Applying Community-Engaged Intervention Mapping to Preparing Nurse Scientists

Abstract

Background:

Preventing and managing chronic illness necessitates multilevel, theory-based interventions targeting behaviors, environmental factors, and personal determinants that increase risk for illness onset, greater burden, and poorer outcomes.

Objectives:

The purpose of this paper is to provide the basis for multilevel interventions, describe community-engaged intervention mapping as an approach to designing theory-based interventions, and discuss potential benefits of applying community-engaged intervention mapping in preparing nurse scientists to build programs of interdisciplinary research in preventing and managing chronic illness.

Methods:

Community-engaged intervention mapping integrates two methodological approaches: intervention mapping and community-engaged research.

Results:

The six-step intervention mapping approach provides a logical structure for preparing nurse scientists in designing, adapting, and implementing multilevel, theory-based interventions. Community-engaged research approaches offer principles and direction for engaging patients, clinicians, community members and other stakeholders throughout the research process. Integrating these methods retains the theoretical integrity of interventions, improves the relevance and timely completion of the research and its products, and enhances intended beneficiaries and the community’s understanding, trust, and use of the results.

Discussion:

Potential benefits of preparation in community-engaged intervention mapping to nurse scientists and nursing science include explicit consideration of multilevel factors influencing health. Additional benefits include guidance for linking relevant constructs from behavior- and environment-oriented theories with evidenced-based methods for affecting desired changes in care and quality of life outcomes. Moreover, enhancement of the theoretical fidelity of the intervention, explication of the mechanisms influencing change in the primary outcome, and improved relevance and feasibility of interventions for intended beneficiaries and potential adopters are other benefits.

Preventing and managing chronic illness necessitates multilevel interventions that target behaviors, environmental factors, and personal determinants (modifiable cognitive and affective factors) contributing to risk for illness onset, greater burden, and poorer care and quality of life outcomes. Multilevel interventions that are well designed and effective are theory-based. Theory aids identification of the array of factors that contribute to health-related problems. Moreover, theory can guide nurse scientists in (a) selecting change methods, (b) devising an evaluation plan, and (c) identifying which components underlie change in the primary outcome and thus must be included when the intervention moves into practice (Bartholomew & Mullen, 2011). Theory can also direct selection of characteristics not expected to change because of the intervention (sex, age group), and baseline measurements (stage of readiness to change) that typify subgroups with distinct responses to interventions (potential moderators). This information can guide development of algorithms for tailoring intervention components and dose in efficacy studies (Sidani & Braden, 2011), and identification of those most likely to benefit when efficacious interventions are moved into practice (Becker et al., 2020). Community-engaged intervention mapping is well suited for directing the preparation of nurse scientists in systematically designing multilevel interventions to prevent and manage chronic illness and explicating the use of theory in that work. In this brief report, we: (a) provide the basis for multilevel interventions in preventing and managing chronic illness; (b) describe the methods (community-engaged research approaches and intervention mapping) that comprise community-engaged intervention mapping; and (c) discuss potential benefits of their application in preparing nurse scientists specifically in designing, testing, and implementing multilevel interventions to prevent and manage chronic illness.

Multilevel Interventions

Over the past decade, nurse scientists and other researchers have adopted socio-ecological models in their pursuit to understand how factors within and across individual and environmental levels contribute to the onset, burden, and management of chronic illness (Whittemore et al., 2004). In the context of health equity research that examines the higher burden of chronic illness among socially and economically disadvantaged populations, social determinants of health (SDOH) consistently explain substantially larger proportions of variance in health than individual level factors (Braveman et al., 2011). With SDOH operating within multiple environmental levels (neighborhood, organizations, society), public health scholars have increasingly pointed to the need to move beyond downstream interventions—that is, those targeting factors operating at the individual level (Castrucci & Auerbach, 2019; Hall et al., 2018). Given the growing body of literature demonstrating that SDOH are consistent and robust predictors of disparities in chronic illness prevalence and outcomes, the empirical and theoretical bases for developing multilevel interventions to prevent and manage chronic illness is compelling (Agurs-Collins et al., 2019; Hall et al., 2018).

