonia compensation from salary to a system based mainly treatment and surgical care improvement) included on productivity. The researchers reported that these hospitals that reported on all measures and recorded changes both facilitated and challenged the imple- at least 30 patients in each topic. Only results in the mentation of advanced access across the system. top quartile are noted. One HealthPartners hospital (Regions Hospital) was evaluated on clinical top- 30 L. I. Solberg, M. V. Maciosek, J. M. Sperl-Hillen et ics and two (Regions and Westfields Hospitals) on al., “Does Improved Access to Care Affect Utiliza- HCAHPS results. The HCAHPS overall rating of tion and Costs for Patients with Chronic Condi- care means a patient rating of 9 or 10 on a 10-point tions?” American Journal of Managed Care, 2004 scale. The analysis did not include Hudson Hospital 10(10):717–22. since it was not part of the organization during the 31 J. M. Sperl-Hillen, L. I. Solberg, M. C. Hroscikoski time periods studied. et al. , “The Effect of Advanced Access Implemen- 37 N. R. Orrick, “Doctors’ Group Knocks Insurers’ tation on Quality of Diabetes Care,” Preventing Performance Plans,” Minneapolis/St. Paul Business Chronic Disease, Jan. 2008 5(1): A16. The study Journal, Nov. 19, 2007. used multilevel logistic regression to predict per- formance on composite measures of performance controlling for patient age, sex, and coronary artery disease status. H ealth P artners : C onsumer -F ocused M ission and C ollaborative A pproach 23 Appendix A. Collaborative Organizations in Minnesota Several Minnesota organizations are active in promoting improvements in health care delivery through information- sharing and collaborative learning. Among them are: Minnesota Community Measurement (http://www.mnhealthcare.org), a nonprofit collaboration between the Minnesota Medical Association and participating medical groups, consumers, businesses, and health plans in Minnesota and surrounding states. The group’s objectives are to improve care and support quality initiatives, reduce reporting-related expenses, and communicate fair, usable, and reliable findings. It publishes information on the quality of care provided by more than 120 physician practices. Measures have been adapted primarily from the Healthcare Effectiveness Data and Information Set (HEDIS) to align with clinical guidelines established by the Institute for Clinical Systems Improvement. The group also has developed composite measures of optimal care for diabetes and coronary artery disease. The Institute for Clinical Systems Improvement (ICSI) (http://www.icsi.org), which promotes evidence-based prac- tice and the redesign of the health care delivery system through the development and dissemination of consensus- driven clinical guidelines and payment models. ICSI also facilitates stakeholders’ collaboration in the development of patient- and value-centered models of care for women’s health, preventive care, and various health conditions. ICSI supports providers in transforming their practices and implementing quality improvement activities through col- laborative learning. ICSI’s membership includes more than 50 medical groups (physician group practices) located in Minnesota and adjacent states, and six health plans that sponsor the organization financially. Bridges to Excellence (BTE) (http://bridgestoexcellence.org), a national collaboration that recognizes and rewards health care providers who reengineer their practices to deliver care consistent with the Institute of Medicine’s aims for the health system. The program is active in 20 states; Minnesota’s effort is led by the Buyer’s Health Care Action Group (http://www.bhcag.com), a coalition of private and public employers, in collaboration with health plans, the Minnesota Medical Society, providers, ICSI, and Minnesota Community Measurement. The BTE pay-for-perfor- mance model focuses on reducing defects, misuse, and waste in health care. Jim Reimann, an independent consultant, kindly shared perspective on the Minnesota market environment. Incentives are based upon publicly reported data consistent with the efforts of the National Committee for Quality Assurance (http://www.ncqa.org) and Minnesota Community Measurement. Appendix B: HealthPartners Health Goals 2010 Health Goals 2010 High-Level Summary for September 2008 Sept 2008 Infra-structure Relative Position Health Goal Results Improvement to Competitors 1 Customers receive amazingly easy to use care, coverage, and service (E) 2 Customers receive maximum quality and affordability in health care (E/H) 3 Patients and members receive equitable care and service (H) 4 Customers feel they are treated as individuals (E) Patients and members have the information they need and understand to be effective decision- 5 makers (E/H) 