Education And Debate

As good as it gets? Chronic care management in nine leading US physician organisations

BMJ 2002; 325 doi: http://dx.doi.org/10.1136/bmj.325.7370.958 (Published 26 October 2002) Cite this as: BMJ 2002;325:958
  1. Thomas G Rundall, professor (trundall{at}uclink.berkeley)a,
  2. Stephen M Shortell, professora,
  3. Margaret C Wang, research assistanta,
  4. Lawrence Casalino, assistant professorb,
  5. Thomas Bodenheimer, clinical professor of family and community medicinec,
  6. Robin R Gillies, research directora,
  7. Julie A Schmittdiel, research assistanta,
  8. Nancy Oswald, presidentd,
  9. James C Robinson, professora
  1. aUniversity of California at Berkeley, School of Public Health, Division of Health Policy and Management, 140 Warren Hall, Berkeley, California, USA 94720-7360
  2. bThe University of Chicago, Department of Health Studies, 5841 South Maryland Avenue, MC 2007, Chicago, Illinois, USA 60637
  3. cUniversity of California at San Francisco, Building 80-83, San Francisco General Hospital, 101 Potrero Avenue, San Francisco, California, USA 94110
  4. dHealthcare Consulting, 61 Rock Lane, Berkeley, California, USA 94708
  1. Correspondence to: T G Rundall

    Innovations in care management processes have improved the care of patients with chronic illnesses, but many patients still do not receive these benefits. The authors have studied the barriers and facilitators to implementing these improvements in leading US physician practices

    About 125 million of the 276 million people living in the United States have some type of chronic illness (table 1).1 Four chronic conditions affect nearly half of Americans with a chronic disease: asthma, depression, and diabetes each affect about 15 million,24 while five million have congestive heart failure.5 In 1999 these four chronic diseases were directly responsible for 140 000 deaths in the United States6 and generated at least $173bn (£108bn, €170bn) in medical and other costs. 5 79

    Over the past decade the effectiveness of care for patients with these and other major chronic illnesses has been improved by innovations in care management processes such as the use of guidelines, disease management techniques, case management, and patient education to improve self management of chronic disease.10 However, many patients are not benefiting from these advances. Recent studies indicate that fewer than half of US patients with asthma, depression, and diabetes receive appropriate treatment.1113 Organisational characteristics of physician practices associated with effective chronic disease care include the use of patient care teams, supportive information systems, and a high volume of patients.14 Hence, we expect that in the United States moderate and large sized, well organised, multispecialty practices are likely to offer chronic disease care that is as good as it gets and provide other physician organisations with benchmarks against which performance can be measured.

    Summary points

    • Many patients with chronic diseases are not benefiting from effective care management processes

    • Among leading physician practices in the United States, examples of comprehensive, evidence based chronic care management processes can be found

    • Among these practices, however, the use of such processes for patients with asthma, congestive heart failure, depression, and diabetes varied greatly

    • The use of computer based information systems to support the care of patients with chronic diseases could be considerably expanded

    • The future agenda for improving physician practices should include redesigning work processes to address physicians' concerns about workload, promotion of a culture that supports quality improvement, diffusion of clinical information systems, and financial incentives to reward practices that improve the care and outcomes of patients with chronic disease

    We assessed the extent to which chronic care management processes and computer based clinical information systems were used to care for patients with asthma, congestive heart failure, depression, and diabetes in nine large, multispecialty practices with national reputations for delivering high quality care. We also identified the barriers and facilitators in these organisations that affected their ability to implement care management processes and clinical information systems.