Multilevel interventions target factors at two or more levels that influence the onset and trajectory of chronic illness. Within and across levels, these factors operate through dynamic, complex, and interdependent mechanisms to affect health and quality of life (QOL; Institute of Medicine, 2012). Biological, physiological, psychological, and behavioral factors interrelate within the individual level, and with sociocultural, economic, and physical factors operating at various levels in peoples’ interpersonal interactions with others (caregivers, social network members, clinicians) and with their physical and social environments, indicating that intervening at more than one level is needed (Braveman et al., 2011). For example, intervening to improve cognitive and behavioral outcomes for children with blood lead levels is inadequate when the community’s drinking water source is at the root of the problem (Hanna-Attisha, 2018). Similarly, to reduce maternal morbidity, preterm birth, and mortality among Black women, simply emphasizing stress reduction strategies during pregnancy will not address the sources of bias and differences in the level or quality of care provided in health care settings by race—a critical factor believed to drive the persistent inequities in maternal outcomes by race (Howell, 2018).

The need for multilevel interventions to address the problems provided by these examples is supported by the premise that factors influencing health are interdependent and have cumulative effects on outcomes across levels. Expansions in thinking about the conceptual basis for multilevel interventions by public health scholars assert that this view is incomplete and overly simplistic, as interventions can be designed to act synergistically across levels to achieve even larger and possibly more durable effects (Weiner et al., 2012). Scholars also warn that synergistic or even cumulative effects of multilevel interventions should not be presumed as the effects of one aspect of the multilevel intervention could inhibit and possible conflict with those at other levels if the theoretical assumptions underlying interactions across interventions are not valid (Weiner et al, 2012).

A growing body of evidence demonstrates the efficacy of multilevel over single-level interventions when risks for poorer care and quality of outcomes are compounded across levels (Breslau et al., 2016), as is the case with socially and economically disadvantaged populations (Paskett et al., 2016). Given their complexity, designing and implementing multilevel interventions pose novel theoretical and practical challenges that go beyond those encountered with single-level interventions (Weiner et al., 2012). Although beneficial for all intervention types, methods that demand both elucidation of the theoretical rationale underlying the intervention and engagement with intended beneficiaries and adopters throughout the research process could be particularly advantageous for designing and testing multilevel interventions. An approach that integrates community-engaged research (CEnR) approaches and Intervention Mapping (IM) holds promise for advancing intervention research to prevent and manage chronic illness. We refer to this integrated approach as community-engaged IM and describe why this approach is well-suited to address intervention development challenges within and/or across levels.

Methods

As shown in Figure 1, community-engaged IM refers to the integration of two methodological approaches—CEnR approaches and IM (an iterative six-step approach widely used in public health to develop health promotion programs) to design, adapt, and/or implement scalable theory- and evidence-based interventions that consider the multilevel factors influencing health. Specifically, community-engaged IM invokes engaging representatives from various stakeholder groups (patients, caregivers, clinicians, organizations, communities) in research processes and decision-making as the research team carries out IM processes.

Figure 1.

Community-Engaged Intervention Mapping

Intervention Mapping

IM involves three steps. Step 1 involves describing a health-related problem, its effects on health, and the population at risk. Step 2 focuses on ascertaining and then selecting modifiable behaviors, environmental factors, and personal determinants (knowledge, attitudes, beliefs, emotions, skills) as targets for interventions to ameliorate the problem and improve outcomes. Step 3 involves using evidence-based taxonomies (Kok et al., 2016; Michie et al., 2005, 2013) and tables (Eldredge et al., 2016) derived from meta-analyses of intervention studies to link theoretical constructs with methods shown to modify those factors and thus promote cascade of changes needed to achieve the program goal. Then, (Step 4) the intervention protocol and other study materials are produced; and (Step 5) a tentative plan for future adoption is crafted; as (Step 6) is a comprehensive plan for evaluating intervention outcomes and study processes using a rigorous design and mixed methods.

Noteworthy features of IM that make it suitable for the complexity of designing multilevel interventions for the prevention and management of chronic illness include: a socio-ecological perspective which acknowledges that relationships within and between individuals and their environments influence health; application of theory and evidence; and early planning for sustainability in the real world (Belansky et al., 2011). IM is not a theoretical framework, nor does it require that an intervention or its components draw on a single theory or theory type. Rather, IM supports an approach in which relevant theoretical constructs are selected from among those prevalent in behavior- and environment-oriented theories, linking those constructs to methods shown to effect change on those constructs in prior studies to design the intervention, and then planning for evaluation and future widespread implementation (Fernández et al., 2005). Engaging representatives from stakeholder groups is integral to IM (Fernandez et al., 2019). CEnR approaches offer direction for effective engagement.

