Operationalizing Surveillance of Chronic Disease Self-Management and Self-Management Support

Sixty percent of US adults have at least one chronic condition, and more than 40% have multiple conditions. Self-management (SM) by the individual, along with self-management support (SMS) by others, are nonpharmacological interventions with few side effects that are critical to optimal chronic disease control. Ruiz and colleagues laid the conceptual groundwork for surveillance of SM/SMS at 5 socio-ecological levels (individual, health system, community, policy, and media). We extend that work by proposing operationalized indicators at each socio-ecologic level and suggest that the indicators be embedded in existing surveillance systems at national, state, and local levels. Without a robust measurement system at the population level, we will not know how far we have to go or how far we have come in making SM and SMS a reality. The data can also be used to facilitate planning and service delivery strategies, monitor temporal changes, and stimulate SM/SMS–related research.

We sought surveillance indicators that were sensitive to change, were actionable, had easily definable numerators and denominators, had face validity, were low-cost, required minimal effort to collect and analyze data, and -most importantly -had an existing system available to collect the data. If following a recommendation from Ruiz et al (3) was not feasible per these criteria, we explored other possible indicators to identify ones with the desired characteristics.
To standardize surveillance, we adopted the definitions in the strategic framework for multiple chronic conditions (7). We debated attempting surveillance of a single chronic disease or all chronic diseases. Should we select a single condition (like diabetes) and ascertain SM/SMS for that "sentinel" as an indicator of what is happening in other conditions? Although single conditions are likely to have specific SM activities (eg, a person with diabetes testing their blood sugar), our notion of SM addresses cross-cutting dimensions rather than disease-specific activities. Thus, we concluded that we needed to measure SM/SMS dimensions relevant across all chronic diseases, because focusing on a single condition might not be representative.
For each indicator proposed by Ruiz et al, we attempted to craft a question and specify the potential data source and numerator and denominator necessary for calculation of the indicator. As we attempted to operationalize the Ruiz et al concepts (3), we encountered consistent feasibility problems. First, we had difficulty defining a "case" of SM or SMS (the numerator) and the relevant denominator. We also lacked existing data-collection systems. The remainder of the article summarizes our deliberations on the indicators we discarded (Table 1) and our recommendations for indicators at each socio-ecological level (Table 2). What Won't and What Might Work for Surveillance of Self-Management and Self-Management Support Self-management at the individual level SM at the individual level involves a variety of activities, including adopting healthy behaviors (eg, not smoking, appropriate diet, taking medications), action planning, self-monitoring, coping with emotions, managing disability, and navigating the health care system (5,6,8). SM objectives (disease control, symptom control, prevention of deterioration) vary by condition (9). In an attempt to identify core cross-cutting elements of SM, attendees at a 2014 Self-Management Alliance meeting were asked to identify observable actions that would indicate that a person was actively and effectively self-managing his or her chronic condition. (See the Alliance's membership list in Ruiz et al [3]). The divergent written an-swers indicated that SM is an abstract concept defined in many ways, including many behaviors that are hard to differentiate from general wellness behavior and that are not necessarily reflective of chronic condition SM (eg, resolutions to lose weight or exercise). The most frequently reported observation reflected taking action to manage the condition. We reasoned that other mentioned items like goal-setting and action-planning are skills taught to help transform someone into a self-manager, that is, more a means to the end rather than the end itself. We felt that the essence of SM is being able to make wise decisions and recognize and respond to changing circumstances, adapting to the changes in the disease's trend and tempo and to the complicated realities of life with chronic disease (10,11). Taking action applies to any chronic disease and encapsulates a range of activities to respond to symptoms or situations as they arise. Action reflects the embodiment and culmination of translating education, plans, and counseling in daily life (12). The general idea of taking action to improve a chronic condition or avoid making it worse offers a comprehensive, action-oriented indicator of SM.
Because many people have multiple chronic conditions, the relative importance of the conditions may shift over time (13), and because SM may differ across conditions (9), we suggest first identifying the condition that is of most concern to consistently anchor responses. We also recommend focusing on the past 3 months to lessen recall problems (Table 2). Such questions could be used in any survey that ascertains whether respondents have a chronic disease.
Several of the recommendations of Ruiz et al (3) related to selfmanagement education, a common cross-cutting strategy to facilitate transformation into a self-manager (10). As a secondary individual-level indicator, we suggest including a question about taking a course or class (Table 2).