    The nine practices studied were the medical groups of the Cleveland Clinic, Group Health Cooperative of Puget Sound, Henry Ford Health System, Intermountain Health Care, Lovelace Clinic, Marshfield Clinic, Mayo Clinic (Rochester, Minnesota Division), Park Nicollet Clinic, and the Permanente Medical Group. The care management processes we examined were practice guidelines (descriptions or statements that guide recommended treatment based on literature and scientific review), population disease management (a care programme that identifies the population of patients with chronic diseases served by the medical group and provides these patients with a symptom management plan, conditioning and drug regimen, education on the disease, and case management by telephone), case management (a programme to intensively manage individual patients with uncontrolled or high cost conditions), and health promotion or disease prevention (an activity occurring outside the clinical encounter that promotes health such as a newsletter containing advice, a series of classes addressing a specific health risk, or screening offered in the community). We also collected information on each medical group's use of seven selected functions of a clinical information system that support chronic disease care management—electronic medical record, electronic recording of health history, recording of tests and procedures, recording of diagnosis and treatment, computerised entry of drug prescriptions, automated reminders, and electronic exchange of information with patients.

    Table 1.

    Burden of chronic illness in the United States*

    View this table:

    These care management processes and clinical information system functions incorporate many of the practices recommended in Wagner's chronic care model, a widely recognised guide to improving chronic care.1518 The chronic care model suggests that effective care requires an appropriately organised delivery system linked with complementary community resources available outside the organisation.

    Methods

    We selected the sample of leading medical groups (table 2) on the recommendations of the study team members and of the members of the study's national advisory committee. We used qualitative methods to collect and analyse the data. 1920 Members of the research team conducted structured interviews with at least three senior leaders in each organisation—the chief executive officer, the chief medical officer, and the chief information officer. A total of 41 interviews were conducted between May and September 2001. Respondents were assured that none of their comments would be attributed to them personally or to their organisation.

    Table 2

    Physician organisations in the study

    View this table:

    Implementation of care management processes and clinical information systems

    Although we found several examples of comprehensive use of care management processes and clinical information systems (see box), some of the medical groups used few, if any, of these care processes. In some cases care management processes were discontinued because of financial and staffing problems. For example, one medical director reported, “In the major downsizing of 1998 the physician running the quality initiatives left and was not replaced. Twelve nurses were laid off. It was bloody.” Even when care management processes are maintained, sustaining a high level of effort can be difficult. As one chief executive officer explained, “It's hard work, and one has to keep at it year after year. If you let up, you can lose the gains.”

    Example of comprehensive care management

    • In one medical group the doctors have implemented clinical guidelines,disease management, case management, and health promotion and disease prevention for asthmatic patients. There is a registry of asthma patients. Doctors assess whether patients are level one (condition controlled),level two (condition uncontrolled), or level three (condition is complex with multiple diagnoses). The primary care team cares for level one patients. Level two patients are referred to an asthma case manager, who provides intensive care for each patient for about six months with the expectation that the patient's condition will be brought under control. Level three patients are assigned a case manager who is usually a registered nurse, often assisted by a social worker, working in the primary care team. Level one and level two patients are asked to attend educational classes. The organisation also has a website linked to a self care handbook that can be accessed by patients. The registry is used to track patients' progress. In particular the ratio of use of inhaled corticosteroids to use of inhaled corticosteroids plus bronchodilators is periodically assessed. The organisation's information system generates prompts that are placed on the front of charts for asthmatic patients. Physicians receive feedback reports every six months.

    Use of care management processes— Nearly all study groups used clinical practice guidelines for all four chronic conditions (asthma, congestive heart failure, depression, and diabetes). However, the use of the other care management processes varied greatly across conditions (table 3). Additional analyses not shown here revealed that, for each disease, fewer than half of the nine groups used all four care management processes: three used all four processes for asthma, as did four for congestive heart failure, none for depression, and four for diabetes.

    Table 3.

    Number of nine leading physician organisations that have implemented care management processes for selected chronic diseases

    View this table:

    Care management functions performed by clinical information systems— The use of clinical information system functions varied greatly among the medical groups (table 4). One group did not use any of the seven functions, and no group used all seven functions. The eight groups with some clinical information system capability used between four and six of the functions. Six of the nine groups used an electronic medical record; only two had developed electronic information exchange with patients. Two of the clinical information system functions (computerised entry of drug prescriptions and automated reminders) have been shown to improve quality of care in well designed studies. 21 22 Only four of the nine practices used computerised entry of drug prescriptions. Automated reminders were used by six of the practices.

