Patient Self-Management Support Programs: An Evaluation
Design Options for a Self-Management Support Program
In both the literature and interviews, we found wide variation in the design of current programs. In particular, programs vary with respect to:
- The location of the program within the health care system; i.e., what entity offers the service and manages it.
- The extent to which the program is provided through personal interaction between coaches and patients.
- Content of the support.
- Patient population served.
- Information support.
- Protocols for how staff is to provide the support.
- Staff training.
- In what manner and how often coaches communicate with patients.
- Nature of the communication between primary care physicians and self-management support staff.
These factors are discussed in more detail below. The programs also vary in terms of the performance measures used. Program evaluation is discussed in the main section that follows.
Program Location and Extent of Personal Interaction Between Coaches and Patients
Programs vary with respect to:
- Where they are located within the health care system; i.e., who has responsibility and authority for their day-to-day operations.
- Whether or not self-management support is provided through some form of personal interaction.
Some programs are located within primary care practices, some are run by other health care organizations such as plans or hospitals, and others are run by commercial vendors. For purposes of this discussion, we will distinguish between those programs that are located within the primary care setting and are under the responsibility of the local provider (e.g., physician group or clinic) and those that are located outside of the local setting (e.g., through a plan or commercial vendor or centralized within an independent delivery system).
In addition, some programs include telephone contact or face-to-face contact but are distinguished from programs that rely entirely on technology or written materials without any person-to-person interface. Personal interaction may be further characterized by whether it involves face-to-face meetings or relies on contact by telephone or computer.
These first two factors combine to create four models of self-management support delivery most frequently seen in the United States today. The four models are summarized in Table 1 and described further below.
Primary Care Model
In the primary care model, self-management support is usually provided directly by local providers' offices and usually includes face-to-face contact in the primary care office setting. The patient may be referred to additional self-management support resources. Action plans are often used as tools for collaborative goal setting, patient activation, and communication, and clinicians have ready access to patients' self-management assessments and goals. Group sessions (group visits, classes, support groups) can be offered in these local settings. The coaches may work on teams with clinicians.
Current database and information technology constraints in many primary care practices may limit the availability of patient information to support the self-management support coaches' decisions. Program objectives are more likely to be framed in terms of quality rather than return on investment. The development of primary care, office-based mechanisms for self-management support frequently occurs as part of efforts to implement the Chronic Care Model in primary care practices. These programs are initiated by the local, primary care practice (or by the delivery system of which the local practice is a part). Plans, employers, and other payers can use contracting mechanisms or possibly incentives, such as pay-for-performance, to encourage this local self-management support programming, but they usually do not purchase it directly. An example of a real-life primary care model is described in the following box.
Example of a Primary Care Model
This private, non-profit corporation of nine community medical centers provides preventive and primary care services in a rural, underserved area. The centers began developing self-management support services in 1999 for patients with diabetes. They now have hired 15 "care managers" who provide self-management support for multiple conditions (such as asthma, cardiovascular disease, depression, and diabetes) and for prevention and lifestyle issues.
The care managers primarily are licensed practical nurses. A care manager meets with a patient in the primary care setting, and they work together to set a self-management goal that is "actionable and time framed," according to the program. Group visits are used as well as office visits and phone calls.
The goal of the program is to have every patient seen by a care manager, and for 97 percent of all patients to set an annual self-management goal. The care managers follow up with the patients (in-person or by telephone) to see how they are doing at meeting their goals. They also will contact the person prior to a scheduled visit to ask what problems he or she would like to discuss during the visit. Problem solving is seen as the foundation of self-management support. The care manager interacts with the patient's primary care clinicians through a team structure that includes regular team "huddles." Program management evaluates the effectiveness of the self-management support program with measures of patients' confidence in their ability to manage their conditions.
These measures use patient-reported data collected with an anonymous written survey.
External On-the-ground and External Call Center Models
In contrast to primary care model programs, programs in the external on-the-ground and external call-center models are managed from outside the local primary care setting and are likely to be referred to as "disease management" programs. Such programs differ from each other in the nature and degree of personal interaction between the self-management support staff and the patient.
