- Email: [email protected]
What Are the Facilitators and Barriers in Physician Organizations’ Use of Care Management Processes?
Joint Commission
Journal on Quality and Safety
Performance Improvement
What Are the Facilitators and Barriers in Physician Organizations’ Use of Care Management Processes?
Thomas Bodenheimer, M.D. Margaret C. Wang, Ph.D. Thomas G. Rundall, Ph.D. Stephen M. Shortell, Ph.D. Robin R. Gillies, Ph.D. Nancy Oswald, Ph.D. Lawrence Casalino, M.D., Ph.D. James C. Robinson, Ph.D.
ajor deficiencies have been reported in the management of common chronic conditions. For example, some 66% of people with high blood pressure are inadequately treated.1 Half of the patients hospitalized with congestive heart failure (CHF) are readmitted within 90 days,2 and 63% of adults with diabetes have glycohemoglobin (A1C) levels > 7.0; 52% have total serum cholesterol levels of ⱖ 200.3 Literature reviews and meta-analyses have begun to elucidate which practice innovations can improve processes or outcomes of chronic illness care. These innovations include disease registries,4 practice guidelines in conjunction with physician education and reminder systems,5–7 performance feedback,7,8 case management,9 and patient self-management education.10 These chronic illness improvement measures can be called care management processes (CMPs).11 Wagner and associates have proposed a Chronic Care Model that combines these processes into a guide for chronic care improvement.12,13 Some studies have examined the factors that serve as the facilitators and barriers in practice innovation in high-performing organizations,14–17 but these factors have not been explored in a sample of higher- and lowerperforming physician organizations. In 2000–2001 we conducted 158 interviews in 15 physician organizations to determine the major facilitators and barriers affecting the adoption of CMPs for asthma, CHF, diabetes, and depression.
M
September 2004
Article-at-a-Glance Background: Care management processes (CMPs) such as disease registries, reminder systems, performance feedback, case management, and self-management education can improve chronic illness care, yet 50% of physician organizations have instituted few if any CMPs. Methods: Site-visit interviews were conducted with 158 leaders at 15 physician organizations, with 3 organizations (1 large medical group, 1 small medical group, and 1 independent practice association [IPA]) chosen randomly in most cases in each of five communities. Results: Seven of the 15 organizations had implemented CMPs minimally or not at all. CMPs were most common for diabetes and least common for depression; no IPAs had achieved significant CMP implementation for any of the conditions. The two most commonly mentioned facilitators were strong leadership and organizational culture valuing quality. The top five barriers were poor financial situation, reimbursement that does not reward high quality, inadequate information technology, physician resistance, and physicians being too busy. Discussion: Strong leadership and a quality-valuing culture are important facilitators of improved chronic care, but if insurers do not reward chronic care improvement, it is unlikely that CMPs will become permanently institutionalized in physician organizations.
Volume 30 Number 9
Copyright 2004 Joint Commission on Accreditation of Healthcare Organizations
505
Joint Commission
Journal on Quality and Safety
This qualitative study was a part of a larger project, the National Study of Physician Organizations and the Management of Chronic Illness (NSPO), which consisted of a telephone survey of 1,040 medical groups and independent practice associations (IPAs) with 20 or more physicians. The survey indicated that 15% of physician organizations did not use CMPs for these four chronic conditions and that 50% had instituted little if any chronic care improvement. Factors associated with the use of more CMPs were greater clinical information technology (IT) capability and external quality incentives such as public recognition or financial bonuses for better performance.11 Whereas the quantitative NSPO uses a wideangle lens to illuminate CMP implementation in physician organizations, the qualitative part of the NSPO employs a zoom lens to focus on the details of CMP facilitators and barriers.
Methods Sample Between September 2000 and October 2001, seven of the members [all authors but M.C.W.] of the NSPO research group conducted in-depth site visits at 15 physician organizations, chosen in the following manner. The 12 communities studied by the Community Tracking Study of the Center for Studying Health System Change18 were ranked according to the level of 1998–1999 health maintenance organization (HMO) penetration. The top six communities were chosen, but one was dropped because of difficulty gaining access to physician organizations. We picked Community Tracking Study communities because a wealth of information about these health care markets has been collected. For each of the five communities selected, we used a stratified random sampling procedure to choose three physician organizations. In each community, the universe of physician organizations with 20 or more physicians was identified, and information on the type of organization—medical group or independent practice association (IPA)—and the number of physicians was obtained. A medical group was defined as an organization composed of physicians who all belong to one practice, operating in one office or at many locations. An IPA was defined as an organization through which
506
September 2004
physicians contract with managed care plans and in which physicians practice solo or in small groups independently from each other. We randomly selected one physician organization within each of the following strata: ■ Large medical groups (more than 50 physicians) ■ Smaller medical groups (20–50 physicians) ■ IPAs or IPA/medical group hybrids In two communities where physician organizations were reluctant to participate in the study, we selected a similar mix of organizations on the basis of a conveniencesampling strategy.
Interviews At each of the 15 physician organizations studied, 60–90-minute semistructured interviews were conducted with each of the following personnel: chief executive officer, medical director, physician board member, chief financial officer, chief information officer, quality improvement (QI) manager, and physicians active in treating patients with asthma, CHF, diabetes, and depression. Not all of the personnel were relevant or available for interviews at all 15 sites. The interview was intended to assess the extent to which CMPs had been adopted for asthma, CHF, diabetes, and depression and elicited factors that constituted facilitators for or barriers to the implementation of CMPs. The interviews asked about the following CMPs: ■ Chronic disease registries ■ Practice guidelines in conjunction with physician education and reminder systems ■ Performance feedback ■ Case management ■ Patient self-management education One or two members of a three-person team conducted each interview at each site; most physician interviewees were paired with physician interviewers. Each of the 15 site-visit teams prepared a report on the basis of detailed interview notes. Respondents’ answers were tape recorded and recorded by hand on an interview report form. For each physician organization, the sitevisit team prepared a report consolidating the information obtained from all informants. To the extent possible, verbatim responses to questions were incorporated in the site-visit report.
Volume 30 Number 9
Copyright 2004 Joint Commission on Accreditation of Healthcare Organizations
Joint Commission
Journal on Quality and Safety
Table 1. Stages of Implementation of Care Management Processes at 15 Organizations for Four Chronic Illnesses* Organization A1 A2 A3 B1 B2 B3 C1 C2 C3 D1 D2 D3 E1 E2 E3
Organization Type CMP Adoption Asthma Diabetes CHF MG More advanced Aware Implementation Implementation Hybrid More advanced Aware Implementation Implementation MG More advanced Implementation Implementation Institutionalized MG More advanced Implementation Implementation Aware IPA Minimal Pre-aware Pre-aware Pre-aware MG Minimal Pre-aware Aware Pre-aware MG More advanced Pre-aware Implementation Institutionalized MG More advanced Implementation Implementation Aware MG Minimal Pre-aware Aware Aware IPA Minimal Pre-aware Pre-aware Pre-aware IPA Minimal Aware Aware Aware Hybrid More advanced Pre-aware Implementation Implementation MG More advanced Institutionalized Institutionalized Institutionalized Hybrid Minimal Aware Aware Aware IPA Minimal Aware Aware Aware
Depression Pre-aware Pre-aware Aware Aware Pre-aware Pre-aware Aware Aware Pre-aware Pre-aware Pre-aware Aware Implementation Pre-aware Pre-aware
* A,B,C,D, and E represent the five communities visited. CMP, care management process; CHF, congestive heart failure; MG, medical group; IPA, independent practice association; hybrid = mix of MG and IPA. The four stage-of-implementation categories are listed below; minimal CMP adoption entails pre-aware or aware categories for all conditions and more advanced CMP adoption entails implementation or institutionalization for at least one condition.