6 Customers are incented and supported for self care and healthy behaviors Customers experience perfect transitions among clinicians, patients, family, payers, and 7 community support (E/H) 8 Customers receive evidence-based care, creating an efficient path to recovery (H) 9 Members and patients will have help to be healthy (H) 10 Members and patients will have help with health/life transitions (H) 11 Members and patients will live well with acute and chronic illness and disease (H) Diabetes Care Vascular Disease Cancer Care Bone & Joint Disease Care Depression Care Asthma Care 24T he  C ommonwealth F und 12 Members and patients will be safe (H) Health Goals 2010 High-Level Summary for September 2008 Health Goal Progress Key: Goal achieved / infrastructure in place with full spread / position relative to competition strong Positive performance trend / infrastructure in place / position relative to competition good Stable performance / infrastructure in design or early implementation / position relative to competition is neutral H ealth P artners : C onsumer -F ocused M ission Measurement development in progress or unstable performance / early infrastructure design in process / position relative and to competition is weak Performance measurement not yet established / infrastructure in the planning stage / not applicable C ollaborative A pproach 25 Appendix C: HealthPartners Performance Incentive Program for Contracted Medical Groups Partners in Excellence – 2009 Primary Care Targets Primary Care Groups: > 1,500 Pi C G 1 500 Excellent Superior Pending MNCM/ICSI Pending MNCM/ICSI decision on HbA1C decision on HbA1C level level MNCM Optimal Diabetes Care DDS MNCM Optimal Vascular Disease Care DDS 55% 60% Optimal Depression Care 45% 50% Evidence-based Cervical Cancer Screening 45% 55% Alcohol Assessment 90% 95% Preventive Services – Adult 93% 98% Generic Drug Use 75% 80% Low back Pain Composite Measure 15% 25% “Informed about your care” Top Band Patient Satisfaction “Talked about pros & cons for any choices for Top Band your treatment or health care” HP Innovations in Health Care Award HP Innovations in Shared Decision Making (tech specs) 26T he  C ommonwealth F und Partners in Excellence – 2009 Primary Care Targets Primary Care Groups between 100 - 1,500 Excellent Superior Pending Pending MNCM/ICSI MNCM Optimal Diabetes Care DDS MNCM/ICSI decision on HbA1C decision on HbA1C level H ealth P artners : C onsumer -F ocused M ission level and MNCM Optimal Vascular Care DDS 55% 60% Evidence-based Cervical Cancer Screening 45% 55% MNCM Breast Cancer Screening 80% 85% Generic Drug Use 75% 80% C ollaborative A pproach HP Innovations in Health Care Award HP Innovations in Shared Decision Making (tech specs) 27 Partners in Excellence – 2009 Pediatric Targets Pediatrics Excellent Superior Part 1: BMI Assessment – Pediatric 80 % 85 % Part 2: Preventive Services - Pediatric 75 % 80 % MNCM Pediatric Immunization Combo 3 90 % 95 % “Informed about your care” y Top Performer Patient Satisfaction: “Talked about pros & cons for any choices Top Performer for your treatment or health care” Generic Drug Use 70 % 75% HP Innovations in Health Care Award HP Innovations in Shared Decision Making (tech specs) 28T he  C ommonwealth F und Partners in Excellence – 2009 Specialty Targets Specialty 2009 Measure Excellent Superior CMS Heart Failure Re-admissions 30 day < or = 10% 30 day < or = 5% Generic Drug Use 75% 80% Cardiology . “Informed about your care” 75% 80% H ealth P artners : C onsumer -F ocused M ission Patient Satisfaction “Talked about pros & cons for any choices for Top Band your treatment or health care” and DVT/PE Infection Measure 0.5 index or lower (Index rate*) “Informed about your care” 80% 85% Ortho Patient Satisfaction “Talked about pros & cons for any choices for Top Band your treatment or health care” Generic Drug Use 75% 80% DVT/PE Infection Measure 0.5 index or lower (Index rate*) C ollaborative A pproach Alcohol Assessment 90% 95% OB/GYN “Informed about your care” 87% 92% Patient Satisfaction “Talked about pros & cons for any choices for Top Band your treatment or health care” Index rate: Rate based on total network average. “Provider group had ≤0.5 (1/2) lower complications compared to total network.” 29 Partners in Excellence – 2009 Specialty Targets Specialty p y 2009 Measure Excellent Superior p Generic Drug Use 75% 80% “Informed about your care” 71% 76% ENT Patient Satisfaction “Talked about pros & cons for any choices for Top Band y your treatment or health care” Behavioral Generic Drug Use 70% 75% Health Optimal Depression Care 55% 60% Initial 75 % See Combined Measure PT Functional Assessment/Oswestry F ti lA t/O t Longitudinal L it di l 35 % See Combined Measure Compliance 75 % Initial and Combined N/A 35 % Longitudinal All HP Innovations in Health Care Award HP Innovations in Shared Decision Making (tech spec) Index rate: Rate based on total network average. “Provider group had ≤0.5 (1/2) lower complications compared to total network.” 