Self-management support at the health system level
The most commonly recognized site for provision of SMS is within health care systems; the Institute of Medicine definition of SMS specifies only health care providers (14). SMS at this level can take the form of motivational interactions (eg, motivational interviewing), collaborative care planning strategies, tools to enhance the patient-provider interaction, self-management education programs, and referrals to community SMS resources (15). Goal-setting and action-planning feature prominently in many primary care quality improvement projects (16) and allow the clinician to help create the self-management plan (17).
Action planning, or collaborating with patients to develop an SM plan, is a key means of SMS and a key activity in most health care provider-based SMS strategies (15). Privacy concerns make it difficult to ascertain what happens in clinical encounters. Consequently, we propose an indicator to capture from health care providers how frequently action planning occurred in the past 7 days. A parallel indicator is available from the National Committee for Quality Assurance's 2017 standards for recognition as a patient-centered medical home, which includes an elective element on developing a self-management plan (18). Because providing SMS is not a traditional part of health care provider training (19), we also recommend an indicator on specialized training in helping patients develop an SM plan (Table 2). These questions could be added to existing surveys of physicians or other health care providers.
Self-management support at the community level Community resources are one of the pillars of quality improvement in the original (2) and in the Expanded Chronic Care Model (20). SMS at this level could be reflected by the availability, accessibility, and awareness of SMS programs or by community structures (coalitions, parks, greenspaces, access to appropriate foods) that support SM (5,6). Clinical-community linkages could be reflected in referrals to community programs.
We encountered difficulties enumerating SMS programs and supportive community structures (Table 1). We considered identifying a sentinel SM program (eg,diabetes SM programs [21] or the Stanford Chronic Disease Self Management Program [22]) and using number of programs listed on a centralized listing of available programs (ie, a program locator) as the indicator. However, as with using SM practices for only a single disease, we were concerned with the representativeness of one program.
To examine geography-based community availability measures, we explored the use of web-based search services (eg, Google Maps) to search for nearby SMS programs. However, searches for "self-management" or "managing X condition" returned little of value. Until there is a roster of SMS programs and robust program locators, developing an indicator based on the geography of SMS programs is not feasible.
We concluded that the most feasible indicator of community-level SMS was the availability of SMS programs in the community. Directly asking people with chronic disease if they were aware of a course appeared to be the most feasible way to measure community-level SMS, because it implicitly integrates awareness and availability. This question, used in combination with a question about taking a course (see individual-level secondary indicator in Table 2), could capture both access and uptake.

Self-management support at the policy level
Policy can be an effective facilitator of SM/SMS at all levels of the socio-ecological model and is highlighted in the Expanded Chronic Care Model (20). Many major changes in public health (eg, decreased smoking, increased use of seat belts) came from policy interventions. Policy can help reduce barriers or enhance facilitators such as SMS program availability or financing. Ruiz et al (3) suggested a focus on health care systems and health insurance policies; however, we identified multiple issues that rendered them infeasible (Table 1). Instead, we identified alternative approaches in health care professional training and certification, professional practice guidelines, and federal rules and grants.

Professional training and certification
Health professionals' promotion of SM can influence patients' use. However, most providers receive no formal training in SM/ SMS (19). A content review of curriculum standards from the American Association of Colleges of Nursing (23) indicated no SMS-related standards. A survey of medical school curricula did not list SMS specifically, although some curricula included counseling for behavior change, which could be a component of SMS (24). Continuing medical education (CME) requirements for physician and nurse relicensure, on the other hand, is a potentially useful metric to follow (baseline = zero), given the ease in searching requirements and variation by state. Although some states dictate CME in certain topics (eg, pain management), none mention SMS (Table 2).