CEnR

CEnR encompasses an array of approaches that commit to inclusivity and recognize the value of engaging representatives of those intended to benefit from the intervention, other stakeholder groups, and potential implementers in the research process (Clinical and Translational Science Awards [CTSA] Consortium, 2011). Engaging with representatives from these groups as equal partners builds on strengths and resources, improving the relevance of the research and its products. This engagement also allows for identification of the complex web of causation of the problem and broad realm of intervention targets. With engagement, there is increased feasibility that the research will be completed and that the population’s or community’s understanding, trust, and use of the results are ensured (Eldredge et al., 2016; Forsythe et al., 2018). These approaches invoke equitable decision-making throughout the research process and building trust among the community of interest, academic researchers, and other stakeholders (Jagosh et al., 2015). Incorporating these principles into intervention research increases the likelihood that community representatives will feel comfortable sharing critical, sometimes sensitive, information about health problems most relevant to them, contributing factors, practical ways of applying selected methods, and preferences for delivery mode and the look, feel, and language used in program materials and delivery platforms (Schulz et al., 2011).

Enacting CEnR entails including stakeholders in steering committees, advisory boards, and other decision-making bodies in research. Usually, to simplify achievement of project goals, numbers of stakeholders are small relative to the size of the communities they represent. We define “community” as a group of people linked by geographic proximity, faith, and/or conditions that affect their health and QOL, with researchers typically traveling to the communities they partner with when using CEnR approaches (CTSA Consortium, 2011). While usual ways of conducting CEnR have clear benefits, challenges exist. First, forming and maintaining meaningful and trusting CEnR partnerships takes time, which can slow timelines (Jagosh et al., 2015). Second, the small numbers of community representatives on advisory boards, etc., are unlikely to embody the full range of experiences and viewpoints within the community. Third, preventing and managing chronic illness can entail wicked problems that necessitate the wisdom of larger numbers of people to generate innovative solutions. Thus, while representatives of communities and other stakeholder groups can provide depth of expertise when designing interventions, intervention prototypes and other products of the research generally need to be vetted with larger samples. Finally, some chronic illnesses are rare or rarely occur in groups defined by age, race/ethnicity, or sex. When prevalence is low, at risk or affected individuals within communities bound by shared illness-related experiences are likely geographically dispersed which, given the associated costs, hampers CEnR.

Community engagement takes many forms. Historically, this has typically included a small, consistent group of stakeholders provide their perspectives through focus groups, interviews, or surveys. Typically, this involves meeting face-to-face, and more recently, via video or phone conference technology. With the advent of technology and social media platforms, one underexplored approach to improving the speed and reach of applying CEnR to intervention research with chronic illness populations is crowdsourcing. Through crowdsourcing, an organization can disperse idea generating, problem-solving, and specific tasks to a motivated and engaged group, producing relevant, community-vetted, mutually beneficial results faster and better (Brabham, 2013; Brabham et al., 2014).

Originally used to advance stakeholder involvement in other sectors (public transportation, urban planning, policy making; Brabham, 2013), crowdsourcing is now used in chronic illness-related intervention research. For example, crowdsourcing was used to design and test web-based mobile interventions to improve condition management by parents of very young children with type 1 diabetes (Wysocki et al., 2018) and to evaluate personal determinants of lung cancer screening (Monu et al., 2020). In our digital age, advantages of crowdsourcing include broad community engagement, more citizen science, implementation-primed interventions, and high-quality research outcomes. Nurse scientists and others who use crowdsourcing in their research must be prepared to address issues such as transparency about their intentions and commitment to the crowd. Researchers must also agree to social media platform terms of use and be prepared to address intellectual property rights. Additionally, researchers will have to work with their institutional review boards (IRB) on determining which tasks asked of the crowd should be considered research activities (Brabham, 2013; Day et al., 2020; Tucker et al., 2019).

Discussion

PhD students, novice nurse scientists, and even those with more experience can struggle with exactly how to use theory and evidence to design interventions, and with how to effectively engage stakeholders throughout the research process (Majid et al., 2018). Published intervention protocols and reports of trial results typically lack adequate description of the rationale and goals for the intervention. Reports may also lack information about content delivered in the intervention as well as how and why the researchers selected the intervention methods. How intervention components and their content link to elements of the intervention theory and study outcomes may also be missing (di Ruffano et al., 2017; Hoffman et al., 2014; Sidani et al., 2020). Preparation in community-engaged IM is ideally suited to provide nurse scientists with the skills they need to systematically undertake and explicate the work involved in designing, adapting, and implementing theory- and evidence-based interventions to prevent and manage chronic illness.

