4.1 PROCESS FOR INTERVENTION DESIGN

Intervention mapping (Bartholomew et al., 2016) is a systematic, well‐structured process that is useful for designing interventions. Intervention mapping has been recently applied to design health interventions aimed at promoting performance of health behaviors including physical activity (e.g. Direito et al., 2018; Krops et al., 2018), engagement in screening tests (e.g. Besharati et al., 2017; Byrd et al., 2012), and self‐management in chronic illness (e.g. Beck et al., 2019; Burrell et al., 2019). Intervention mapping has also been applied to design public health interventions (e.g. Abbey et al., 2017, Belansky et al., 2013) and implementation interventions that facilitate the adoption of clinical guidelines or evidence‐based interventions in practice (e.g. Caminiti et al., 2017; Chambers et al., 2019), as well as to adapt evidence‐based treatments to the local context (e.g. Koutoukidis et al., 2018; Perry et al., 2017).

Intervention mapping consists of a clearly described process for developing interventions that focus on changing health behaviors (Bartholomew et al., 2016; Kok et al., 2016). The process begins with an analysis of the health behavior to identify what needs to be changed, followed by the creation of matrices that combine change objectives with determinants of the health behavior, selection of theory‐based intervention methods, translation of these methods into practical application, and integration of the practical applications into an organized program. In addition to theory, relevant evidence synthesized from the literature and input from stakeholder groups including clients, are integrated in the process. The application of the mapping process culminates in the design of interventions that are informed by theory and that match or are responsive to the target client population’s experience of the health problem (Beck et al., 2019; Brendryen et al., 2013; Dalager et al., 2019).

The mapping process described next is adapted to enable the design of interventions addressing a range of health problems such as symptoms and cognitions, in addition to behaviors. The application of the steps comprising the process is illustrated with the design of an intervention for the management of insomnia.

Step 1 – Clarify the Health Problem

Clarification of the health problem is done by reviewing the theory or the logic model of the problem (detailed in Chapter 3). It is important to be familiar with the conceptual definition of the problem, as well as its indicators, level of severity, duration, determinants, and consequences, and to understand the direct and indirect relationships among the determinants and the problem experience. Special attention is given to the problem’s indicators and determinants, and the explanations of the pathways linking the determinants to the problem. A lucid understanding of the problem is essential to guide the next steps of the intervention design process. The theory of the problem, exemplified for insomnia, is illustrated in Table 3.1. The theory summarizes the determinants, indicators, and consequences of insomnia.

Step 2 – Analyze the Health Problem

Analysis of the health problem is a foundational step in the process of designing interventions. The analysis consists of critically reviewing the theory of the problem (see Chapter 3) to determine “what about the problem needs to and can be changed” in order to prevent, manage, or resolve the problem. The analysis involves a critical and meticulous examination of: (1) the conceptual definition of the health problem, which highlights the nature of the problem and provides a general hint on its amenability to change; for example, a problem that is genetic in nature may be difficult, if not impossible, to change whereas unhealthy behaviors or cognitions are potentially modifiable; (2) the operational definition of the problem into attributes that are specified in respective indicators, which point to indicators that are potentially changeable; and (3) the determinants of the problem, which identify those potentially modifiable. The meticulous examination contributes to a judgment as to what can be actually changed (also referred to as aspects of the problem): the overall problem, some or all its indicators, or some or all its determinants (Araújo‐Soares et al., 2018; Bello & Pillay, 2019; Besharati et al., 2017; Bleijenberg et al., 2018; Wight et al., 2016).

The judgment is based on logical thinking relative to the amenability of the problem, its indicators and determinants to change. The judgment is also informed and endorsed by the propositions of the middle range theory underpinning the conceptualization of the problem (selected if the theoretical approach is used to gain an understanding of the problem—see Chapter 3), the empirical evidence integrated to support the experience of the problem and its indicators and the association with its determinants (as is done if the empirical approach is used), and/or the explanations provided by stakeholder groups (as is done if the experiential approach is used).

The middle range theory of the problem provides statements about the conceptualization of the health problem and its associations with determinants. The theory also points to specific aspects of the problem that are malleable and have the greatest scope of change (Bleijenberg et al., 2018; Wight et al., 2016). These changeable aspects become the target of the intervention; that is, the intervention is designed to manage these aspects, with the ultimate goal of successfully addressing the health problem. For example, the social cognitive theory is frequently used to inform the conceptualization of health behaviors and the design of behavioral interventions (e.g. Durks et al., 2017; Lamort‐Bouché et al., 2018). The social cognitive theory posits that (non)engagement in a behavior is influenced by: personal determinants such as cognitions (beliefs, attitudes, expected outcomes of the behavior); perceived behavioral control (self‐efficacy); and social determinants such as norms and peer influence. The theory highlights cognitions and perceived behavioral control as determinants most malleable to change (e.g. Ball et al., 2017; Dalager et al., 2019; Direito et al., 2018).