Professional practice guidelines
Treatment guidelines can reflect the integration of SMS into routine care delivery. PubMed provides a standardized searchable source for peer-reviewed literature, where physicians likely look for guidelines. Using the strategy shown in Table 2, we found mentions of SMS in practice guidelines for diabetes (n = 28), asthma (n = 6), arthritis (n = 4), chronic obstructive pulmonary disease (n = 1), and HIV (n = 1).

US government policy as reflected in the Federal Register
We found no ongoing inventory of federal activities or funding on SMS (Table 1). Monitoring issuance of federal government rules on SM/SMS over time can reflect policy changes. The Federal Register has a searchable database containing rules, notices, and proposed federal rules. Two searches (Table 2) of the database identified modest variability in numbers by year, but numbers were very low (Table 3).

Self-management support at the media level
Media coverage of SM/SMS could help set social norms and influence people with chronic disease to explore SM/SMS (5,10). We identified several impediments to exposure-based indicators proposed by Ruiz et al (3) (Table 1) but found that measuring volume of coverage was possible. Several services routinely monitor media streams (eg, newspapers, wire feeds). We searched ProQuest (ProQuest LLC) for English-language articles containing "selfmanagement" and "condition" (specifying the list of chronic conditions [7]) from 2010-2016 in newspapers, wire feeds, or magazines. On the basis those results (Table 4), we concluded that using such a database with a consistent search strategy (Table 2) can reflect the volume of SM-related content.
Advancing Surveillance of Self-Management and Self-Management Support We propose operationalized indicators for surveillance of SM/ SMS at 5 levels of the socio-ecological model ( Table 2). Measuring both SM and SMS is important, as is measuring SMS at multiple levels. SMS indicators could be considered process indicators for the eventual outcome of SM. If society effectively promotes SMS at multiple levels, more people with chronic disease may actively self-manage.
Identifying indicators that are a reasonable reflection of SM/SMS at each level was challenging (Table 1). A recurring constraint was our assumption that resources for such surveillance would be minimal. Thus, we limited ourselves to using existing surveys, systems, and databases rather than creating new surveillance systems as would have been required by the indicators proposed by Ruiz et al (3). Our proposed indicators provide only small glimpses of the big picture. Nevertheless, we believe they can provide useful information on SM/SMS status. Survey-dependent indicators will require cognitive testing.
Our proposed indicators of individual-level SM and communitylevel SMS rely on surveys of individuals with chronic diseases, such as the Behavioral Risk Factor Surveillance System. Local communities, health systems, or other organizations could gather data to reflect SM/SMS within their nexus of control. Surveillance of community-based SMS programs would be strengthened by development of consistent, accurate program-locator services, which could also facilitate program participation.
Two of our proposed indicators at the health system level (co-development of an SM plan and specialized training in SMS) require surveys of health care providers. Local health systems could survey their affiliated providers to assess SMS in their systems. One way to both strengthen surveillance and facilitate SMS delivery at the health system level would be the creation of Current Procedural Terminology codes within the Healthcare Common Procedure Coding System specific to chronic disease SMS (25).
We were unable to identify a strong surveillance indicator of SMS at the policy level. However, the indicators we proposed are promising because they are searchable and likely to detect change if consistent search strategies are used. Finally, we suggest surveillance of media coverage that includes mention of SM, and we identified a potential database for national-level surveillance. Similar media monitoring could be conducted at the state or local level by adding geographic restrictions to the search.
Consistent with the idea that "what gets measured gets done," surveillance of SM/SMS can serve as both a progress report and a motivator for expansion (26). Gathering data directly from people with chronic conditions is essential to national and state surveillance to capture SM and SMS provided beyond the health care system. Inclusion of the proposed indicators into surveys like the National Health Interview Survey, Behavioral Risk Factor Surveillance System, and Medicare Current Beneficiary Survey and adding provider-based indicators to provider surveys would advance SM/SMS surveillance. Achieving the improved functional and clinical outcomes predicted by the Chronic Care Model (2,20,26) is unlikely without leadership and investment in promoting SM/SMS. Without surveillance, we will not know how far we have come or how far we have to go in making SM/SMS an integral part of health and health care.  Need to determine whether focus should be on 1) intention or action; 2) optimal SM versus some SM versus no SM; 3) whether was done within a recent period or ever done; and 4) SM for all conditions, each condition, or just one condition (SM practices might vary by condition).
• Proportion and characteristics of individuals who can articulate setting a health-related SM goal and related action plans (3) Requires ascertaining whether the person has a goal and an action plan. What if a person had one but not the other?