■ Implementation (ⱖ two CMPs have been developed)
Content Analysis The statements in each site visit report were coded, and qualitative content analyses were performed to determine the extent of CMP adoption, assess IT, and identify barriers and facilitators of CMP implementation. Content analyses were independently completed by three members of the study team [T.B., M.C.W., T.R.], with two members analyzing interview data for each site visit. These coders agreed on the classification of interviewees’ responses into thematic categories in > 80% of classification decisions; inconsistencies were resolved in discussions between the two coders. Given the small number of site visit reports analyzed, manual content analytic techniques19 were chosen. The extent to which the organizations had developed CMPs (Table 1, above) was evaluated using a modified stages-of-implementation analysis20 with the following four categories: ■ Pre-awareness (no thinking or planning about CMPs) ■ Awareness (believe CMPs are needed and may have implemented one CMP)
September 2004
■ Institutionalization (ⱖ two CMPs are permanently
embedded in the organization) It was determined that the introduction of only one CMP was not sufficient to locate a physician organization in the implementation stage; two CMPs were considered necessary for an organization to be categorized as a serious adopter of care management. The sophistication of the organizations’ IT was estimated using the following three categories: ■ Computerized billing and scheduling only (level 1) ■ An electronic medical record (level 2) ■ An electronic medical record plus at least one disease registry or other CMP function such as reminder systems (level 3) Members of the research team shared expectations that medical groups would report more advanced implementation of CMPs than IPAs and that organizations with greater IT would have more CMPs. Our analyses confirmed the former preconception but not the latter.
Volume 30 Number 9
Copyright 2004 Joint Commission on Accreditation of Healthcare Organizations
507
Joint Commission
Journal on Quality and Safety
Results Basic CMP Data Of the 15 physician organizations visited, 8 were medical groups, 4 were IPAs, and 3 were hybrid (mixed medical group and IPA) organizations. Four of the physician organizations had long histories (founded in 1970 or before), whereas 11 were relatively new (founded in 1985 or after). For asthma, 3 organizations had implemented CMPs but these practices were not sufficiently engrained in the organization to ensure their permanence; 1 had institutionalized asthma CMPs, permanently embedding these practices into its clinical structure. Six organizations were in the pre-awareness stage, not having considered developing asthma CMPs, and 5 were in the awareness stage, considering or starting one CMP (Table 1). For CHF, 6 out of 15 organizations had implemented or institutionalized CMPs. Diabetes CMPs were more prevalent, with 8 organizations having achieved implementation or institutionalization stages. The condition for which CMPs were least adopted was depression. Only 1 organization had implemented depression CMPs, whereas 9 were in the pre-awareness stage and 5 were in the awareness stage. The five communities demonstrated a variety of patterns regarding CMP adoption. Each community had 12 CMP-adoption data points—three organizations each reporting on four chronic conditions. The three physician organizations studied in one of the communities had implemented or institutionalized CMPs in 7 out of 12 opportunities; organizations in two communities had implemented CMPs in only 2 out of 12 possibilities. In another of the communities, one of the three physician organizations demonstrated highly advanced CMP development, whereas the other two sites had minimal adoption of CMPs. The community with the greatest level of CMP development is an elite center of medical education populated by several innovative health care institutions. Seven of the 15 organizations, including all four IPAs, demonstrated minimal CMP adoption; they were in the pre-awareness or awareness stage for all four chronic conditions. All of these organizations were relatively young (founded after 1984), whereas all four of the older organizations had relatively advanced CMP development.
508
September 2004
Eleven of the 15 sites had level 1 IT capability: whereas, 4 sites had an electronic medical record (levels 2 and 3). Only 2 sites—older, well-respected physician organizations associated with group model HMOs—had level-3 IT capability, including not only an electronic medical record but also reminder systems, performance feedback systems, or chronic disease registries. These two sites also had the best developed CMPs.
Facilitators and Barriers Table 2 (page 509) and Table 3 (page 510) show the number of organizations whose interviewees mentioned specific facilitators and barriers affecting the adoption of CMPs. The two most commonly mentioned facilitators were strong leadership and an organizational culture that valued quality. Other facilitators mentioned were IT, having physician champions, and capitation payment. The top five barriers were lack of resources, reimbursement that does not reward high quality, inadequate IT, physician resistance, and physicians being too busy. The facilitators and barriers included a mixture of intraorganizational (leadership, culture, and IT) and broader health system factors (mode of reimbursement). Some facilitators and barriers can be viewed as two sides of the same coin. For physician organizations succeeding at CMP implementation, the facilitators were good leadership, a quality-oriented culture, and advanced IT, whereas for organizations without CMPs, poor leadership, the lack of a quality culture, and poor IT were barriers. In the quantitative NSPO, IT capability was associated with greater use of CMPs.11 In the qualitative study reported here, only three of eight organizations more advanced in CMPs had advanced IT. One organization with well-developed IT had adopted virtually no CMPs. Although sophisticated IT makes chronic care improvements easier to achieve, it appears that improvements can be made without advanced IT. Moreover, having IT does not guarantee greater use of CMPs. Acknowledgement of the role of leadership and culture varied markedly between the eight sites that implemented CMPs and the seven sites with minimal CMP adoption. All eight sites with more advanced CMPs cited leadership and/or physician champions as facilitators; of
Volume 30 Number 9
Copyright 2004 Joint Commission on Accreditation of Healthcare Organizations
Joint Commission
Journal on Quality and Safety
Table 2. Facilitators Mentioned by Interviewees* Organization A1 A2 A3 B1 C1 C2 D3 E1 B2 B3 C3 D1 D2 E2 E3 Total
CMP Adoption More Advanced More Advanced More Advanced More Advanced More Advanced More Advanced More Advanced More Advanced Minimal Minimal Minimal Minimal Minimal Minimal Minimal
Leadership Quality Culture Advanced IT Champions X X X X X X X X X X X X X X X X X X X X X X X X X X X
Have Financial Risk X X X
X X
9
8
6
6
X 4
* A,B,C,D, and E represent the five communities visited. CMP, care management process; IT, information technology.
the seven less-advanced sites, only two cited those facilitators. Of the eight more advanced sites, six mentioned quality culture as a facilitator; of the seven less advanced sites, only two mentioned culture as facilitating. The issue of busy physicians yielded surprising findings. Seven of the eight more advanced sites noted that overworked physicians was a barrier; whereas, only one less advanced site mentioned overworked physicians as a barrier. This suggests that good leadership and culture that values quality might trump physician overwork in determining whether organizations implement CMPs.
Case Studies: Interaction of Facilitators and Barriers Case studies for 4 of the 15 organizations elucidate the interaction of facilitators and barriers (organization names are fictitious).
Southern IPA: An IPA with Little Interest in CMP Development Southern IPA includes 400 physicians who practice in their private offices. The IPA’s main function is to contract with health plans on behalf of its physicians. The
September 2004
IPA itself has a tiny budget and two part-time employees. Southern IPA does not gather clinical data about its physicians. Chronic care improvement is not part of Southern IPA’s strategic plan. Many Southern physicians do not trust clinical practice guidelines nor performance feedback reports sent by health plans. One physician noted that practice algorithms assume that “all the patients who have asthma or high blood pressure are the same. As if we were all identical twins.” When asked about a diabetes registry, a physician replied, “I receive capitation payments. Why should I send patients letters and contact them to make appointments?” When asked about a CHF program, a physician answered that it does not matter if CHF patients go into the hospital because he would not get paid under capitation for hospital visits. One health plan pays a small bonus to physicians with good Health Plan Employer Data and Information Set (HEDIS) scores, but the amount is insignificant and physicians pay no attention. Southern IPA’s culture is to shield physicians from intrusion by health plans. The physician leadership focuses on health plan contracts and has little interest in care management. One leader admitted that although he
Volume 30 Number 9
Copyright 2004 Joint Commission on Accreditation of Healthcare Organizations
509
Joint Commission
Journal on Quality and Safety
Table 3. Barriers Mentioned by Interviewees* Organization A1 A2 A3 B1 C1 C2 D3 E1 B2 B3 C3 D1 D2 E2 E3 Total
CMP Adoption More Advanced More Advanced More Advanced More Advanced More Advanced More Advanced More Advanced More Advanced Minimal Minimal Minimal Minimal Minimal Minimal Minimal
In Poor Financial Plans Don’t Pay Shape for Quality X X X X X X X X X
Physician Resistance
Poor IT X X
X X X
X X X X
X X X X X X X 11
X X
X X
X
X X
X 9
9
Physicians Are Too Busy X X X X X X X
X X X X X X 9
8
* A,B,C,D, and E represent the five communities visited. CMP, care management process; IT, information technology.