The empirical evidence synthesizes the results of quantitative and/or qualitative studies pertaining to the experience if the health problem and its determinants. The evidence indicates aspects of the problem, in particular determinants, that could be potentially targeted by the intervention. Specifically, relevant evidence shows that the determinants (1) are consistently (across studies) and significantly associated with the experience of the problem; (2) are prioritized or considered important in contributing to the problem by the target client population; and (3) change over time, either normally or following treatment. Evidence of change confirms that the determinants are potentially modifiable. For example, incorrect beliefs and expectations about sleep have been found to perpetuate insomnia, and to be modified as a result of cognitive therapy (e.g. Eidelman et al., 2016; Morin et al., 2007).

In the absence of a middle range theory and empirical evidence on the health problem, the judgment is formed on the basis of systematic analysis of the problem and logical thinking; both are done either by the researchers alone or in collaboration with experts, including health professionals and clients. For example, the theory of insomnia presented in Chapter 3 is used here to illustrate the application of this analysis. As mentioned in Table, insomnia is conceptualized as a learned behavior. It is manifested as difficulty initiating and/or maintaining sleep (indicators), and influenced by predisposing, precipitating, and perpetuating factors (determinants). The analysis begins by questioning the extent to which insomnia, as a learned behavior, can be altered directly. Logically, this may be possible but not easy due to the complexity of the behavior. The conceptualization of insomnia as a learned behavior suggests that it can be “unlearned” and substituted with other behaviors that promote sleep. This perspective points to the need for behaviorally based interventions to manage this problem, but it does not indicate the specific behaviors to be changed. The analysis then moves to other aspects of insomnia to determine their amenability to change.

The analysis involves a review of the indicators and determinants of insomnia. There is no theoretical, empirical, or clinical/practical proposition that suggests that the indicators of insomnia (e.g. difficulty falling and/or staying asleep) can be directly manipulated or changed. However, a review and critical analysis of the three categories of determinants points to the following logic: (1) predisposing factors are innate characteristics of persons with insomnia and accordingly, they are not modifiable; (2) precipitating factors are often out the persons’ control because they initiate or trigger poor sleep but dissipate once the persons start to experience insomnia; therefore they may not be changed at the time the persons start to experience insomnia and seek treatment; (3) perpetuating factors, representing behaviors and use of strategies or techniques that maintain insomnia are potentially modifiable—these behaviors can be unlearned and the strategies can be substituted with sleep‐promoting ones.

The analysis results in the identification of what about the problem or aspects are modifiable. The identified aspects are defined at the conceptual and operational levels, based on the information presented in the theory of the problem, or relevant theoretical and empirical literature. These definitions are useful in specifying the desired changes in the next step.

Step 3 – Identify Desired Changes

In this step, desired changes are delineated for each aspect of the health problem identified as modifiable (Burrell et al., 2019). The changes represent alterations that should take place in the respective aspects of the problem and, subsequently, contribute to the prevention, management, or resolution of the problem. The changes are expected to occur at the level (e.g. intraindividual, environmental) at which the respective aspects of the problem are experienced. Two types of desired changes are specified. The first type, referred to as “change objectives” in the intervention mapping process described by Bartholomew et al. (2016), defines what clients need to do or alter to induce changes in the respective aspect of the problem (Beck et al., 2019); they reflect behaviors, activities, or actions in which clients engage to modify the aspect of the problem. The second type of desired changes, referred to as “performance objectives” in the intervention mapping process, defines the alterations in the aspects of the problem (Ball et al., 2017) that clients are expected to experience. Thus, desired changes reflect significant milestones, that is, changes in condition that clients experience and in behaviors in which clients engage, in the pathway to prevent, manage, or resolve the health problem (Brendryen et al., 2013; Czajkowski et al., 2015).

The desired changes are stated clearly, concisely, and in observable terms that accurately depict what is the specific alteration to be experienced and what is to be done, and who should make the change. The changes may be hypothesized to take place sequentially, whereby the occurrence of one leads to another, ultimately resulting in the prevention of or improvement in the experience of the health problem. The specification of the desired changes is critical (1) for designing interventions; the changes inform the delineation of the active ingredients (in step 4) and (2) for understanding the mechanism of action (represented in the sequence of desired changes) that explains how the intervention yields improvement in the health problem (Brendryen et al., 2013).