• Articulating a goal is not the same as doing (intention vs action). • Uncertain which individual characteristics to select and why. • Proportion of individuals attending a series of SM education sessions in a health care setting that help solve health-related problems (3) Uncertain whether a series of sessions is necessary or whether one session is enough.
• Unclear why sessions are restricted to a health care setting. • Proportion of individuals who report receiving support for or assistance with their SM goals in the past year (3) Must first establish whether the person has a goal(s Unclear whether solving a problem related to health is different from solving a problem related to a chronic disease. Doesn't necessarily equate to self-managing a chronic condition. • Impact of solving just one problem unclear. • Long recall period. • Proportion of individuals with chronic disease who regularly self-monitor their chronic disease symptoms Unclear how to define "regularly." • Meaning of "self-monitoring" is unclear; for hypertension it may mean taking blood pressure, for diabetes it may mean measuring blood sugar, for conditions associated with pain it may mean quantifying pain on a scale. Unclear what to look for in a patient health record and whether it needs to be updated at each visit.    (3) High percentage of providers choose not be a part of the PCMH quality improvement process. • PCMH standards were established in 2011 and revised in 2014 and 2017; SM elements have changed in each revision. The 2017 standards contain 2 core elements (Team-based Care-9, and Performance Measurement-14) and one elective element (Knowing and Managing-22) that are too inclusive to assure that they reflect SMS.
• Proportion of health care systems that link to community resources offering SM support (eg, direct referral to programs, follow-up to see if an individual attended) (3) What if resources were offered by the system and there was no need for a community link? • Unclear what would qualify as a link or which community resources it should be linked to. • Unclear how data would be ascertained. • Proportion of individuals who engaged in a process with a health system that significantly changed their ability to manage their health problem (3) Implies a survey of patients in a system. • Unclear what individuals would be asked in order to ascertain whether change was significant. • Unclear what is considered a process. • Proportion of health care professionals who received training on working with patients to set and monitor selfmanagement goals (3) Unclear what type of training should be included or whether it would differ by specialty. If focused on a "sentinel" delivery system (eg, YMCAs, senior centers), unclear how to determine if these facilities offer SMS programs.
• Proportion of SM education/SM support programs by organization types in given counties (3) Unclear what defines an SM education program or an SM support program. • Unclear which organizations should be included and how they would be found and enumerated. How would a program be defined (eg, does an ongoing exercise class count the same time as a onetime offering of an SMS class)?
• Unclear denominator. • Implies multiple surveys of community organizations. • Proportion of communities actively promoting the construction of supportive environments that encourage people to be active (3) Unclear how to define active promotion, construction, a supportive environment, and encouraging being active.
• Unclear how to define a community and if communities of different sizes counted equally. • Unclear data source. • Proportion of communities that actively promote programs that offer affordable healthy foods (3) Unclear what specifically defines active promotion and a program that supports healthy affordable food.
• Unclear what defines affordable or what happens if healthy food is offered but expensive. • Unclear data sources. • Proportion of communities that have infrastructure/ partnerships for organizations in the community to work together to foster SM among people with chronic diseases (3) Unclear how to define "fostering," "community," "infrastructure," "partnership," or "working together" (once? ever? ongoing?).   Proportion of individuals being encouraged to attend community programs (3) Unclear how "being encouraged to attend" and "community programs" are defined. • Unclear how data could be collected reliably. • Proportion of population with chronic disease with SM support or education programs available within 5 miles of home (20 miles for rural areas) Difficulty geo-locating people with chronic disease.
• People with chronic diseases may not be distributed the same as total population. • Need geo-location of person and program to measure distance from person to program. • Access to programs has multiple possible definitions. If access is an issue, is it immediate proximity (distance between a location and closest program), availability in an area (number of programs within an area), availability in immediate area (number of programs within a given distance of a point), or average distance between a location and all (or individual) SMS programs?
• Proportion of community benefit surveys that address the availability and/or quality of SM support programs Lack of standard data collected by hospital associations.    Table 1. Surveillance Concepts and Associated Measurement Challenges in Operationalizing Surveillance of Self-Management and Self-Management Support