In the external on-the-ground model, self-management support is provided by an organizational entity external to the local provider and usually outside of the primary care setting. The support generally is more intensive than in the other models and involves face-to-face and telephone interaction. The coaches have their "boots on the ground" in that they may go to where the patients are, providing support in homes, primary care offices, and community settings. More medical care management and coordination may be provided. This model usually uses large electronic databases, allowing identification of full populations with given diagnoses and often patient-specific data on visit and prescription utilization to guide self-management support interactions. Although outreach and some self-management support, such as written materials, an educational Web site, or possibly a hotline or call-in number, may be offered to everyone identified with the diagnosis, only a defined subset of the patients receive the services of the health coach. Program objectives are likely to be framed in terms of return on investment and quality improvement. External call-center programs may be developed internally in plans and independent delivery systems or purchased from vendors by plans, independent delivery systems, employers, or government payers. Purchasers of health care also can use contracting mechanisms to obtain such services. The box below presents an example of this model.
Example of an External on-the-Ground Model
An example of an external on-the-ground program is a diabetes disease management program developed and run by an independent delivery system. During the period under study, the program employed 51 RNs as primary care nurse educators and case managers. Each nurse was responsible for 1-15 primary care sites. The nurses provided self-management support at the primary care clinic in one-on-one encounters or in group sessions. A nurse saw each patient from one to four times, depending on disease severity and patient and physician preference.
Self-management education focused on "the appropriate use of a glucose meter, the role of diet and exercise, the importance of HbA1c testing, medication management, the management of hypoglycemia, and teaming closely with physicians in the use of staged diabetes management clinical guidelines to achieve optimum blood glucose control."59 The nurses documented every encounter in the patient's medical record, and physicians were asked to review and co-sign the entries. Each nurse also collected information from the patient and the medical record for entry into a registry.
This support was part of a larger disease management program that also included nurse promotion of diabetes clinical practice guidelines in their day-to-day interactions with patients and physicians, specialty clinic referrals, and CME sessions.38,59
External Call Center Model
In this model, self-management support is provided under the authority of an organizational entity external to the local provider and usually outside of the primary care setting. Self-management support is provided by phone from a centralized call center, rather than through face-to-face interaction. The self-management support provider, usually a commercial vendor or an independent system or plan, is external to the local health care provider, and communication with the patient's primary care clinician varies. In most other respects, this model is similar to the external on-the-ground model. Self-management support staff have access to large databases. All identified patients may be offered some form of support, but only a small subset receives personal coaching or support. In some cases, calls may also be "inbound"—that is, patients may call the center. A high return on investment is an important objective for the program sponsor. Such programs are usually offered at the initiative of a large payer and are purchased from a vendor (the following box contains an example of this model).
Example of an External Call-Center Model
This external call-center program was established by a health plan and a disease management vendor to offer self-management support to the plan's members. Members with asthma are identified through claims data and provider referrals. All individuals identified are given immunization reminders and educational materials. Claims data are used to identify a high-risk subset of members defined by recent emergency room utilization or inpatient admissions for asthma. This subset is offered the vendor's telephone support program on an opt-out basis. These services include 24-hour access to telephone consultation with a registered nurse, as well as an initial assessment by the nurse, an individualized care plan, regularly scheduled monitoring for early signs of problems, and assessment of asthma-related knowledge, behavior, and health status.
According to an article reporting research on this program, the registered nurses were employed by the disease management organization, averaged more than 20 years of experience in health care, and received specialized training focused on one or more chronic diseases. The program was described as based on the National Heart, Lung, and Blood Institute's clinical practice guidelines for asthma and emphasizes "teaching appropriate self-management behavior that includes the avoidance of triggers; the correct use of medications, inhaler, and peak flow meter; understanding of the signs and symptoms of exacerbations; knowing when to seek medical assistance; smoking cessation; avoidance of secondary smoke; and adherence to treatment plans."50
The nurses use computerized care manager software that contains standard queries and response sets. Primary care providers are sent a summary of the most current clinical practice guideline and alerts summarizing areas of concern. The disease management organization sends the health plan monthly and quarterly reports of care management measures, such as the percentage of individuals who have an action plan, flu vaccinations, a rescue inhaler, and who use daily controller medications.50
With their person-to-person interactions, the primary care, external, on-the-ground, and call-center models are the primary focus of this report. However, we mention the remote model to complete the picture of self-management support programs. Programs in the remote model are characterized by use of the Internet and/or electronic databases, scripted content, limited focus, and little feedback from patients. The support is provided via technology (e.g., computer-generated mail or automated phone calls) with no personal interaction. Reflecting their considerably lower cost, these programs usually are offered to the entire identified population, rather than a subset. In employer settings, this model may extend to people identified (often through health risk assessments) as at-risk for disease or complications.