Potential benefits of preparing nurse scientists in community-engaged IM include that the logic model of change (IM Step 2) illuminates what is expected to happen in the “black box” (Astbury & Leuw, 2010; Lipsey, 1993); that is, the theory-based mechanisms through which the intervention is expected to generate desired changes. Explicating the logical model of change—also referred to the theory of the intervention—has enormous potential to address an identified gap in nursing research vis-à-vis use of theory (Pickler, 2018). Nursing science may be advanced by promoting theoretical fidelity, the alignment between the components and activities that make up an intervention and the constructs that comprise the intervention theory (Ibrahim & Sidani, 2016). Mapping allows researchers to pry the box open to examine mechanisms of change and identify the components essential to change for implementation into practice. In unsuccessful trials, where the logic model change stalled, mapping helps to understand the environmental conditions associated with results as well as identify modifications in the model and the intervention (Ashbury & Leeuw, 2010; Bartholomew & Mullen, 2011). IM acknowledges that intervention creation is complex and time-consuming and offers guidance about how to move work forward by consistently using core processes—posing questions, reviewing literature, considering various theories, and assessing the need for additional data (Eldredge et al., 2016). Moreover, IM provides a robust framework that, in addition to the personal and interpersonal levels, explicitly considers multiple types and levels of environments (cultural, physical, organizational, social, political) that influence health (Fernandez et al., 2019), and in which health care is delivered and nursing practiced. This feature provides means for linking influential environmental factors with constructs from environment-oriented theories at various levels, which can be woven into a series of studies within programs of transdisciplinary intervention research targeting various levels with the overarching goal of improving health.

Preparing nurse scientists in the multilevel perspective of IM reflects beliefs that interventions targeting factors at the multiple levels, which interact in complementary or synergistic ways, will likely produce larger and more durable effects than interventions targeting one level (Weiner et al., 2012). IM can highlight potential intervention targets at multiple levels when mapping a logic model of the health problem of interest and its effects on care and QOL outcomes (IM Step 1). IM also demands theory-based precision in selecting evidence-based methods to address needed changes in influential personal determinants and environmental factors to achieve desired improvements. The strengths that IM brings to intervention research are relevant to each intervention targeting a particular level; IM prompts investigators to be intentional when considering mechanisms underlying potential synergies across levels.

PhD students, novice, and more experienced nurse scientists can also struggle with how to effectively engage stakeholders in intervention design and implementation (Majid et al., 2018). Using CEnR approaches throughout the process offers ways to involve members of the target population and stakeholders representing different levels more effectively. The inclusive and diverse nature of CEnR stakeholder participation facilitates designing multilevel interventions with the pragmatic lens needed to make interventions at each level relevant and feasible for intended beneficiaries and potential adopters (CTSA Consortium, 2011; Forsythe et al., 2018, Jagosh et al., 2015). Relevance and feasibility speed translation—those responsible for using, adopting, implementing, and maintaining interventions will have had input from the outset. Moreover, emphasizing preparation in CEnR can assist nurse scientists who focus on problems for clinical populations in tertiary care settings in developing the background and skills needed to engage patients, caregivers, clinicians, etc., as partners in research processes and decision-making.

We recognize that intervening with individuals in the target population and at multiple environmental levels to improve population health is challenging even for highly experienced nurse scientists. Multilevel interventions are expensive and applications for funding cap allowable direct costs. Furthermore, nurse scientists with research interests related to preventing and managing chronic illness at the person/patient level can have difficulty considering the bigger picture when conceptualizing health problems. As previously mentioned, engaging representatives from stakeholder groups is integral to IM (Fernandez et al., 2019). Nurse scientists with person/patient-level interests can also lack experience with CEnR approaches that public and community health nurses tend to have. They can also conflate stakeholder engagement with collecting data from samples comprised members of the target population.

Conclusion

The goal of preparation in community-engaged IM is not to push nurse scientists to focus on population health. Rather, the goal is to challenge nurse scientists, during their preparation years, to consider the array of environmental conditions that can influence health. To achieve this goal, it is important that nurse scientists learn to develop transdisciplinary teams to design and implement complex multilevel interventions or multiple compatible interventions so they are equiped with the knowledge, perspectives, experiences, and skills needed to improve health.