would welcome organized care for chronically ill patients, there are no incentives, no clinical meetings, and no interest. When asked about CMPs, a front-line physician said, “All these things make more work for me, so I don’t do them.” In 2001, when our interviews were conducted, Southern IPA’s community was beset with conflict, according to the Community Tracking Study. Hospitals were consolidating and obtaining better contracts with health plans, which in turn became less profitable. As a result, health plans attempted to reduce physician payments, which escalated physician anger. This was an environment unfriendly to QI. The case of Southern IPA demonstrates how barriers build on one another. The physician leadership was not interested in QI, thereby reinforcing a culture of physician independence. The NSPO demonstrated that CMPs are associated with clinical information systems, better health plan contracts for quality, and external recognition of quality care. Even if a quality champion emerged at Southern IPA, it would have been difficult to institute CMPs because no clinical information systems existed,
510
September 2004
physician organizations did not receive recognition for quality, and health plans did not significantly reward quality. Because the IPA’s leadership wanted health plan dollars to flow to physicians, the IPA itself had almost no budget and was unable to build a clinical information system or to institute CMPs.
Eastern Health System: A Medical Group with Advanced CMP Implementation Eastern Health System is a large integrated delivery system whose reputation enables it to attract excellent physicians. According to the Community Tracking Study, the health care market in Eastern’s community was in turmoil at the time of our interviews, but Eastern’s stable patient enrollment and employed-physician workforce shielded it from neighboring bankruptcies and dislocations. For a decade, Eastern’s leaders had been at the forefront of chronic care improvement and had built an IT system that facilitates CMP development. Care management enjoys a distinct budget and expert physicians are paid to work on CMPs. The financial incentive inherent in the capitated system encourages programs for
Volume 30 Number 9
Copyright 2004 Joint Commission on Accreditation of Healthcare Organizations
Joint Commission
Journal on Quality and Safety
such conditions as asthma and CHF, which can reduce hospital admissions and cut Eastern’s expenditures. Organizational leaders take pride that the culture of “doing the right thing” is more important than the prospect of financial savings in developing CMPs. Eastern’s diabetes program illustrates a well-designed CMP. A registry of several thousand diabetics has been compiled, giving physicians computerized access to their diabetic panel. All diabetic patients receive a letter on their birthday indicating which screening exams they are missing; they are also called by phone. Diabetes classes are available. For every diabetic visit, physicians receive a reminder prompt including laboratory and medication data and missing screening tests. Patients whose conditions are poorly controlled are referred to case managers. Practice guidelines are available to all physicians on the intranet, and diabetes care is discussed at some monthly meetings. Physicians receive quarterly feedback reports on their diabetic panels. Because it is both a health plan and a medical group, Eastern Health System receives public recognition for quality through the National Committee for Quality Assurance, although it does not receive better contracts for quality. Whereas CMPs are highly developed, the extent to which they are used by front-line physicians is unclear. Physicians are extremely busy and may view reminder prompts and performance feedback reports as more work for no reward. Whereas Eastern has the leadership, culture, financial resources and IT to implement CMPs, the barrier of a lack of physician time limits the CMPs’ influence on each physician-patient encounter.
Western Medical Group: A Medical Group Attempting to Initiate CMPs Western Medical Group is a 60-physician primary care group whose leaders want to develop CMPs but have not yet succeeded in doing so. Western recently rolled out an electronic medical record but has not yet used it to produce clinical data. A medical director explained, “Our investment in computers has been enormous. Now we must reap the benefit by using it to improve care.” According to the group’s leaders, Western Medical Group’s organizational culture emphasizes physician independence, decentralization of decision making, and
September 2004
low group cohesiveness. Implementing centralized CMPs would be expected to meet considerable resistance. One senior manager said, “Some doctors are concerned that if there is a practice guideline and they do not follow it exactly they could be sued.” The leadership has informed physicians that Western will be monitoring several diabetes and CHF process measures and will pay physicians bonuses if they score satisfactorily on those measures. Leaders hope that business will trump culture. The Community Tracking Study places Western Medical Group in a market formerly with medium-high HMO and capitation penetration, more recently retrenching toward weaker managed care and fee-forservice reimbursement. Because some large physician organizations have failed financially, Western’s culture has more strongly embraced physician productivity to improve its finances by increasing the volume of visits. Western has been losing money and is unable to hire a QI staff person. In summary, Western Medical Group has leaders interested in QI but would need a culture shift to sell CMPs to many of its physicians. A robust IT system has not been sufficient to institutionalize registries, reminders, and performance feedback. Western’s shaky financial situation has made it impossible to hire a quality manager to activate those IT functions and sell them to physicians. Lack of health plan bonuses for improved quality and no public reporting of quality indicators adds to an environment unfavorable for CMP development. Leadership is a necessary but not a sufficient condition for QI; cultural and financial barriers can stymie even the most committed leaders.
Northern Clinic: A Medical Group with Substantial CMP Development That Is Losing Some of Its Gains Northern Clinic is a highly respected multispecialty clinic in a market characterized by the Community Tracking Study as highly competitive among health plans and physician organizations, with capitation giving way to fee for service. In contrast to Western Medical Group, Northern Clinic has implemented a number of CMPs but because of financial pressures some of these innovations have dropped from clinicians’ radar screens. Though a quality culture has a long
Volume 30 Number 9
Copyright 2004 Joint Commission on Accreditation of Healthcare Organizations
511
Joint Commission
Journal on Quality and Safety
history at Northern, some leaders are cynical about how deep that culture penetrates. Whereas one leader claimed, “We will not sacrifice quality for cost,” another complained, “Evidence-based medicine is not a required course here; it is an elective.” The external environment is having a major impact on Northern Clinic’s commitment to chronic care quality. One physician executive admitted, “If care management practices don’t result in cost reduction, they should probably be eliminated.” Another added, “The movement away from capitation to fee for service makes the need for implementing CMPs less urgent” and explained, “The organization has experienced several difficult years financially, making it hard to allocate FTEs [full-time equivalents] for needed CMP work.” At Northern Clinic, a clash has developed between the culture of quality and the reality of financial solvency.
Discussion The qualitative component of the NSPO adds to the findings of the quantitative survey of 1,040 physician organizations.11 The large survey found that external incentives and IT are associated with the adoption of CMPs but did not examine issues of organizational leadership and culture. Leaders of the 15 physician organizations receiving in-depth site visits frequently confirmed that the presence or absence of quality-enhancing incentives from payers and adequate IT were important facilitators or barriers to improved chronic care. The in-depth interviews provided the additional insight that strong leadership and a quality-oriented physician culture are powerful facilitators for chronic care improvement. Similar conclusions were reached by several other studies.14,16,17,21 In the 15 sites, the majority of facilitators to CMP adoption appear to be factors internal to the organization rather than conditions created by the financial environment impinging on the organization. The five most common facilitators were mentioned 33 times by at least one interviewee at a site. Of these 33 mentions, 29 involved factors internal to the organization—leadership and champions, a culture valuing quality, and adequate IT; only 4 noted an external factor—the organization’s assumption of financial risk. Leadership and a quality culture may be the primary facilitator; these traits corre-
512
September 2004
lated strongly with a high level of CMPs in the 15 sites. In contrast, external factors appeared to have greater importance as barriers to chronic care improvement; the external environment was cited 20 times out of 47 mentions of barriers. In organizations with strong leadership and a qualityfocused culture, the most frequently mentioned barriers—inadequate finances, payers not rewarding quality, inadequate IT, and resistance or overwork of physicians—did not prevent the adoption of CMPs. Strong leadership and quality culture can sometimes overcome barriers to CMP development. Care management implementation (CMI) can be simplified to a formula: CMI = D × A (desire × ability to get it done). Both D and A are needed for success. The desire is expressed by the facilitators of strong leadership and a quality-valuing culture. The ability to get it done can be thwarted by the barriers discussed above. One might hypothesize that the facilitators of strong leadership and a quality-oriented culture (the desire) trump the barriers in initiating CMPs, while the major barriers (the inability to get it done) can overcome the facilitators in sustaining CMPs over time. Longitudinal data collection would be needed to explore this hypothesis. One area requiring further research is that of physician overwork. Sites mentioning physician overwork as an important barrier tended to be strong CMP adopters. Do some CMPs (for example, case management) reduce physician work? Do CMPs create more physician work, while strong leadership and a quality-oriented culture overcome physician overwork in the mix of factors determining CMP adoption? Can CMPs be sustained in an organization of overworked physicians? This article has several limitations. First, given the small sample size, the study should be considered exploratory and suggestive of the challenges involved in implementing chronic care improvement. Second, the study did not measure clinical outcomes but examined clinical processes, some of which have been correlated with improved outcomes.4–10 Third, the pattern of findings cannot address the issue of causality. For example, does sophisticated IT create CMP adoption, or does strong leadership create the development of both sophisticated IT and CMPs?