30T he  C ommonwealth F und H ealth P artners : C onsumer -F ocused M ission and C ollaborative A pproach 31 A bout the A uthors Douglas McCarthy, M.B.A., president of Issues Research, Inc., in Durango, Colorado, is senior research adviser to The Commonwealth Fund. He supports The Commonwealth Fund Commission on a High Performance Health System’s scorecard project, conducts case studies on high-performing health care organizations, and is a contributing editor to the bimonthly newsletter Quality Matters. He has more than 20 years of experience working and consulting for government, corporate, academic, and philanthropic organizations in research, policy, and operational roles, and has au­hored or coauthored reports and peer-reviewed articles on a range of health care– t related topics. Mr. McCarthy received his bachelor’s degree with honors from Yale College and a master’s degree in health care management from the University of Connecticut. During 1996–1997, he was a public policy fellow at the Hubert H. Humphrey Institute of Public Affairs at the University of Minnesota. Kimberly Mueller, M.S., is a research assistant for Issues Research, Inc., in Durango, Colorado. She earned an M.S. in social administration from the Mandel School of Applied Social Sciences at Case Western Reserve University and an M.S. in public health from the University of Utah. A licensed clinical social worker, she has over 10 years’ experience in end-of-life and tertiary health care settings. She was most recently a project coordinator for the Association for Utah Community Health, where she supported the implementation of chronic care and quality improvement models in community-based primary care clinics. Ingrid Tillmann, M.S., M.P.H., is an independent consultant who conducts research and analysis on health policy issues. She was formerly a senior vice president of the Economic and Social Research Institute in Washington, D.C. The primary focus of her 20 years in health policy has been the promotion of quality and accountability in health care. She has written on the subjects of health system reform, quality improvement initiatives, performance measurement, value-based health care purchasing, disease and disability management, alternative delivery models for high-risk patients, and the use of information technology in health care. Ms. Tillmann received an M.S. in nursing from Pace University and an M.P.H. in maternal and child health from the University of North Carolina at Chapel Hill. A cknowledgments The authors gratefully acknowledge the following individuals who kindly provided information for the case study: Mary Brainerd, HealthPartners’ CEO; George Isham, M.D., medical director and chief health officer; Beth Waterman, R.N., M.B.A., vice president of primary care and clinic operations; Beth Averbeck, M.D., associate medical director for primary care; Michael Trangle, M.D., associate medical director, Behavioral Health Division; Karen Lloyd, senior director of behavioral health strategy and operations; Nico Pronk, Ph.D., vice president and health science officer, JourneyWell; and Lief Solberg, M.D., clinical director for care improvement research at the HealthPartners Research Foundation. The authors also thank the staff at The Commonwealth Fund for advice on and assistance with case study preparation. Editorial support was provided by Joris Stuyck. This study was based on publicly available information and self-reported data provided by the case study institution(s). The Commonwealth Fund is not an accreditor of health care organizations or systems, and the inclusion of an institution in the Fund’s case studies series is not an endorsement by the Fund for receipt of health care from the institution. The aim of Commonwealth Fund–sponsored case studies of this type is to identify institutions that have achieved results indicating high performance in a particular area of interest, have undertaken innovations designed to reach higher performance, or exemplify attributes that can foster high performance. The studies are intended to enable other institutions to draw lessons from the studied institutions’ experience that will be helpful in their own efforts to become high performers. It is important to note, however, that even the best-performing organizations may fall short in some areas; doing well in one dimension of quality does not necessarily mean that the same level of quality will be achieved in other dimensions. Similarly, performance may vary from one year to the next. Thus, it is critical to adopt systematic approaches for improving quality and preventing harm to patients and staff.