What is the State of Patient Self-Management Support Programs? An Evaluation
What is the State of Patient Self-Management Support Programs?
Staffing
Staffing refers to the question of, "Who are the people who actually provide the support to patients?"
Within the primary care model, self-management support responsibilities often are delegated to a non-physician member of the staff, a nurse or possibly a medical assistant, who can devote more time and attention to self-management support. The primary care physician often plays a role as well. An action plan frequently is used to facilitate collaborative decisionmaking between the primary care physician and the patient and to facilitate follow-up in subsequent primary care visits. In many cases, limited office resources for self-management support are supplemented by referral or collaboration with other resources in the community.
Many external programs staff their call centers with nurse care managers, but the programs differ with respect to the qualifications sought for these positions, seemingly reflecting uncertainty in the field regarding care manager core competencies. While some programs seek nurses with backgrounds in the specific disease, e.g., two years in acute care cardiac settings, other interviewees stressed the importance of interpersonal skills over medical knowledge—or the ability to focus on a patient's goals rather than solely on the disease. Some interviewees distinguished care management skills (with the focus on coaching and behavior change) from prescriptive educational skills and/or more traditional nursing skills. Others distinguished between generalists and specialists, seeking primary care and "broad rather than deep" backgrounds rather than masters-level nurses. Some emphasized motivational skills and the ability to foster self-efficacy rather than patient dependency. Empathy and compassion were also stressed, as were computer and telephone communication skills.
A Request for Proposal for a self-management support contract may determine staffing levels and qualifications, including if and how non-nursing staff are used. Some programs use staff with other backgrounds to help provide self-management support. Staff mentioned included social workers, pharmacists, dieticians, behavioral or mental health specialists, licensed practical nurses, medical assistants, and patient coordinators. Programs may use non-nursing staff in the call centers, along with nursing staff; staff with less medical knowledge focus on tasks such as prompting or "nagging" and transition the call to one of the nurses in the center when a medical care issue comes up. Some programs specifically look for people with roots in the community—nurses who know the community resources, in some cases, and lay health workers or "promotoras," in others.
Staffing decisions involve a number of tradeoffs. Nurses have greater medical knowledge, on the one hand, and may be viewed as more acceptable to primary care physicians whose support or cooperation the programs wish to obtain. On the other hand, nurses are expensive and in short supply. Some interviewees also argued that nursing training frequently does not emphasize behavior change interventions. Some characterized nursing education and the nursing profession as hierarchical and prescriptive and questioned whether other backgrounds might not be better suited to work collaboratively with patients. One informant stressed that staffing should be based on an assessment of major problems in the population and areas needing improvement. More than one informant discussed the need or potential for greater specialization in self-management support—differentiating self-management support tasks and looking for people with different skills for different tasks. The argument here is that coaching people on exercise takes different skills than discussing medication side effects; remote self-management support interaction requires different skills than face-to-face self-management support. Staffing levels and qualifications are directly linked to cost issues, and the argument in favor of specialization emphasized the possibility of greater reach for the same amount of investment. Smaller caseloads also are more expensive but are needed for high levels of personal interaction.
Content
The multiple staffing considerations reflect the wide variety of tasks that self-management support staff are frequently expected to perform and the variety of services patients may receive. Together, these comprise the content of the program.
The first column in Table 2 shows the different kinds of content commonly ascribed to self-management support programs. These content areas reflect the dual purpose of most programs:
- To educate people and thereby increase their knowledge.
- To coach people to change their behavior.
Educational content (at the top of column 1) primarily consists of information on the disease, its treatment, medications, and self-management. The second column shows some of the commonly mentioned prerequisites for providing this content: medical knowledge of the disease, its symptoms, treatment, medications, side effects, outcomes, and medical terminology, etc.; knowledge of the self-management tasks needed and their benefits; knowledge of the applicable clinical practice guidelines; knowledge of resources that are useful and available; and the ability and resources to communicate such information so that people will understand it. Much of the information provided is disease-specific.