Volume 30 Number 9
Copyright 2004 Joint Commission on Accreditation of Healthcare Organizations
Joint Commission
Journal on Quality and Safety
Conclusion The information gathered in this study suggests several health system policy and practice changes that might encourage chronic care improvement. In terms of practice changes, health science schools and residency programs should add physician leadership training and education on the Chronic Care Model to existing curricula. Redesigning primary care and delegating routine chronic care management functions to nonphysician personnel may help address the barrier of physician overwork.22 Physician organizations need financial and technical assistance in developing IT. In a demonstration program recently initiated through Medicare, such assistance will be offered to physician organizations that are initiating computerized quality performance measurement.23 In terms of policy, reforms needed in the external environment include the devising of reimbursement formulas by Medicare, Medicaid, and private insurers which pay for care management processes and reward superior performance. Early pay-for-performance programs are sprouting up across the United States.24 The consequences of the trend from capitation payment back to fee for service need to be examined because fee-for-service payment promotes a culture valuing quantity rather than quality. It is essential to have inspiring leaders to initiate chronic care improve-
ments, but sustaining these improvements requires insurers to pay for them. J Grant support was provided by the Robert Wood Johnson Foundation (grant no. 038690). The authors also acknowledge the fellowship support of Margaret Wang provided by the Health Research and Educational Trust (HRET).
Thomas Bodenheimer, M.D., is Adjunct Professor, Department of Family and Community Medicine, University of California at San Francisco, San Francisco. Margaret C. Wang, Ph.D., is a Researcher, School of Public Health, University of California at Berkeley, Berkeley, California, Thomas G. Rundall, Ph.D., is Professor, Division of Health Policy and Management, School of Public Health, University of California at Berkeley. Stephen M. Shortell, Ph.D., is Dean, School of Public Health, University of California at Berkeley. Robin R. Gillies, Ph.D., is Project Director, Division of Health Policy and Management, School of Public Health, University of California at Berkeley. Nancy Oswald, Ph.D., is Consultant, Healthcare Consulting, Berkeley. Lawrence P. Casalino, M.D., Ph.D., is Assistant Professor, Department of Health Studies, University of Chicago. James C. Robinson, Ph.D., is Professor, Division of Health Policy and Management, School of Public Health, University of California at Berkeley. Please address correspondence to Thomas Bodenheimer, M.D., [email protected].
References 1. Chobanian A.V., et al.: The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: The JNC 7 report. JAMA 289:2560–2572, May 21, 2003. E-pub May 14, 2003. Erratum in JAMA 290:197, Jul. 9, 2003. 2. Ni H., Nauman D.J., Hershberger R.E.: Managed care and outcomes of hospitalization among elderly patients with congestive heart failure. Arch Intern Med 158:231–1236, Jun. 8, 1998. 3. Saydah S.H., Fradkin J., Cowie C.C.: Poor control of risk factors for vascular disease among adults with previously diagnosed diabetes. JAMA 291:335–342, Jan. 21, 2004. 4. Griffin S., Kinmouth A.L.: Systems for routine surveillance for people with diabetes mellitus. Cochrane Library, Issue 2, 2004. 5. O’Brien T., et al.: Educational outreach visits: Effects on professional practice and health care outcomes. Cochrane Database Syst Rev 2:CD000409, 2000. 6. O’Brien T., et al.: Local opinion leaders: Effects on professional practice and health care outcomes. Cochrane Database Syst Rev 2:CD000125, 2000.
September 2004
7. Weingarten S.R., et al.: Interventions used in disease management programs for patients with chronic illness—which ones work? Meta-analysis of published reports. BMJ 325:925–932, Oct. 26, 2002. 8. Thomson O’Brien MA, et al: Audit and feedback: Effects on professional practice and health care outcomes. Cochrane Database Syst Rev 2:CD000259, 2000. 9. Ferguson J.A., Weinberger M.: Case management programs in primary care. J Gen Intern Med 13:123–126, Feb. 1998. 10. Bodenheimer T., et al.: Patient self-management of chronic disease in primary care. JAMA 288:2469–2475, Nov. 20, 2002. 11. Casalino L., et al.: External incentives, information technology, and organized processes to improve health care quality for patients with chronic diseases. JAMA 289:434–441, Jan. 22, 2003. 12. Bodenheimer T., Wagner E.H., Grumbach K.: Improving primary care for patients with chronic illness. JAMA 288:1775–1779, Oct. 9, 2002. 13. Bodenheimer T., Wagner E.H., Grumbach K.: Improving primary care for patients with chronic illness: The Chronic Care Model, Part 2. JAMA 288:1909–1914, Oct. 16, 2002.
continued
Volume 30 Number 9
Copyright 2004 Joint Commission on Accreditation of Healthcare Organizations
513
Joint Commission
Journal on Quality and Safety
References, continued 14. Solberg L.I., et al.: Key issues in transforming health care organizations for quality: The case of advanced access. Jt Comm J Qual Safety 30:15–24, Jan. 2004. 15. Goldberg H.I., et al.: Evidence-based management: Using serial firm trials to improve diabetes care quality. Jt Comm J Qual Improv 28:155–166, Apr. 2002. 16. Weber V., Joshi M.S.: Effecting and leading change in health care organizations. Jt Comm J Qual Improv 26:388–399, Jul. 2000. 17. Nelson E.C., et al.: Microsystems in Health Care: Part 1. Learning from high-performing front-line clinical units. Jt Comm J Qual Improv 28:472–493, Sep. 2002. 18. Ginsburg P.B., et al.: The Community Tracking Study analyses of market change: Introduction. Health Serv Res 35(1 Pt 1):7–16, Apr. 2000. 19. Miles M.B., Huberman A.M.: Qualitative Data Management: An Expanded Sourcebook. Thousand Oaks, CA: Sage Publications, 1994.
514
September 2004
20. Rundall T.G., Starkweather D.B., Norrish B.R.: After Restructuring: Empowerment Strategies at Work in America’s Hospitals. San Francisco: Jossey Bass, 1998. 21. Rundall T.G. et al.: As good as it gets? Chronic care management in nine leading US physician organisations. BMJ 325:958–961, Oct. 26, 2002. 22. Huber T.P. et al.: Microsystems in Health Care: Part 8. Developing people and improving work life: What front-line staff told us. Jt Comm J Qual Saf 29:512–522, Oct. 2003. 23. DOQ-IT [Doctor Office Quality-Information Technology] Collaborative: DOQ-IT: Practice Performance Improvement Initiative Accelerating EHR Use to Improve Chronic Care. http://www.cmri-ca.org/physicianoffices/collaboratives/doqit.asp (last accessed Jun. 23, 2004). 24. Epstein A.M., Lee T.H., Hamel M.B.: Paying physicians for highquality care. N Engl J Med 350:406–410, Jan. 22, 2004.
