Desktop version

Home arrow Engineering arrow Behavioral Intervention Research: Designing, Evaluating, and Implementing



Currently, thousands of behavioral interventions have been developed, evaluated, and found to be effective for a very wide range of populations, purposes, public health issues, and outcomes. It is impossible to summarize this vast body of promising and proven behavioral i nterventions. Nevertheless, considered collectively, common characteristics can be discerned of effective interventions that are designed to address behavior change and complex health and social issues. These are shown in Table 1.1, although this list should not be construed as exhaustive. Effective approaches may differ by the specific purpose or intent of an intervention, its mode, and context of delivery.

However, one apparent shared characteristic of effective interventions for behavioral change is that most tend to involve multiple components each of which targets a different aspect of a presenting problem and pathway for effecting positive change. This is not surprising as most issues targeted by behavioral interventions are complex and multidimensional. For example, an intervention designed to prevent and treat delirium in hospital settings targets factors intrinsic to the person such as medication profile and pain, and extrinsic factors such as staff training and the physical environment (e.g., noise, lighting, cues for orientation) (Inouye et al., 1999). The Get Busy Get Better program to help African Americans address depressive symptoms seeks to improve mood by impacting various potential contributory factors including a person’s anxiety, knowledge of depression, and ability to detect his/her symptoms; reducing stressors in the external environment including financial strain and unmet social, housing, and medical needs; and by helping people re-engage in activities that are meaningful to them (Gitlin, Roth, & Huang, 2014).

Effective interventions also appear to tailor or customize content and strategies to key risks, needs, or specific profiles of target populations and contexts (Richards et al., 2007). For example, the REACH II intervention for families caring for persons with dementia modified the intensity (time spent) and amount of exposure (dose) to each of its five treatment components based upon a caregiver’s initial (baseline) risk profile (Belle et al., 2006; Czaja et al., 2009). More time was spent on one component versus the other, depending on the risk profile of the individual caregiver; a caregiver with a home safety risk received greater attention in this area than a caregiver without this risk, although both received a minimal dose of this treatment component. The Get Busy Get Better program for depressive symptoms included five treatment components (care management, referral and linkages, stress reduction, depression education and symptom recognition, and behavioral activation). Although all participants received all five treatment components at equivalent dosage and intensity, the content covered in each component was tailored to the participant’s specific care needs; the person’s preferred stress reduction techniques; housing, financial, and unmet medical needs; and self-identified preferred activity and behavioral goals (Gitlin et al., 2013).

Another shared element of many effective interventions is their flexible delivery schedule. Interventions that do not have rigid dosing requirements have a greater likelihood of being adopted by, and integrated into, clinical settings and by end users

TABLE 1.1 Characteristics of Effective and Ineffective Intervention Approaches

Effective Approaches

Ineffective Approaches

  • ? Intervention and its characteristics are grounded in theory
  • ? Multicomponent such that different strategies are used to address distinct factors contributing to the identified problem area
  • ? Multimodal such that different pathways (e.g., physical exercise, cognitive stimulation) are targeted to impact the identified problem area
  • ? Strategies are tailored to participant needs, characteristics, cultural preferences
  • ? Participant-centered in that it integrates the client perspective
  • ? Participant-directed in that intervention addresses self-identified needs
  • ? Use of active engagement of participants and/or problem solving
  • ? Flexible delivery characteristics to accommodate differences in practice settings
  • ? Outcomes are closely aligned with and reflect intervention intent
  • ? Oriented toward building skills and problem solving to bring about behavior change
  • ? Criteria for participant inclusion reflect intent of intervention
  • ? Involving end users (participants) and/or stakeholders in the development of the intervention
  • ? No theoretical basis for the design of the intervention
  • ? Focus on a singular aspect of a complex set of factors contributing to a particular problem area
  • ? One pathway is targeted although multiple factors contribute to the identified problem area
  • ? Use of a "one size fits all" approach
  • ? Prescriptive, didactic, standard approach regardless of participant perspective
  • ? Participant needs are assumed a priori
  • ? Use of didactic, prescriptive approach
  • ? Fixed dose and intensity
  • ? Outcomes are too distal from content or focus of intervention
  • ? Providing education to enhance knowledge when goal is to change behavior
  • ? Mismatch between intervention intent and participant inclusion criteria
  • ? Not considering the participant or stakeholder perspectives early on in designing an intervention