Socio-Ecological Level and Concept Indicator Measurement Challenges
Proportion of nursing schools with SM support included in curricula No centralized repository of curricula.
Proportion of physician specialty certifications requiring SM support training Exploration of new certification exam content (because need for exams on recertification is not consistent) in the specialties most likely to manage patients with chronic conditions (ie, internal medicine, family medicine, and preventive medicine) showed that SM is not explicitly mentioned per se in any of these exams.
Proportion of employers who provide SM support as an employee wellness benefit Examined Kaiser Employee Health Benefits Survey (http://kff.org/report-section/ehbs-2015-summaryof-findings/); questions not specific enough to be sure they represent SMS.
Federally funded research or programs on SM Number of results in grants.gov (www.grants.gov/web/grants/search-grants.html) was low, highly influenced by choice of keyword, and produced different results on different days.

Media Level
Broad issues with concepts proposed for this level Unclear how to determine whether an individual was exposed and whether exposure was relevant to the chronic disease.
• Recall period unclear (eg, ever? last month? last week?). • Proportion of individuals exposed to media campaigns locally, regionally, or nationally that promote SM, including collaborative goal-setting (3) Unclear whether collaborative goal-setting must be included if other aspects of SM were covered. • Unclear how to determine whether an individual was exposed. • Denominator unclear (adults, adults with chronic condition, adults exposed to any media campaign?).
• Proportion of newspaper columns or radio/television stories on SM support (3) Unclear what elements must be present to constitute SMS content and how those elements would be detected (ie, what specifically would we search for?) Would it be disease-specific?
• Unclear denominator (the sum of all TV, radio, and newspaper stories?) and uncertain how those sums would be ascertained.
• Proportion of individuals exposed to public health campaigns promoting SM Unclear what constitutes a public health campaign or how to determine whether an exposure occurred.
Proportion of individuals exposed to social media campaigns promoting SM Unclear what constitutes a social media campaign or how to determine whether an exposure occurred.
Proportion of product commercials that articulate SM as part of their product's use Unclear how SM would be defined in this context. • Unclear how to enumerate products and commercials (billboard, print advertisement, television/ radio advertisement?) • Abbreviations: CPT, current procedural terminology; DSMT, diabetes self-management training; HCPCS, Healthcare Common Procedure Coding System; PCMH, patient-centered medical home; PCP, primary care provider; SM, self-management; SMS, self-management support.  Responses to Q2 can also be used to calculate patterns of disease trajectory (eg, among people who report arthritis as their most concerning condition, what proportion report better, worse, or stayed the same during the reference period) and show how these proportions change over time for a given condition. Q2 responses can also be combined with Q3 responses to compare proportions taking self-management action given the condition's getting better or worse or remaining the same. Both of these secondary analyses can be used to further target self-management support interventions.