Volume 30 Number 9
Copyright 2004 Joint Commission on Accreditation of Healthcare Organizations
Journal on Quality and Safety
Performance Improvement
What Are the Facilitators and Barriers in Physician Organizations’ Use of Care Management Processes?
Thomas Bodenheimer, M.D. Margaret C. Wang, Ph.D. Thomas G. Rundall, Ph.D. Stephen M. Shortell, Ph.D. Robin R. Gillies, Ph.D. Nancy Oswald, Ph.D. Lawrence Casalino, M.D., Ph.D. James C. Robinson, Ph.D.
ajor deficiencies have been reported in the management of common chronic conditions. For example, some 66% of people with high blood pressure are inadequately treated.1 Half of the patients hospitalized with congestive heart failure (CHF) are readmitted within 90 days,2 and 63% of adults with diabetes have glycohemoglobin (A1C) levels > 7.0; 52% have total serum cholesterol levels of ⱖ 200.3 Literature reviews and meta-analyses have begun to elucidate which practice innovations can improve processes or outcomes of chronic illness care. These innovations include disease registries,4 practice guidelines in conjunction with physician education and reminder systems,5–7 performance feedback,7,8 case management,9 and patient self-management education.10 These chronic illness improvement measures can be called care management processes (CMPs).11 Wagner and associates have proposed a Chronic Care Model that combines these processes into a guide for chronic care improvement.12,13 Some studies have examined the factors that serve as the facilitators and barriers in practice innovation in high-performing organizations,14–17 but these factors have not been explored in a sample of higher- and lowerperforming physician organizations. In 2000–2001 we conducted 158 interviews in 15 physician organizations to determine the major facilitators and barriers affecting the adoption of CMPs for asthma, CHF, diabetes, and depression.
M
September 2004
Article-at-a-Glance Background: Care management processes (CMPs) such as disease registries, reminder systems, performance feedback, case management, and self-management education can improve chronic illness care, yet 50% of physician organizations have instituted few if any CMPs. Methods: Site-visit interviews were conducted with 158 leaders at 15 physician organizations, with 3 organizations (1 large medical group, 1 small medical group, and 1 independent practice association [IPA]) chosen randomly in most cases in each of five communities. Results: Seven of the 15 organizations had implemented CMPs minimally or not at all. CMPs were most common for diabetes and least common for depression; no IPAs had achieved significant CMP implementation for any of the conditions. The two most commonly mentioned facilitators were strong leadership and organizational culture valuing quality. The top five barriers were poor financial situation, reimbursement that does not reward high quality, inadequate information technology, physician resistance, and physicians being too busy. Discussion: Strong leadership and a quality-valuing culture are important facilitators of improved chronic care, but if insurers do not reward chronic care improvement, it is unlikely that CMPs will become permanently institutionalized in physician organizations.
Volume 30 Number 9
Copyright 2004 Joint Commission on Accreditation of Healthcare Organizations
505
Joint Commission
Journal on Quality and Safety
This qualitative study was a part of a larger project, the National Study of Physician Organizations and the Management of Chronic Illness (NSPO), which consisted of a telephone survey of 1,040 medical groups and independent practice associations (IPAs) with 20 or more physicians. The survey indicated that 15% of physician organizations did not use CMPs for these four chronic conditions and that 50% had instituted little if any chronic care improvement. Factors associated with the use of more CMPs were greater clinical information technology (IT) capability and external quality incentives such as public recognition or financial bonuses for better performance.11 Whereas the quantitative NSPO uses a wideangle lens to illuminate CMP implementation in physician organizations, the qualitative part of the NSPO employs a zoom lens to focus on the details of CMP facilitators and barriers.
Methods Sample Between September 2000 and October 2001, seven of the members [all authors but M.C.W.] of the NSPO research group conducted in-depth site visits at 15 physician organizations, chosen in the following manner. The 12 communities studied by the Community Tracking Study of the Center for Studying Health System Change18 were ranked according to the level of 1998–1999 health maintenance organization (HMO) penetration. The top six communities were chosen, but one was dropped because of difficulty gaining access to physician organizations. We picked Community Tracking Study communities because a wealth of information about these health care markets has been collected. For each of the five communities selected, we used a stratified random sampling procedure to choose three physician organizations. In each community, the universe of physician organizations with 20 or more physicians was identified, and information on the type of organization—medical group or independent practice association (IPA)—and the number of physicians was obtained. A medical group was defined as an organization composed of physicians who all belong to one practice, operating in one office or at many locations. An IPA was defined as an organization through which
506
September 2004
physicians contract with managed care plans and in which physicians practice solo or in small groups independently from each other. We randomly selected one physician organization within each of the following strata: ■ Large medical groups (more than 50 physicians) ■ Smaller medical groups (20–50 physicians) ■ IPAs or IPA/medical group hybrids In two communities where physician organizations were reluctant to participate in the study, we selected a similar mix of organizations on the basis of a conveniencesampling strategy.
Interviews At each of the 15 physician organizations studied, 60–90-minute semistructured interviews were conducted with each of the following personnel: chief executive officer, medical director, physician board member, chief financial officer, chief information officer, quality improvement (QI) manager, and physicians active in treating patients with asthma, CHF, diabetes, and depression. Not all of the personnel were relevant or available for interviews at all 15 sites. The interview was intended to assess the extent to which CMPs had been adopted for asthma, CHF, diabetes, and depression and elicited factors that constituted facilitators for or barriers to the implementation of CMPs. The interviews asked about the following CMPs: ■ Chronic disease registries ■ Practice guidelines in conjunction with physician education and reminder systems ■ Performance feedback ■ Case management ■ Patient self-management education One or two members of a three-person team conducted each interview at each site; most physician interviewees were paired with physician interviewers. Each of the 15 site-visit teams prepared a report on the basis of detailed interview notes. Respondents’ answers were tape recorded and recorded by hand on an interview report form. For each physician organization, the sitevisit team prepared a report consolidating the information obtained from all informants. To the extent possible, verbatim responses to questions were incorporated in the site-visit report.
Volume 30 Number 9
Copyright 2004 Joint Commission on Accreditation of Healthcare Organizations
Joint Commission
Journal on Quality and Safety
Table 1. Stages of Implementation of Care Management Processes at 15 Organizations for Four Chronic Illnesses* Organization A1 A2 A3 B1 B2 B3 C1 C2 C3 D1 D2 D3 E1 E2 E3
Organization Type CMP Adoption Asthma Diabetes CHF MG More advanced Aware Implementation Implementation Hybrid More advanced Aware Implementation Implementation MG More advanced Implementation Implementation Institutionalized MG More advanced Implementation Implementation Aware IPA Minimal Pre-aware Pre-aware Pre-aware MG Minimal Pre-aware Aware Pre-aware MG More advanced Pre-aware Implementation Institutionalized MG More advanced Implementation Implementation Aware MG Minimal Pre-aware Aware Aware IPA Minimal Pre-aware Pre-aware Pre-aware IPA Minimal Aware Aware Aware Hybrid More advanced Pre-aware Implementation Implementation MG More advanced Institutionalized Institutionalized Institutionalized Hybrid Minimal Aware Aware Aware IPA Minimal Aware Aware Aware
Depression Pre-aware Pre-aware Aware Aware Pre-aware Pre-aware Aware Aware Pre-aware Pre-aware Pre-aware Aware Implementation Pre-aware Pre-aware
* A,B,C,D, and E represent the five communities visited. CMP, care management process; CHF, congestive heart failure; MG, medical group; IPA, independent practice association; hybrid = mix of MG and IPA. The four stage-of-implementation categories are listed below; minimal CMP adoption entails pre-aware or aware categories for all conditions and more advanced CMP adoption entails implementation or institutionalization for at least one condition.