    Table 4.

    Information technology functions available to support management of care for patients with chronic diseases implemented by nine physician organisation

    View this table:

    Barriers and facilitators to use of care management processes

    Why have some of these medical groups implemented a full complement of chronic care management processes whereas others have not? We uncovered important barriers and facilitators to the use of care management processes (table 5). Frequently mentioned barriers were lack of financial and staff resources, inadequate clinical information systems, doctors' heavy workload, compensation not being related to quality of care, and doctors' resistance to change. Doctors' resistance to change and heavy workload are related since overworked doctors fear that change may make things even more difficult. One medical director said, “We have a major problem with overwork in primary care, and it's getting worse. It is impossible to launch any programme that gives physicians more work.”

    These barriers can be overcome by facilitators, in particular a group culture oriented to quality and supportive managerial and medical leadership. Group culture is often identified as an important factor determining the service orientation of health organisations.23 Our respondents often noted that at the heart of their group's efforts to provide high quality chronic care is the belief that using care management processes “is the right thing to do.” One chief executive officer explained, “We have a culture, a tradition, of partnership between clinician and patient. The patients have high expectations of our organisation, and we feel an impatience to get things improved.”

    Table 5

    Five most frequently mentioned barriers and facilitators for implementing chronic care management processes (opinions of a total of 41 chief executive officers, chief medical officers, and chief information officers from nine physician organisations)

    View this table:

    Culture can also be a barrier and can vary with changes in leadership. One medical director reflected, “We did not harness our intellectual capital to support innovations. The focus was on expense reduction. Now [with new leaders] we have a renaissance of concern about quality; we are having a cultural turnaround.”

    Facilitators can come from outside as well as inside the organisation. A frequently mentioned facilitator was “support from external organisations such as health plans.” A quality oriented culture can be greatly enhanced if health plans reward physician organisations by paying more for high quality.

    Building chronic care management capabilities

    Some care management processes are easier to achieve than others. The implementation of clinical practice guidelines and health promotion or disease prevention activities for chronic disease is now common among leading physician organisations, and probably feasible for all large group practices. Implementation of population disease management and case management seems to be more difficult. The most common pattern was for the medical groups to implement two or three of the four care management processes, often leaving out population disease management or case management, or both. Surprisingly, there was no relation between the level of sophistication of groups' clinical information systems and their degree of implementation of care management processes. However, the medical groups that were profitable and had a clinical information system, supportive clinical and administrative leadership, a group culture that promoted quality improvement, and incentives from health plans and other external organisations to improve quality were more likely to implement chronic disease care management processes.

    If our findings about the barriers and facilitators for use of care management processes are supported by large sample studies of medical practices, the future agenda for restructuring practices should include a redesign of work processes to address doctors' concerns about workload, promotion of a culture that supports quality improvement, diffusion of clinical information systems, and a change in the design of providers' financial incentives.

    Acknowledgments

    We thank the many professional associations that formally endorsed the study, including the American Academy of Family Physicians, American Academy of Pediatrics, American College of Surgeons, American Medical Association, American Medical Group Association, American Society of Internal Medicine, California Association of Physician Organisations, California Medical Association, Illinois State Medical Society, IPA Association of America, Massachusetts Medical Society, Medical Group Management Association, Medical Society of the State of New York, Michigan State Medical Society, Minnesota Medical Association, National IPA Coalition, and Texas Medical Association.

    Contributors: TGR, SMS, RRG, JCR, LC, TB, and NO contributed to the conceptual development of the study, study design, data collection and analysis, and writing the manuscript. MCW contributed to the study design, data collection and analysis, and writing the manuscript. JAS contributed to the data collection and analysis. DB, EW, and PG contributed comments and consultation. TGR is guarantor for the study.

    Footnotes

    • Funding This work was supported by the Robert Wood Johnson Foundation, grant No 038690.

    • Conflict of interest None declared..

    References

    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.
    15. 15.
    16. 16.
    17. 17.
    18. 18.
    19. 19.
    20. 20.
    21. 21.
    22. 22.
    23. 23.