Content areas frequently described as needed in order to support and coach people's behavior change are listed under "Supportive Interventions" in column 1. Much of the specific content of these supportive activities is patient-specific and relies on the results of the assessments. Thorough assessment of multiple aspects of the patient (i.e., perceptions, knowledge, motivations, confidence, skills, needs, and goals) is needed to support many of the subsequent tasks. While many of the prerequisites for the educational content are disease-specific, the prerequisites for the supportive interventions tend to be described as a core set of skills applicable across diseases. A common terminology for describing these skills was lacking, but they generally were described as psychosocial skills such as the ability to motivate, persuade, emotionally support, reinforce, build confidence, problem-solve, and work collaboratively. Programs frequently use self-management support protocols, software and information system support, and staff training to promote these skills and tasks.
Programs appear to vary in the relative emphasis they put on these different content areas. Some programs are primarily educational and information is the main content of the support. While all programs include some education, some place considerably more emphasis on supportive interventions directed towards behavior change. Even within these programs, there is variation in the types of supportive interventions. The most notable variation with respect to self-management support content had to do with emphasis on collaborative decisionmaking. Some program interviewees stressed the centrality of collaborative decisionmaking, while others never mentioned it. Where collaborative decisionmaking is key, the program (often through the care management software) is specifically designed to involve the patient in selecting and prioritizing the problems to work on, as well as the intensity and sequencing of the activities. The problem most pressing to the patient, rather than the clinical priority, is addressed. While traditional education alone may emphasize the provider's agenda, patient compliance, and provider decisionmaking, supportive interventions, particularly those involving collaborative goal setting and shared decisionmaking, move health care interactions to a focus on the patient's agenda, the patient's self-efficacy and confidence in his or her ability to change, and collaborative decisionmaking between patients and providers (Bodenheimer et al. 2005).
Patient Population Served
Who is to receive the self-management support? Most of the primary care model programs target everyone with a specific diagnosis or diagnoses. Some may offer more support to those who are more ill, but predictive modeling is not common. One primary care program, however, is developing a stratification approach, based on patient-report data, which divides patients with chronic disease into three categories and distributes resources differently to each category. The strata are defined by patients' financial status, confidence in self-care, and presence of bothersome pain and psychosocial problems (Wasson et al. 2006).
Within the external models, there is variation in determining who is eligible for self-management support. Most interviewees asserted that their programs were designed to target a subset of a population with a particular diagnosis but the definition of the subset appeared to differ. A number said that their programs focused on those with the highest severity of illness, but others described their target in different terms, such as those at highest risk, with greatest potential cost savings, "likely to have the greatest benefit," "whom they can help," "most impactable and engageable," or most receptive to change. Most use proprietary algorithms to identify patients with a chronic condition from claims data and employ commercial predictive modeling software to triage patients for services. (Some programs even develop and maintain their own predictive modeling software.) The specific combination of methodology for identifying and triaging patients, however, is considered the "secret sauce" and said to vary from one program to the next. The proportion of the population that is included in the targeted subset varies too. One informant said that the number of patients selected for personal contact—"how far down the list they go"—depends on the specifications of each contract.
Self-management support programming is customized to the categories identified through predictive modeling. The external programs differ in whether they limit their self-management support services to the subset or provide something to everyone in the identified population but with more intensive self-management support only to those in the top subset. Many programs may offer written materials, Web site access, or possibly one phone contact to those below the threshold. For those above the threshold, the intensity of the outreach, engagement contacts, and coaching contacts may vary with the severity of illness. As an example, an external call-center program for heart failure patient support stratified participants (people identified with heart failure who agreed to enroll) into three categories of high, medium, and low risk. The protocols specified that high risk individuals were to receive 16 calls per year; medium risk patients were to receive seven calls and low risk patients two calls (Berg et al. 2005).