The identification of desired changes is illustrated in the example of insomnia. Of the factors that perpetuate insomnia and that are amenable to change, two health behaviors, physical inactivity and smoking, are reported to influence sleep. The following sequence of changes in condition is desired to help clients avoid these behaviors:

1. Heightened awareness of the general health‐related behaviors that affect sleep.
   
2. Improved understanding of when, how, and why these behaviors interfere with sleep and contribute to insomnia
   
3. Increased knowledge of recommended techniques for handling these behaviors and, hence, mitigating their interference with sleep.
   
4. Enhanced self‐confidence or self‐efficacy in applying the recommended techniques.
   
5. Enhanced ability/skills in applying the recommended techniques.
   
6. Appropriate and consistent use of the recommended techniques.

Step 4 – Delineate Intervention’s Active Ingredients

The active ingredients of a health intervention are the specific therapies or techniques (called “intervention methods” in the intervention mapping process) that are hypothesized to bring about the desired changes and, consequently, induce improvement in the health problem (Bleijenberg et al., 2018; Wight et al., 2016). The specific techniques encompass information that is relayed to clients and behaviors that clients engage in to address the problem. The techniques are delineated for each aspect of the problem identified as malleable, and relative to the respective desired changes. The techniques should be consistent with the aspect of the problem and capable of inducing the desired changes. The conceptual correspondence or match among the aspect of the problem, active ingredient or technique, and desired changes is essential to ensure the design of interventions with great potentials for being effective in addressing the health problem in the client population and context of interest (Dohnke et al., 2018; Mesters, et al., 2018).

The active ingredients or techniques can be identified from relevant theory, relevant empirical evidence, and/or consultation with experts or stakeholder groups, as explained in Section 4.2.

Delineation of the intervention’s active ingredients is founded on logical reasoning. Logical reasoning is based on a thorough understanding of the nature of the aspect of the problem amenable to change and critical thinking of how it can be modified. The goal is to generate new or select available therapies or techniques that conceptually correspond with the nature of the problem or its aspects. This implies that the techniques should address the problem or its relevant aspects directly, effectively, and efficiently by triggering the respective desired changes.

In the example of insomnia, engagement in health behaviors (i.e. physical inactivity and smoking) perpetuates this sleep problem. To induce the desired changes identified for the behaviors in the step 3, it is logical to provide education (active ingredient 1) about the behaviors and how they interfere with sleep, and recommend techniques for handling them to produce desired changes 1–3, as well as to offer instrumental support (active ingredient 2) in applying the recommended techniques to induce desired changes 4–6.

Step 5 – Operationalize the Active Ingredients

In step 4, the intervention’s active ingredients are delineated at the conceptual level; as such they are broadly defined (e.g. provide education about factors contributing to the health problem and offer support in changing these factors). While important, conceptually delineated active ingredients may not give specific instructions for how to deliver them; therefore, they should be diligently operationalized in a manner that maintains correspondence between their conceptual definition and delivery. This correspondence is necessary to enhance construct validity of the intervention (Sidani, 2015).

Operationalization of the intervention’s active ingredients consists of:

1. Specifying the components that represent the active ingredients: This is accomplished by operationally defining each active ingredient in a component. The operationalization of each component involves detailing the content or information to be relayed to clients, the specific behaviors or activities in which clients engage, and the recommendations that clients are to apply in their daily life in order to attain the desired changes specified relative to the aspects of the problem amenable to change.
   

   In the example of insomnia, education and support were delineated as active ingredients for addressing the health behaviors that perpetuate insomnia. To operationalize these ingredients, one should ask the questions: education about what in particular and what type of support is useful for changing these two determinants of insomnia? The answers should be very specific detailing:

   
   
   
   
   • The content of education: (1) inform clients of the health behaviors that perpetuate insomnia (what are the behaviors?)—for example, physical inactivity; (2) explain how the behaviors interfere with sleep (what is the pathway through which these behaviors lead to poor sleep and insomnia)—for example, physical inactivity influences circadian rhythm “synchronizers,” daytime physical and mental health, arousal and body temperature around bedtime, and sleep architecture (Chennaoui et al., 2015; Irish et al., 2015); (3) describe, in detail, the recommended techniques for handling the behaviors—for example, develop a regular schedule of physical activity during the day, engage in physical activity in the later afternoon or early evening to help ward off feeling of early sleepiness or drowsiness, and avoid rigorous exercise immediately before bedtime.
     
