SOURCES INFORMING DISCOVERY
Three important sources can inform the considerations described previously: systematic literature reviews, prospective studies, and simulation or modeling techniques. Similar to all phases and activities related to behavioral intervention research, it is always essential to begin the construction of an intervention by conducting systematic reviews of existing literature including gray matter, examining the epidemiological record, and identifying meta-analyses and/or systematic reviews. Epidemiological studies or research using large-scale population-based representative samples is particularly helpful for: identifying and quantifying the scope of a problem at both individual- and population-based levels; providing the evidence that the identified problem is a public health concern; and showing the magnitude of the problem. For example, epidemiological studies may be particularly helpful for identifying risk factors for a particular health problem. Either identifying findings from existing population-based studies or conducting secondary analyses with existing population-based data sets can inform how many people are affected, their characteristics, the impact of the problem area, and potential targets for intervention (Ebbeling et al., 2007).
For example, through an examination of the epidemiological literature, both the direct and indirect effects of alcohol misuse have been identified as major contributors to the risk for infection with HIV and transmission of HIV/AIDS. It may be that no level of alcohol consumption is appropriate for those infected by HIV. Thus, developing culturally suitable strategies that provide education to this effect may be a worthwhile intervention (Bernal et al., 2009).
Unfortunately, there continues to be large gaps between the epidemiological record and the targets selected in interventions in many fields (Gitlin, Marx, Stanley, & Hodgson, 2015; McBeth & Cordingley, 2009). An example of this gap can be found with caregiving interventions. Whereas population-based studies suggest that financial strain and physical burdens of care prompt nursing home placement (Spillman & Long, 2009), caregiver interventions to date have not attempted to intervene on these two contributing factors. Nevertheless, an important caveat is this. Although epidemiological research has much to offer, it is unclear whether interventions are indeed more effective if their targets are more aligned with those identified from such studies.
Discovery may involve small pilot testing or dynamically overlap with activities in Phase I (see Chapter 2, Figure 2.2). For example, conducting focus groups, surveys (online, face-to-face, mailings, Facebook), Delphi surveys, or needs assessments, or, as discussed earlier, engaging in secondary data analyses may help to more fully understand a problem area, identify what is meaningful to affected persons, and explore potential approaches to address the issue. Employing quantitative and qualitative methodologies (see Chapter 11 on mixed methods) including ethnographic, narrative approaches, or semistructured or open-ended in-depth interviews can capture the lived experience of individuals and inform the direction by which to develop an intervention (Zatzick et al., 2011).
Finally, applying simulation and modeling techniques to large data sets can be helpful for identifying potential targets for an intervention and examining under hypothetical scenarios associated costs and outcomes. Such approaches can assist in identifying whom or what to target prior to the prospective development of an intervention in order to maximize its impact (Anderson et al., 2014).