Patient engagement in the program is generally considered a major challenge. External programs usually use phone calls and mailings to get people to participate in the program. Programs differ in the perseverance which is demanded of the staff doing the outreach, most notably in the number of contact attempts. They likewise differ in the speed of engagement. While some programs required participants to "opt-in," others use a more passive "opt-out" approach. Programs also differ in approach. At least one self-management support provider, for example, begins using motivational interviewing techniques during the initial contact to encourage program participation.
Incentives are considered useful for encouraging program participation and increasing self-management effort and activities. Small incentives, such as water bottles, totes, and baby shoes, were mentioned. One informant expressed the opinion that substantial dollar amounts would be needed to have a substantial impact. Employers have begun to offer reductions in health insurance contributions to employees who participate in such programs. Distinct from the self-management support programs, some employers are offering financial incentives to employees to encourage better medication compliance. Examples include change in formularies to insure that critical chronic care drugs are on the least expensive tier and eliminating co-pays for generic drugs.
Information Support
Information databases for self-management support decision support and population identification may vary considerably and may be non-existent in primary care model programs. Registries are a common approach to developing a database and vary from paper-and-pencil versions to sophisticated electronic registries. When registries or electronic databases are missing, the reach of the program may be limited towards those seeking care.
In all models, data commonly are collected by interaction with patients (e.g., signs and symptoms of the disease). External model programs, on the other hand, characteristically rely on large electronic databases to identify the target population. Most, if not all, use claims data, which are convenient but frequently limited by inconsistencies in reliability and validity and time lags in reporting. Some access pharmacy records, lab reports, hospital admissions records, health risk assessments, and/or other data as well. These same data may be used for decision support.
As with identification of the eligible population, the data available affects the type and amount of decision support available to the self-management support providers for their interactions with patients. Our interviews suggested that there is great variation in decision support provided within the external models. While not all programs may have computerized decision support, a number of our interviewees stressed the critical importance of utilizing timely patient information in self-management support contacts. Timely access to patient information is one of the factors which distinguish self-management support from traditional didactic education. The kind of patient information used varies from visit, testing, and prescription utilization data that allow reminders and follow-up of the patient's management of health care processes to clinical data, such as lab results, used to guide contact frequency and coaching content to patient assessment data that guide collaborative decisionmaking and problem-solving or motivational interviewing or other self-management support content. In some programs, the care managers input patient assessment data into care management software which then helps to define or prioritize the patient's self-management problems, design a care plan, implement self-management support activities, and guide reassessment.
Some care managers have access to data from remote monitoring devices. Some have data from electronic medical records. One informant said the program shared these care management data with the patient's primary care physician.
Protocols
What are the guidelines for self-management support programs? When asked about protocols, many interviewees refer to clinical practice guidelines which they see as underlying their self-management support programs. Self-management support programs vary in the degree to which they build on clinical practice guidelines, and, for the most part, the clinical guidelines call for self-management support activities such as patient education, but do not fully prescribe how these self-management support activities should be conducted. Box 4 provides the sections of the American Diabetes Association (ADA) Guidelines that pertain to self-management support, as an example. National guideline development efforts have yet to focus on detailed self-management support guidelines.
Box 4. Protocols Self-Management Support Provisions of the American Diabetes Association (ADA) Guidelines ADA guidelines for diabetes care state that any diabetes management plan should recognize diabetes self-management education (DSME) as an integral component of care. A variety of strategies and techniques should be used to provide adequate education and development of problem-solving skills in the various aspects of diabetes management. Implementation of the management plan requires that each aspect is understood and agreed on by the patient and the care providers and that the goals and treatment plan are reasonable. (American Diabetes Association 2006). The ADA's specific recommendation for DSME follows:
The national standards are structural criteria. ADA-recognized DSME programs have staff that includes at least a registered nurse and a registered dietitian; these staff must be certified diabetes educators or have recent experience in diabetes education and management. The curriculum of ADA- recognized DSME programs must cover all areas of diabetes management, with the assessed needs of the individual determining which areas are addressed. All ADA-recognized DSME programs utilize a process of continuous quality improvement to evaluate the effectiveness of the DSME provided and to identify opportunities for improvement. (American Diabetes Association 2006). |
Self-management support protocols vary by contract, but many contract negotiators know little about protocols. Purchasers may leave a number of the protocol specifications up to the vendor (or the protocol decisions essentially may follow from the choice of vendor), but a number of vendors also made reference to specifications included in a Request for Proposal.