Comparing the Models
Key experts in this field identify different strengths and limitations with each of these models. Since care management is the sole focus and core competence of the external programs, self-management support may be better planned and executed than it is in primary care settings where it is only one of a myriad of tasks that need to be accomplished, often within severe time constraints. External program staff have access to large plan or employer databases with claims, pharmacy, and sometimes laboratory data. These data allow external programs to readily identify and reach out to whole populations of patients with specified diagnoses and risks and thereby to offer a population-based framework for self-management support. Often these databases, in combination with technological investments such as predictive modeling software, offer greater economies of scale to external programs.39 In the case of external programs, patients are identified by their plan membership or employer, rather than their primary health care provider. As the plan or employer is at financial risk for the patients' health care, the external models provide a way for them to target for self-management support those individuals (and only those individuals) for whom they hold risk.
The primary care model differs from the external models primarily with respect to its integration within the primary care practice. This integration offers the possibility for the self-management support staff and the physician to more closely coordinate the self-management support with medical care and to provide support and reinforcement in both directions. Local providers have face-to-face access to patients. Medical records are housed in the practice setting, and self-management support staff there may have direct access to better clinical information (e.g., test results, recorded symptoms, physical findings, and treatment plans), as well as to discussions with clinicians. These data offer extensive decision support to staff for providing and customizing self-management support but are more difficult to use for population identification if electronic medical records are not in use. Practices implementing the chronic care model frequently create and maintain registries (of varying capacities) of populations with specified diagnoses. Without an electronic medical record or registry, self-management support in primary care settings may be limited to patients who present to the office and thus fall short of a population basis. Self-management support in the primary care model is less likely to be focused on a population subset established through predictive modeling.
The fact that the patient populations in the external models usually are defined through health plan membership may make it difficult for local providers to interface with and support the external self-management support efforts. Given that their patients may be covered by many different plans and some of these plans may use different external programs for different diagnoses, a local provider organization might have 20 or more external programs serving their patients. Some observers argue that care may be further fragmented when employers change health plans and external programs change as a consequence. The primary care model also may allow self-management support staff to be better informed about local community resources and thus encourage better patient utilization of such resources.40
A major challenge for the primary care model is the lack of clear-cut mechanisms for primary care practices to be reimbursed for the staff time and other resources needed to provide self-management support. While reimbursement for external self-management support is structured through contractual mechanisms, no similar contract arrangements are currently negotiated with most organizations providing primary care. Current reimbursement mechanisms do not easily accommodate direct reimbursement of primary care staff tasked with self-management support. Local providers, in fact, usually are not reimbursed for many care support services. Moreover, while employer purchasers may use their contracts with plans to obtain self-management support services or purchase such services directly from disease management vendors, they do not contract directly with local physician groups and clinics and thus cannot purchase the primary care model of self-management support directly. Although plans do contract directly with local providers, one plan's contract often covers only a small proportion of the provider's patients.
The tradeoffs between the two external models revolve around the advantage of in-person communication versus the disadvantage of its cost. Proponents of the external on-the-ground model stress the effectiveness of face-to-face interaction in getting people to participate. Arranging settings for face-to-face support and sending health coaches out to the various settings, on the other hand, can be considerably more expensive than a centralized call-center approach. A key consideration in the choice of models is the targeted population, as the higher intensity models are typically only used for high-risk patients.
The primary care, external, and remote models are not necessarily mutually exclusive. It has been suggested that more combinations might be seen in the future. As an example, all of a primary care practice's patients with chronic disease might be encouraged to participate in self-management support sessions offered through the practice, while those needing more support could also receive additional coaching from an external program.