(e.g., interventionists and participants who may benefit). For example, the Adult Day Plus intervention provides care management, education, support, and skill building on an “as needed” basis for family caregivers who use adult day services for a relative for whom they provide care. Sessions initially occur biweekly for 3 months and at the time when a family member drops off or picks up their relative at the adult day service. Following this initial phase, ongoing contact is periodic and can be initiated by the service provider or family member (Reever, Mathieu, Dennis, & Gitlin, 2004).

In addition, repeated exposures to an intervention appear to yield better outcomes such as reducing nursing home placement or maintaining independence at home. For example, the Maximizing Independence at Home for persons with cognitive impairment provides ongoing care management for 18 months; pilot data with over 300 persons showed that this approach resulted in the reduction of some home safety risks and more days at home (versus nursing home placement or death) (Samus et al., 2014).

Finally, interventions that actively involve participants in the treatment process and the learning of new skills may be more effective than prescriptive, didactic approaches (Belle et al., 2003) when the intent is to change behavior and redesign lifestyles for healthier living. Self-paced programs, approaches in which participants have opportunity to practice and integrate behavioral change strategies, afford more positive outcomes than approaches that do not provide such opportunities. Similarly, if the goal is to improve self-management, certain strategies appear to be more effective than others. For example, using behavioral activation techniques (Hopko, Lejuez, Ruggiero, & Eifert, 2003) that involve individuals self-selecting personal life, health, or daily goals, or providing control-oriented strategies to help people achieve their daily activity goals, afford increased control over daily life events and result in better health outcomes (Heckhausen, Wrosch, & Schulz, 2010). Correspondingly, having participants codesign their own action plans for achieving healthier lifestyles (Lorig et al., 1999) are all strategies rooted in complementary theoretical frameworks that result in enhanced self-efficacy and health-related benefits. This is not to say that behavioral interventions must embrace each or all of these treatment elements to be effective. However, it does suggest that consideration be given to these characteristics in order to maximize the impact on certain types of behavioral change outcomes (Zarit & Femia, 2008). Each of these characteristics is rooted in various theories, best practices such as adult learning principles, and research evidence concerning what works and what does not in changing behavior and personal health practices.

We also have fairly good knowledge of what is ineffective when the goal is to change behavior and health care practices through an intervention. Although an implicit goal of an intervention is to have the biggest impact on the largest number of persons as possible, given the heterogeneity and diversity of populations, a “one size” approach typically does not work. For example, the REACH II intervention, overall, was more effective for Hispanic and White/Caucasian caregivers than for African Americans. However, further analyses showed that, within the African American sample, it was more effective for caregivers who were spouses and older. Devising ways of introducing choice and tailoring an intervention to preferences or situations is, in general, preferred (Belle et al., 2006).

Using prescriptive approaches or providing education alone when the goal is behavioral change has also been shown to be mostly ineffective. Fixed dosing requirements may be important, but this also limits the translation and implementation potential as clinical settings and other end users of an intervention may need greater flexibility in the delivery of such a program. Finally, developing interventions without fully understanding the context in which they will be implemented (see discussion below and Figure 1.2) limits its ultimate usability and acceptability. Involving immediate end users and stakeholders (e.g., interventionists, administrators, payors, participants themselves) early on in the intervention development process is emerging as a best practice. This systems-oriented approach integrates a usability testing and iterative process for developing interventions from the start, to optimize the fit between the intervention and the context in which it is designed for implementation if it is proven to be effective (see Chapter 2).

Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >

Related topics