PREVENTING CHRONIC DISEASE
Questions: Only those with 1 or more chronic diseases should be asked this question. Q1. Earlier you told me that you had (fill in list of chronic diseases [eg, diabetes, arthritis, heart disease]). Which of these conditions concerns you the most? Q2. Do you feel as if your (name of condition of most concern) has gotten better, worse, or stayed the same in the last 3 months?* Better A. Worse B.
Stayed the same C.
Not sure/don't know D.
Refused E. *Three-month reference period can be reduced (eg, to one month) to match other reference periods used on the survey. Q3. Based on what you just told me, have you changed anything about how you manage your (name of condition of most concern)? Yes (optional to go to Q4) A.
No B.
Not sure/don't know C.
Refused D. Numerator: Q3 = A (yes) Denominator: Q2 = A + B + C + D +E (better, worse, same, don't know/not sure, refused) Optional 4th question if more detailed information is desired Q4. What did you do? __________(Record verbatim). Probe: Did you do anything else? Interviewer note: type of actions could be avoiding certain foods or environmental situations, performing specific types of activities, etc.
Secondary Indicator: Proportion of individuals with one or more chronic conditions who report they have taken a course or class to help manage their most concerning chronic condition. Rationale: Provides estimate of proportion of people attending a selfmanagement program. Can be asked in conjunction with community-level course awareness indicator question to compute proportion who are aware and attended (if they know, will they come?). Attending a class is a selfmanagement behavior that cuts across chronic diseases; attendance is one means to the end of becoming a good self-manager. Proposed data source: Surveys that assess chronic disease status such as National Health Interview Survey (https://www.cdc.gov/nchs/nhis/index.htm), Behavioral Risk Factor Surveillance System (https://www.cdc.gov/brfss/ index.html), or Medicare Current Beneficiary Survey (https://www.cms.gov/ Research-Statistics-Data-and-Systems/Research/MCBS/index.html).
Question: Only those with 1 or more chronic diseases should be asked this question. Q1. Have you ever attended a course or class on how to manage your (name of condition of most concern)? Yes A.
No (Optional to go to Q2) B.
Not sure/don't know C. Refused D. Numerator: Q1 = A (Yes) Denominator: Q1 = A + B + C (Yes, No, Not sure/don't know) Optional second question if more detailed information is desired Q2. There are many reasons people do not participate in a course or class to help them manage their (name of condition of most concern). Why have you not attended a course or class to help you manage that condition? _________ _(Record verbatim). Probe: Anything else?

Health System Level
Indicator: Proportion of chronic disease patients seen by clinician in past 7 days who have a jointly developed self-management plan.
Question: Some clinicians think it advisable for patients with chronic diseases to develop their own plan of action for managing their condition(s). Think about the patients with chronic disease you saw in the past 7 days. With what percentage of them have you or a member of your practice team Abbreviations: CME, continuing medical education; NCQA, National Committee for Quality Assurance; SM, self-management; SMS, self-management support. www.cdc.gov/pcd/issues/2018/17_0475.htm • Centers for Disease Control and Prevention (continued)  Indicator: Proportion of clinicians reporting they or a member of their practice team received specialized training on how to help patients develop a self-management plan. Rationale: Health care providers traditionally receive little or no specialized training in SMS techniques such as motivational interviewing. Potential data source: Survey of health care providers, such as those conducted by the Physicians Foundation (https://physiciansfoundation.org/ research-insights/biennial-physician-surveys-patient-surveys/) or other state or local organizations.

Question:
Have you or a member of your practice team received specialized training in helping patients develop a self-management plan? Such specialized training includes motivational interviewing, stages of change, or other patient activation techniques. Yes A.
No B.
Don't know C. Refused D. Numerator: A (yes) Denominator: A + B+ C (yes, no, or don't know)

Community Level
Indicator: Proportion of individuals with 1 or more chronic conditions who report that they are aware of courses or classes in their community to help manage their most concerning chronic condition. Rationale: Awareness of classes is a proxy for community availability and also the communication of such availability. If classes exist but people are unaware of them, they are effectively not available. Proposed data source: Surveys that assess chronic disease status such as National Health Interview Survey (https://www.cdc.gov/nchs/nhis/index.htm), Behavioral Risk Factor Surveillance System (https://www.cdc.gov/brfss/ index.html), or Medicare Current Beneficiary Survey (https://www.cms.gov/ Research-Statistics-Data-and-Systems/Research/MCBS/index.html).
Questions: Only those with 1 or more chronic diseases should be asked this question. Q1. Earlier you told me that you had (fill in chronic diseases [eg, diabetes, arthritis, heart disease]). Which of these conditions concerns you the most? Note: This question can be eliminated if individual-level self-management questions have already identified the most concerning condition. Q2. Do you know of any courses or classes in your community to help people manage (name of condition of most concern)? Yes A.
No B.
Not sure/don't know C. Measurement: Number of stories in English containing the words "selfmanagement" AND "condition" in a given year in newspapers, wire feeds, or magazines. Replace term "condition" with the individual chronic condition. Search by year. Numerator: Number of stories by year Denominator: None Abbreviations: CME, continuing medical education; NCQA, National Committee for Quality Assurance; SM, self-management; SMS, self-management support.