■ Implementation (ⱖ two CMPs have been developed)
Content Analysis The statements in each site visit report were coded, and qualitative content analyses were performed to determine the extent of CMP adoption, assess IT, and identify barriers and facilitators of CMP implementation. Content analyses were independently completed by three members of the study team [T.B., M.C.W., T.R.], with two members analyzing interview data for each site visit. These coders agreed on the classification of interviewees’ responses into thematic categories in > 80% of classification decisions; inconsistencies were resolved in discussions between the two coders. Given the small number of site visit reports analyzed, manual content analytic techniques19 were chosen. The extent to which the organizations had developed CMPs (Table 1, above) was evaluated using a modified stages-of-implementation analysis20 with the following four categories: ■ Pre-awareness (no thinking or planning about CMPs) ■ Awareness (believe CMPs are needed and may have implemented one CMP)
September 2004
■ Institutionalization (ⱖ two CMPs are permanently
embedded in the organization) It was determined that the introduction of only one CMP was not sufficient to locate a physician organization in the implementation stage; two CMPs were considered necessary for an organization to be categorized as a serious adopter of care management. The sophistication of the organizations’ IT was estimated using the following three categories: ■ Computerized billing and scheduling only (level 1) ■ An electronic medical record (level 2) ■ An electronic medical record plus at least one disease registry or other CMP function such as reminder systems (level 3) Members of the research team shared expectations that medical groups would report more advanced implementation of CMPs than IPAs and that organizations with greater IT would have more CMPs. Our analyses confirmed the former preconception but not the latter.
Volume 30 Number 9
Copyright 2004 Joint Commission on Accreditation of Healthcare Organizations
507
Joint Commission
Journal on Quality and Safety
Results Basic CMP Data Of the 15 physician organizations visited, 8 were medical groups, 4 were IPAs, and 3 were hybrid (mixed medical group and IPA) organizations. Four of the physician organizations had long histories (founded in 1970 or before), whereas 11 were relatively new (founded in 1985 or after). For asthma, 3 organizations had implemented CMPs but these practices were not sufficiently engrained in the organization to ensure their permanence; 1 had institutionalized asthma CMPs, permanently embedding these practices into its clinical structure. Six organizations were in the pre-awareness stage, not having considered developing asthma CMPs, and 5 were in the awareness stage, considering or starting one CMP (Table 1). For CHF, 6 out of 15 organizations had implemented or institutionalized CMPs. Diabetes CMPs were more prevalent, with 8 organizations having achieved implementation or institutionalization stages. The condition for which CMPs were least adopted was depression. Only 1 organization had implemented depression CMPs, whereas 9 were in the pre-awareness stage and 5 were in the awareness stage. The five communities demonstrated a variety of patterns regarding CMP adoption. Each community had 12 CMP-adoption data points—three organizations each reporting on four chronic conditions. The three physician organizations studied in one of the communities had implemented or institutionalized CMPs in 7 out of 12 opportunities; organizations in two communities had implemented CMPs in only 2 out of 12 possibilities. In another of the communities, one of the three physician organizations demonstrated highly advanced CMP development, whereas the other two sites had minimal adoption of CMPs. The community with the greatest level of CMP development is an elite center of medical education populated by several innovative health care institutions. Seven of the 15 organizations, including all four IPAs, demonstrated minimal CMP adoption; they were in the pre-awareness or awareness stage for all four chronic conditions. All of these organizations were relatively young (founded after 1984), whereas all four of the older organizations had relatively advanced CMP development.
508
September 2004
Eleven of the 15 sites had level 1 IT capability: whereas, 4 sites had an electronic medical record (levels 2 and 3). Only 2 sites—older, well-respected physician organizations associated with group model HMOs—had level-3 IT capability, including not only an electronic medical record but also reminder systems, performance feedback systems, or chronic disease registries. These two sites also had the best developed CMPs.
Facilitators and Barriers Table 2 (page 509) and Table 3 (page 510) show the number of organizations whose interviewees mentioned specific facilitators and barriers affecting the adoption of CMPs. The two most commonly mentioned facilitators were strong leadership and an organizational culture that valued quality. Other facilitators mentioned were IT, having physician champions, and capitation payment. The top five barriers were lack of resources, reimbursement that does not reward high quality, inadequate IT, physician resistance, and physicians being too busy. The facilitators and barriers included a mixture of intraorganizational (leadership, culture, and IT) and broader health system factors (mode of reimbursement). Some facilitators and barriers can be viewed as two sides of the same coin. For physician organizations succeeding at CMP implementation, the facilitators were good leadership, a quality-oriented culture, and advanced IT, whereas for organizations without CMPs, poor leadership, the lack of a quality culture, and poor IT were barriers. In the quantitative NSPO, IT capability was associated with greater use of CMPs.11 In the qualitative study reported here, only three of eight organizations more advanced in CMPs had advanced IT. One organization with well-developed IT had adopted virtually no CMPs. Although sophisticated IT makes chronic care improvements easier to achieve, it appears that improvements can be made without advanced IT. Moreover, having IT does not guarantee greater use of CMPs. Acknowledgement of the role of leadership and culture varied markedly between the eight sites that implemented CMPs and the seven sites with minimal CMP adoption. All eight sites with more advanced CMPs cited leadership and/or physician champions as facilitators; of
Volume 30 Number 9
Copyright 2004 Joint Commission on Accreditation of Healthcare Organizations
Joint Commission
Journal on Quality and Safety
Table 2. Facilitators Mentioned by Interviewees* Organization A1 A2 A3 B1 C1 C2 D3 E1 B2 B3 C3 D1 D2 E2 E3 Total
CMP Adoption More Advanced More Advanced More Advanced More Advanced More Advanced More Advanced More Advanced More Advanced Minimal Minimal Minimal Minimal Minimal Minimal Minimal
Leadership Quality Culture Advanced IT Champions X X X X X X X X X X X X X X X X X X X X X X X X X X X
Have Financial Risk X X X
X X
9
8
6
6
X 4
* A,B,C,D, and E represent the five communities visited. CMP, care management process; IT, information technology.
the seven less-advanced sites, only two cited those facilitators. Of the eight more advanced sites, six mentioned quality culture as a facilitator; of the seven less advanced sites, only two mentioned culture as facilitating. The issue of busy physicians yielded surprising findings. Seven of the eight more advanced sites noted that overworked physicians was a barrier; whereas, only one less advanced site mentioned overworked physicians as a barrier. This suggests that good leadership and culture that values quality might trump physician overwork in determining whether organizations implement CMPs.
Case Studies: Interaction of Facilitators and Barriers Case studies for 4 of the 15 organizations elucidate the interaction of facilitators and barriers (organization names are fictitious).
Southern IPA: An IPA with Little Interest in CMP Development Southern IPA includes 400 physicians who practice in their private offices. The IPA’s main function is to contract with health plans on behalf of its physicians. The
September 2004
IPA itself has a tiny budget and two part-time employees. Southern IPA does not gather clinical data about its physicians. Chronic care improvement is not part of Southern IPA’s strategic plan. Many Southern physicians do not trust clinical practice guidelines nor performance feedback reports sent by health plans. One physician noted that practice algorithms assume that “all the patients who have asthma or high blood pressure are the same. As if we were all identical twins.” When asked about a diabetes registry, a physician replied, “I receive capitation payments. Why should I send patients letters and contact them to make appointments?” When asked about a CHF program, a physician answered that it does not matter if CHF patients go into the hospital because he would not get paid under capitation for hospital visits. One health plan pays a small bonus to physicians with good Health Plan Employer Data and Information Set (HEDIS) scores, but the amount is insignificant and physicians pay no attention. Southern IPA’s culture is to shield physicians from intrusion by health plans. The physician leadership focuses on health plan contracts and has little interest in care management. One leader admitted that although he
Volume 30 Number 9
Copyright 2004 Joint Commission on Accreditation of Healthcare Organizations
509
Joint Commission
Journal on Quality and Safety
Table 3. Barriers Mentioned by Interviewees* Organization A1 A2 A3 B1 C1 C2 D3 E1 B2 B3 C3 D1 D2 E2 E3 Total
CMP Adoption More Advanced More Advanced More Advanced More Advanced More Advanced More Advanced More Advanced More Advanced Minimal Minimal Minimal Minimal Minimal Minimal Minimal
In Poor Financial Plans Don’t Pay Shape for Quality X X X X X X X X X
Physician Resistance
Poor IT X X
X X X
X X X X
X X X X X X X 11
X X
X X
X
X X
X 9
9
Physicians Are Too Busy X X X X X X X
X X X X X X 9
8
* A,B,C,D, and E represent the five communities visited. CMP, care management process; IT, information technology.