Guidelines for how to provide self-management support appear in training materials, care management software, other decision support tools (e.g., scripts), and documentation formats for assessments, contact logs, goal summaries, action plans, treatment plans, etc. Possibly because they are somewhat "buried" in the software and materials, protocol variations may be difficult for prospective purchasers to discern. They also may be considered proprietary information. Protocol specifications may vary with the goals of the program and the contract under which the program is provided (e.g., how many outbound calls does an employee want to pay a vendor/health plan to provide).
Programs particularly vary in the frequency and intensity of contact, the degree of scripting they provide for support staff, and the content of the communicated support. Frequency and intensity of contact are affected by the goal of the program (e.g., wellness, risk prevention, or disease management) and the contract/business arrangements of the program including the intended reach of the program. The interviews and literature review suggest that most programs have at least one initial outbound call. Subsequent calls are largely determined by the program contract and the alerts raised by monitoring of patient data, often assisted by decision support tools. Higher-risk patients or patients with more severe/complex diseases generally get more contacts.
Interviewees rarely spoke in detail about specific content of sequential sessions. Some literature was more explicit. A seven-week, small-group self-management support program for patients with one or more chronic diseases, for example, is guided by a detailed manual and covers the following topics: overview of self-management and chronic health conditions, making an action plan, relaxation/cognitive symptom management, feedback/problem solving, anger/fear/frustration, fitness/exercise, fatigue management, healthy eating, advance directives, communication, medications, making treatment decisions, depression, informing the healthcare team, and working with healthcare professionals (Lorig et al. 2001).
In addition to assessment forms and questionnaires, many programs provide scripts to their self-management support staff. While some programs provide explicit scripts to be used during phone calls with patients, the majority of programs have scripts that include typical dialogue for a given interaction (e.g., high-risk patient, or early vs. late-stage interaction), and support staff are encouraged to "use their own words" and personalize the conversation to the patient. In general, there appear to be differences in the degree of prescriptiveness of program protocols. While some emphasize that protocols are useful for monitoring self-management support staff performance and insuring uniformity in program implementation, concern was also expressed that prescriptive protocols may reduce flexibility in a service that needs tailoring and personalization.
Staff Training
While a number of the primary care model pioneers received training through the Institute for Healthcare Improvement or government-sponsored quality improvement collaboratives, non-collaborative training opportunities are not numerous and much training is on-the-job. Training is available for facilitators of some structured workshops, such as the Chronic Disease Self-Management Program and Tomando Control de su Salud (Stanford Patient Education Research Center 2007). One informant stated that a good self-management support training school with a good curriculum and trained instructors is sorely needed.
While some experts voiced the opinion that external model staff training is critical because much of the self-management support content and skills are not taught in nursing school (or other professional schools), there was uncertainty about the length and format of training needed. The content appears to parallel the staffing qualifications emphasized: care management skills, motivational interviewing, cognitive skill-building, disease knowledge and clinical updates (including common comorbidities), medications, self-management skills, program protocols and standards, care management software, and computer skills. Additional content items mentioned included exercise, nutrition, sensitivity to patient issues, lower literacy skills, and behavioral health information. Subcontracting out for training was infrequently mentioned. Descriptions of internal training varied considerably. One informant said that self-management support provider training came under the Nursing Director's nurse education responsibilities. Two interviewees said their programs provided a six-week training program plus subsequent mentoring. Others mentioned ongoing presentations, monthly meetings, onsite mentors, call monitoring, shadowing, and feedback of nurse-specific outcomes data. One informant claimed that there are lots of care manager training programs on the Internet, but their content and quality vary.