   • The type of support: (1) engage clients in an active discussion to select the physical activity they enjoy and afford performing, and to think through a plan or procedures to put in place to regularly perform the selected physical activity; (2) encourage clients to engage in the planned physical activity in daily life and monitor its impact on sleep; (3) review barriers and enablers, and generate or exchange ideas of what can be done to overcome barriers and reinforce enablers.
   
   
2. Selecting the mode for delivering the active ingredients: This consists of specifying the manner in which the active ingredients, as operationalized into respective components, are to be given to the client population in the context of interest. This exercise is referred to as practical applications in the intervention mapping process. Practical applications are concerned with the way the techniques are translated for practical use or delivery to fit the context of the client target population (Dohnke et al., 2018; Kok et al., 2016; Mesters et al., 2018). As described in Chapter 2, mode of delivery is characterized by different media and formats, some of which may be more appropriate for the operationalization of particular active ingredients than others. Therefore, the selection of delivery modes is based on an understanding of the active ingredients and of the range of media and formats.
   

   The mode of delivery is carefully selected to: (1) be consistent with the nature of the active ingredient, as operationalized in specific content, activities, and treatment recommendations; and (2) facilitate the delivery of the active ingredients in a way that is efficient yet maintains integrity of the active ingredient. Selection of the delivery mode is informed, where available, by evidence of the effectiveness of different media and formats, in different contexts, as well as evidence of their acceptability to the client population of interest and feasibility of use in the context of interest.

   
   

   In the example of insomnia, education, as an active ingredient, can be delivered in the written medium and in the format of an online module that covers the information on the health behaviors, the pathway through which they interfere with sleep, and the treatment recommendations. This mode of delivery (online module) is considered efficient, reaching a large proportion of adults with insomnia, with minimal human (e.g. therapist time) and material (e.g. costs of printing) resource expenses, and burden on clients who can assess and review the module at their convenience (i.e. reduced burden associated with time and cost of transportation). However, the online mode of delivery may not be accessible to some clients such as those with low reading and computer skills, vision problems, limited understanding of the language in which the module is written, and those who do not have access to a computer at home. Further, it may not be helpful to persons with low motivation and/or low self‐confidence in the ability to carry out the treatment recommendations (Beall et al., 2014; Free et al., 2013; Murray, 2012). Lastly extant evidence, synthesized in systematic reviews, indicates that the effectiveness of online or web‐based modes for delivering interventions in achieving the desired changes in a range of health problems is rather limited. Specifically, van Straten et al. (2018) found that the effects of self‐help, web‐based modes of providing cognitive behavioral therapy on insomnia severity and other sleep indicators, were smaller than those of person‐dependent modes of delivery like group or individual sessions.

   
   
3. Determining the optimal structure and intervention dose: Once the active ingredients are operationalized into components and the most appropriate and effective mode of delivery is selected, the optimal structure and dose should be determined for delivering the intervention, comprised of all ingredients. The structure has to do with the approach and sequence for providing the intervention. The decision to use a standardized or tailored approach for giving the intervention is made based on a number of factors discussed in Section 4.5 related to the design of tailored interventions. The nature, complexity (i.e. type and number of components), sequence, and mode of delivery guide the specification of the optimal dose. If available, empirical evidence obtained in studies that compared the effectiveness of different dose levels for the same intervention is used to further justify the specified dose.

In the example of insomnia, two active ingredients, education and support, are delineated. It is logical to estimate that (1) education can be offered in one, 90‐minute session; this session length provides ample time for the interventionist or therapist to present the detailed information at a pace suitable to clients who may vary in their ability or speed of grasping and processing the information, and for the clients to ask for clarification as needed; (2) support can be offered in four, 60‐minute sessions; having four sessions provides clients the opportunity to apply the treatment recommendations in their daily life, and to reflect on their impact on sleep in the time interval between sessions, as well as to discuss ways or strategies to overcome barriers and reinforce enablers in subsequent sessions; thus, the plan is for four sessions (one for education and three for support). The selection of this dose is supported by empirical evidence showing that behavioral therapy delivered in four sessions is effective in managing insomnia (van Straten et al., 2018).

Step 6 – Production of the Intervention Protocol and Materials

After delineating all intervention components that operationalize the active ingredients, mode of delivery, and optimal structure and dose, it is important to develop the intervention protocol and produce relevant intervention materials (Ball et al., 2017; Byrd et al., 2012). The intervention protocol organizes the sequence for delivering the components, and within each component, the order for relaying the content and for engaging clients in the planned activities; it also describes the specific procedures to be followed, and the human and material resources required for delivering the intervention (see Chapter 7). The intervention materials refer to those used by the therapist in delivering the intervention (e.g. slide presentation) or provided to clients as a reference for application of the treatment recommendations in daily life (e.g. booklet).