would welcome organized care for chronically ill patients, there are no incentives, no clinical meetings, and no interest. When asked about CMPs, a front-line physician said, “All these things make more work for me, so I don’t do them.” In 2001, when our interviews were conducted, Southern IPA’s community was beset with conflict, according to the Community Tracking Study. Hospitals were consolidating and obtaining better contracts with health plans, which in turn became less profitable. As a result, health plans attempted to reduce physician payments, which escalated physician anger. This was an environment unfriendly to QI. The case of Southern IPA demonstrates how barriers build on one another. The physician leadership was not interested in QI, thereby reinforcing a culture of physician independence. The NSPO demonstrated that CMPs are associated with clinical information systems, better health plan contracts for quality, and external recognition of quality care. Even if a quality champion emerged at Southern IPA, it would have been difficult to institute CMPs because no clinical information systems existed,
510
September 2004
physician organizations did not receive recognition for quality, and health plans did not significantly reward quality. Because the IPA’s leadership wanted health plan dollars to flow to physicians, the IPA itself had almost no budget and was unable to build a clinical information system or to institute CMPs.
Eastern Health System: A Medical Group with Advanced CMP Implementation Eastern Health System is a large integrated delivery system whose reputation enables it to attract excellent physicians. According to the Community Tracking Study, the health care market in Eastern’s community was in turmoil at the time of our interviews, but Eastern’s stable patient enrollment and employed-physician workforce shielded it from neighboring bankruptcies and dislocations. For a decade, Eastern’s leaders had been at the forefront of chronic care improvement and had built an IT system that facilitates CMP development. Care management enjoys a distinct budget and expert physicians are paid to work on CMPs. The financial incentive inherent in the capitated system encourages programs for
Volume 30 Number 9
Copyright 2004 Joint Commission on Accreditation of Healthcare Organizations
Joint Commission
Journal on Quality and Safety
such conditions as asthma and CHF, which can reduce hospital admissions and cut Eastern’s expenditures. Organizational leaders take pride that the culture of “doing the right thing” is more important than the prospect of financial savings in developing CMPs. Eastern’s diabetes program illustrates a well-designed CMP. A registry of several thousand diabetics has been compiled, giving physicians computerized access to their diabetic panel. All diabetic patients receive a letter on their birthday indicating which screening exams they are missing; they are also called by phone. Diabetes classes are available. For every diabetic visit, physicians receive a reminder prompt including laboratory and medication data and missing screening tests. Patients whose conditions are poorly controlled are referred to case managers. Practice guidelines are available to all physicians on the intranet, and diabetes care is discussed at some monthly meetings. Physicians receive quarterly feedback reports on their diabetic panels. Because it is both a health plan and a medical group, Eastern Health System receives public recognition for quality through the National Committee for Quality Assurance, although it does not receive better contracts for quality. Whereas CMPs are highly developed, the extent to which they are used by front-line physicians is unclear. Physicians are extremely busy and may view reminder prompts and performance feedback reports as more work for no reward. Whereas Eastern has the leadership, culture, financial resources and IT to implement CMPs, the barrier of a lack of physician time limits the CMPs’ influence on each physician-patient encounter.
Western Medical Group: A Medical Group Attempting to Initiate CMPs Western Medical Group is a 60-physician primary care group whose leaders want to develop CMPs but have not yet succeeded in doing so. Western recently rolled out an electronic medical record but has not yet used it to produce clinical data. A medical director explained, “Our investment in computers has been enormous. Now we must reap the benefit by using it to improve care.” According to the group’s leaders, Western Medical Group’s organizational culture emphasizes physician independence, decentralization of decision making, and
September 2004
low group cohesiveness. Implementing centralized CMPs would be expected to meet considerable resistance. One senior manager said, “Some doctors are concerned that if there is a practice guideline and they do not follow it exactly they could be sued.” The leadership has informed physicians that Western will be monitoring several diabetes and CHF process measures and will pay physicians bonuses if they score satisfactorily on those measures. Leaders hope that business will trump culture. The Community Tracking Study places Western Medical Group in a market formerly with medium-high HMO and capitation penetration, more recently retrenching toward weaker managed care and fee-forservice reimbursement. Because some large physician organizations have failed financially, Western’s culture has more strongly embraced physician productivity to improve its finances by increasing the volume of visits. Western has been losing money and is unable to hire a QI staff person. In summary, Western Medical Group has leaders interested in QI but would need a culture shift to sell CMPs to many of its physicians. A robust IT system has not been sufficient to institutionalize registries, reminders, and performance feedback. Western’s shaky financial situation has made it impossible to hire a quality manager to activate those IT functions and sell them to physicians. Lack of health plan bonuses for improved quality and no public reporting of quality indicators adds to an environment unfavorable for CMP development. Leadership is a necessary but not a sufficient condition for QI; cultural and financial barriers can stymie even the most committed leaders.
Northern Clinic: A Medical Group with Substantial CMP Development That Is Losing Some of Its Gains Northern Clinic is a highly respected multispecialty clinic in a market characterized by the Community Tracking Study as highly competitive among health plans and physician organizations, with capitation giving way to fee for service. In contrast to Western Medical Group, Northern Clinic has implemented a number of CMPs but because of financial pressures some of these innovations have dropped from clinicians’ radar screens. Though a quality culture has a long
Volume 30 Number 9
Copyright 2004 Joint Commission on Accreditation of Healthcare Organizations
511
Joint Commission
Journal on Quality and Safety
history at Northern, some leaders are cynical about how deep that culture penetrates. Whereas one leader claimed, “We will not sacrifice quality for cost,” another complained, “Evidence-based medicine is not a required course here; it is an elective.” The external environment is having a major impact on Northern Clinic’s commitment to chronic care quality. One physician executive admitted, “If care management practices don’t result in cost reduction, they should probably be eliminated.” Another added, “The movement away from capitation to fee for service makes the need for implementing CMPs less urgent” and explained, “The organization has experienced several difficult years financially, making it hard to allocate FTEs [full-time equivalents] for needed CMP work.” At Northern Clinic, a clash has developed between the culture of quality and the reality of financial solvency.
Discussion The qualitative component of the NSPO adds to the findings of the quantitative survey of 1,040 physician organizations.11 The large survey found that external incentives and IT are associated with the adoption of CMPs but did not examine issues of organizational leadership and culture. Leaders of the 15 physician organizations receiving in-depth site visits frequently confirmed that the presence or absence of quality-enhancing incentives from payers and adequate IT were important facilitators or barriers to improved chronic care. The in-depth interviews provided the additional insight that strong leadership and a quality-oriented physician culture are powerful facilitators for chronic care improvement. Similar conclusions were reached by several other studies.14,16,17,21 In the 15 sites, the majority of facilitators to CMP adoption appear to be factors internal to the organization rather than conditions created by the financial environment impinging on the organization. The five most common facilitators were mentioned 33 times by at least one interviewee at a site. Of these 33 mentions, 29 involved factors internal to the organization—leadership and champions, a culture valuing quality, and adequate IT; only 4 noted an external factor—the organization’s assumption of financial risk. Leadership and a quality culture may be the primary facilitator; these traits corre-
512
September 2004
lated strongly with a high level of CMPs in the 15 sites. In contrast, external factors appeared to have greater importance as barriers to chronic care improvement; the external environment was cited 20 times out of 47 mentions of barriers. In organizations with strong leadership and a qualityfocused culture, the most frequently mentioned barriers—inadequate finances, payers not rewarding quality, inadequate IT, and resistance or overwork of physicians—did not prevent the adoption of CMPs. Strong leadership and quality culture can sometimes overcome barriers to CMP development. Care management implementation (CMI) can be simplified to a formula: CMI = D × A (desire × ability to get it done). Both D and A are needed for success. The desire is expressed by the facilitators of strong leadership and a quality-valuing culture. The ability to get it done can be thwarted by the barriers discussed above. One might hypothesize that the facilitators of strong leadership and a quality-oriented culture (the desire) trump the barriers in initiating CMPs, while the major barriers (the inability to get it done) can overcome the facilitators in sustaining CMPs over time. Longitudinal data collection would be needed to explore this hypothesis. One area requiring further research is that of physician overwork. Sites mentioning physician overwork as an important barrier tended to be strong CMP adopters. Do some CMPs (for example, case management) reduce physician work? Do CMPs create more physician work, while strong leadership and a quality-oriented culture overcome physician overwork in the mix of factors determining CMP adoption? Can CMPs be sustained in an organization of overworked physicians? This article has several limitations. First, given the small sample size, the study should be considered exploratory and suggestive of the challenges involved in implementing chronic care improvement. Second, the study did not measure clinical outcomes but examined clinical processes, some of which have been correlated with improved outcomes.4–10 Third, the pattern of findings cannot address the issue of causality. For example, does sophisticated IT create CMP adoption, or does strong leadership create the development of both sophisticated IT and CMPs?