Communication with Patients
Program components in the primary care model may include various combinations of communication—face-to-face education, phone follow-up, Web site access, group visits or sessions, and referrals to classes and/or community resources. Patients may receive action plans, visit reports, and/or tools such as calendars or diaries which combine recording opportunities for self-management monitoring (e.g., recording of daily weights) with information to support decisions on when to take specific actions.
Many external model programs offer Web site access and written education materials in addition to telephone contact and face-to-face contact (in the case of the external on-the-ground programs). Some also use other modes of communication, such as E-mail and telephone hotlines. Group visits, classes, and home visits are less frequently utilized.
Communication between Physicians and Self-Management Support Staff
Given its location within the primary care setting, the primary care model usually involves communication between the self-management support staff and the primary care physician. The primary care physician frequently plays a collaborative role in the self-management support (through use of action plans, for example, or reinforcement of goals and efforts to change behavior). The self-management support providers and primary care physicians may share treatment plans, as well as action plans. The nature and degree of their interaction may vary from a team structure with extensive collaboration to exchange of information through charting or electronic records.
External model programs differ considerably in their approach to communicating with the primary care physicians. Some programs foster little or no communication between the self-management support staff and the patient's primary care physician, while others strive for a collaborative relationship. One of the arguments voiced by programs that minimized communications with primary care physicians is that, to foster independence, the patient, not the self-management support program, should communicate with the primary care physician. If the patient is having difficulty communicating with the primary care physician, the appropriate role for the self-management support provider, according to this argument, is to coach the patient and help him or her build the skills needed for good communication. A different reason for minimizing communication with the primary care physician rests with the claim that physicians do not want such communication; they already are bombarded with paper and telephone calls and, with only a few minutes with each patient, they do not want to use part of that time to talk about the self-management support program. On the other hand, proponents of communication with the primary care physician argue that physician participation is critical; patients trust their physicians and physician reinforcement of self-management support leads to better outcomes.
The increasing prevalence of pay-for-performance programs that reward physicians for compliance with guidelines creates an incentive for primary care physicians to collaborate with external self-management providers, as those providers may be able to point out patients that require certain tests or drugs.
In those programs that encourage communication with the primary care physician, the form of the communication may differ. Not all programs notify the primary care physician when patients enter the self-management support program. Some program protocols call for sending information to the primary care physician in the form of assessment summaries, relevant clinical practice guidelines, identified risk factors, gaps in care, acute symptoms, or self-management issues being addressed. Communication may be by mail, fax, telephone messages, or E-mail. Some externally-run programs provide self-management support in primary care settings. Other communication strategies mentioned include:
- Provide the primary care physician with access to patient data. (One program is putting its system on the Internet so the provider has access to data gathered by the program.)
- Send reminders and alerts about possible treatment gaps.
- Support the primary care physician's treatment plan.
- Give decision-support software to community providers.
- Request patient data from physicians.
- Put administrative personnel in physician offices to check clinic schedules, print patient data, and put data in charts.
As part of the self-management support provided, some programs prepare the patient for their primary care physician visit. One program encourages patients to prepare to ask three questions at every primary care physician visit and works with them to prioritize their questions beforehand. Another program will occasionally conduct three-way calls (with patient and primary care physician) to facilitate the patient's communication. Others simply review issues with the patient before a primary care physician visit. In one program, the nurse may accompany the patient into the exam room to help with the patient-primary care physician communication.
In programs with physician communication, self-management support staff typically build relationships with non-physician personnel in primary care physician offices. While two-way communication may be a goal for some programs, most of what communication exists at this point is one-way; self-management support providers generally receive little information or response from the primary care physicians. Some programs are considering pay-for-performance or "pay-for-participation" incentives to get physicians to collaborate with self-management support providers.