Volume 30 Number 9
Copyright 2004 Joint Commission on Accreditation of Healthcare Organizations
Joint Commission
Journal on Quality and Safety
Conclusion The information gathered in this study suggests several health system policy and practice changes that might encourage chronic care improvement. In terms of practice changes, health science schools and residency programs should add physician leadership training and education on the Chronic Care Model to existing curricula. Redesigning primary care and delegating routine chronic care management functions to nonphysician personnel may help address the barrier of physician overwork.22 Physician organizations need financial and technical assistance in developing IT. In a demonstration program recently initiated through Medicare, such assistance will be offered to physician organizations that are initiating computerized quality performance measurement.23 In terms of policy, reforms needed in the external environment include the devising of reimbursement formulas by Medicare, Medicaid, and private insurers which pay for care management processes and reward superior performance. Early pay-for-performance programs are sprouting up across the United States.24 The consequences of the trend from capitation payment back to fee for service need to be examined because fee-for-service payment promotes a culture valuing quantity rather than quality. It is essential to have inspiring leaders to initiate chronic care improve-
ments, but sustaining these improvements requires insurers to pay for them. J Grant support was provided by the Robert Wood Johnson Foundation (grant no. 038690). The authors also acknowledge the fellowship support of Margaret Wang provided by the Health Research and Educational Trust (HRET).
Thomas Bodenheimer, M.D., is Adjunct Professor, Department of Family and Community Medicine, University of California at San Francisco, San Francisco. Margaret C. Wang, Ph.D., is a Researcher, School of Public Health, University of California at Berkeley, Berkeley, California, Thomas G. Rundall, Ph.D., is Professor, Division of Health Policy and Management, School of Public Health, University of California at Berkeley. Stephen M. Shortell, Ph.D., is Dean, School of Public Health, University of California at Berkeley. Robin R. Gillies, Ph.D., is Project Director, Division of Health Policy and Management, School of Public Health, University of California at Berkeley. Nancy Oswald, Ph.D., is Consultant, Healthcare Consulting, Berkeley. Lawrence P. Casalino, M.D., Ph.D., is Assistant Professor, Department of Health Studies, University of Chicago. James C. Robinson, Ph.D., is Professor, Division of Health Policy and Management, School of Public Health, University of California at Berkeley. Please address correspondence to Thomas Bodenheimer, M.D., [email protected].
References 1. Chobanian A.V., et al.: The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: The JNC 7 report. JAMA 289:2560–2572, May 21, 2003. E-pub May 14, 2003. Erratum in JAMA 290:197, Jul. 9, 2003. 2. Ni H., Nauman D.J., Hershberger R.E.: Managed care and outcomes of hospitalization among elderly patients with congestive heart failure. Arch Intern Med 158:231–1236, Jun. 8, 1998. 3. Saydah S.H., Fradkin J., Cowie C.C.: Poor control of risk factors for vascular disease among adults with previously diagnosed diabetes. JAMA 291:335–342, Jan. 21, 2004. 4. Griffin S., Kinmouth A.L.: Systems for routine surveillance for people with diabetes mellitus. Cochrane Library, Issue 2, 2004. 5. O’Brien T., et al.: Educational outreach visits: Effects on professional practice and health care outcomes. Cochrane Database Syst Rev 2:CD000409, 2000. 6. O’Brien T., et al.: Local opinion leaders: Effects on professional practice and health care outcomes. Cochrane Database Syst Rev 2:CD000125, 2000.
September 2004
7. Weingarten S.R., et al.: Interventions used in disease management programs for patients with chronic illness—which ones work? Meta-analysis of published reports. BMJ 325:925–932, Oct. 26, 2002. 8. Thomson O’Brien MA, et al: Audit and feedback: Effects on professional practice and health care outcomes. Cochrane Database Syst Rev 2:CD000259, 2000. 9. Ferguson J.A., Weinberger M.: Case management programs in primary care. J Gen Intern Med 13:123–126, Feb. 1998. 10. Bodenheimer T., et al.: Patient self-management of chronic disease in primary care. JAMA 288:2469–2475, Nov. 20, 2002. 11. Casalino L., et al.: External incentives, information technology, and organized processes to improve health care quality for patients with chronic diseases. JAMA 289:434–441, Jan. 22, 2003. 12. Bodenheimer T., Wagner E.H., Grumbach K.: Improving primary care for patients with chronic illness. JAMA 288:1775–1779, Oct. 9, 2002. 13. Bodenheimer T., Wagner E.H., Grumbach K.: Improving primary care for patients with chronic illness: The Chronic Care Model, Part 2. JAMA 288:1909–1914, Oct. 16, 2002.
continued
Volume 30 Number 9
Copyright 2004 Joint Commission on Accreditation of Healthcare Organizations
513
Joint Commission
Journal on Quality and Safety
References, continued 14. Solberg L.I., et al.: Key issues in transforming health care organizations for quality: The case of advanced access. Jt Comm J Qual Safety 30:15–24, Jan. 2004. 15. Goldberg H.I., et al.: Evidence-based management: Using serial firm trials to improve diabetes care quality. Jt Comm J Qual Improv 28:155–166, Apr. 2002. 16. Weber V., Joshi M.S.: Effecting and leading change in health care organizations. Jt Comm J Qual Improv 26:388–399, Jul. 2000. 17. Nelson E.C., et al.: Microsystems in Health Care: Part 1. Learning from high-performing front-line clinical units. Jt Comm J Qual Improv 28:472–493, Sep. 2002. 18. Ginsburg P.B., et al.: The Community Tracking Study analyses of market change: Introduction. Health Serv Res 35(1 Pt 1):7–16, Apr. 2000. 19. Miles M.B., Huberman A.M.: Qualitative Data Management: An Expanded Sourcebook. Thousand Oaks, CA: Sage Publications, 1994.
514
September 2004
20. Rundall T.G., Starkweather D.B., Norrish B.R.: After Restructuring: Empowerment Strategies at Work in America’s Hospitals. San Francisco: Jossey Bass, 1998. 21. Rundall T.G. et al.: As good as it gets? Chronic care management in nine leading US physician organisations. BMJ 325:958–961, Oct. 26, 2002. 22. Huber T.P. et al.: Microsystems in Health Care: Part 8. Developing people and improving work life: What front-line staff told us. Jt Comm J Qual Saf 29:512–522, Oct. 2003. 23. DOQ-IT [Doctor Office Quality-Information Technology] Collaborative: DOQ-IT: Practice Performance Improvement Initiative Accelerating EHR Use to Improve Chronic Care. http://www.cmri-ca.org/physicianoffices/collaboratives/doqit.asp (last accessed Jun. 23, 2004). 24. Epstein A.M., Lee T.H., Hamel M.B.: Paying physicians for highquality care. N Engl J Med 350:406–410, Jan. 22, 2004.
Volume 30 Number 9
Copyright 2004 Joint Commission on Accreditation of Healthcare Organizations