The best magazine
What to Address in Adherence Interventions?
In considering strategies to promote ART adherence, it is helpful to first identify the targets of a given intervention (what will be changed or supported in order to improve or sustain high rates of adherence) and then consider how to impact those targeted factors (specific intervention strategies or approaches). Obviously, the ultimate target is adherence and its persistence over time. Short of direct delivery and observation of each dose or the development of some delivery mechanism that bypasses self-directed dosing behaviors, interventions must work with the factors that lead to good adherence or the mitigation of factors that deter adherence. Thus, identifying what to target in an intervention precedes consideration of how to actually do this, and the extent to which targeted factor/factors actually relate to adherence in many ways sets the upper limit for ultimate efficacy of an intervention approach.
There are a number of discrete factors that have been associated with ART adherence. Across populations, these include accessibility of experience of side effects, availability of resources (including food, finances, and transport to HIV care), knowledge/literacy concerning the condition and its treatment, stability of daily schedule/routine, social support and stigma, positive and negative beliefs about medications, coping, depression, competing priorities and substance use, medication management skills, executive memory and cognitive functioning, simply "forgetting," and consistent source for medications and use of them. Arguably, these individual level correlates are contextualized by the social–cultural context of living with HIV and community-level beliefs more generally about medical establishments treating HIV. Further, with so many discrete correlates of adherence identified, higher order organization of these correlates into models of health behavior offers multiple benefits, such as allowing for more targeted intervention approaches and appropriately contextualizing individual behavior in the context of larger social, cultural, and structural systems.
Models of adherence that combine discrete correlates and situate individual behavior in larger contexts have been proposed and offer some guidance toward identifying common "drivers" of adherence. A basic taxonomy suggested in the late 1990s continues to provide a useful heuristic for considering determinants of adherence: aspects of the patient, aspects of the provider and patient–provider relationship, aspects of the regimen and illness, and the contextual (environmental, structural, community) factors that influence adherence. Health behavior models typically draw from across these general areas to identify core factors that while housed within the individual are the net result of previous and ongoing interactions with others (important others, HIV care providing staff, communities) and with systems of care. While no single behavioral model has been identified as effective in predicting adherence (see for example the study by Amico et al), when considering intervention strategies the articulation of an underlying model is essential for identifying what to target (what aspects of the person, social interactions, system of care, or regimen should be addressed by the intervention) and why it would be expected to positively impact adherence. Across behavioral models, common critical determinants of adherence include knowledge, attitudes, beliefs, perceptions of benefits and costs, social support, past and current experiences with care providers, real and perceived burdens to adherence and several executive functions like developing skills, negotiation of behavior in common situations, coping, and regulation of perceived threats to health. Determinants of adherence likely have commonalities between cultures and geographies but also unique factors emerging from unique culture and community systems (see for example the work of Simoni et al).
Behavioral models typically attempt to explain the behavior, whereas "how" to influence change on any of the identified core determinants of adherence requires additional consideration. Numerous models identifying strategies to promote change on the individual, community, and structural level are available (eg, motivational interviewing, cognitive behavioral therapy strategies, peer delivery models, community health workers (CHWs), or technology). In order to identify effective and not-effective strategies for promoting adherence, both the explanatory and the intervention delivery models adopted need to be identified. Failed intervention approaches may be ineffective because they did not change the targeted factor (eg, participants or patients did not develop new skills) or the intervention was effective in promoting change in the targeted factor, but the factor itself was not sufficiently or uniquely related to adherence for a given population (eg, new skills were developed but the development of new skills was not related to improved adherence). Additionally, the needs of a given population (eg, are rates of adherence currently suboptimal, are the barriers targeted by an intervention actual barriers in the study sample?) or lack thereof effects outcomes and must be carefully considered. Tremendous resources are allocated to rigorous research on approaches to adherence support. Whereas much of the literature currently available determines effects of a given intervention approach on the basis of intent to treat and additionally as treated, the identification of what the intervention changed or failed to change within what kind of cohort and whether success, or failure, can be attributable to one or more of these factors is less often included (see for example the study by de Bruin et al). However, we would argue that this level of granularity is essential to promoting systematic progress toward identifying strategies that are effective in promoting high and consistent adherence. At minimum, research should clearly articulate the targets of their interventions and present data relevant to the success of the intervention in influencing those targets; strategies adopted in practice should similarly identify what a strategy or approach addresses, why it would be expected to influence adherence in a given population, and monitor for success in influencing the targeted factors.
The process of identifying what explanatory models appear to be more reflective of adherence in a certain patient population and which delivery model would offer greatest benefit can involve substantial adaptation. Intervention mapping offers a systematic process for intervention development that can provide guidance for this process with the advantages of producing clear delineation of what will be targeted for change or support and how. Adapting models developed to explain adherence predominantly for one population (eg, PLWH in the United States) to a different population (eg, PLWH in Uganda) can be guided by Ware et al's suggested process where 4 basic questions provide an evaluation of validity for a model: (1) Are the model concepts relevant to the new setting? (2) Are the concepts relevant in the new setting well represented in the model? (3) Do the definitions of the concepts in the model have accuracy in the new setting? and (4) Is the model comprehensive enough to explain the full complexity of adherence in the new setting? Intervention mapping and vetting model validity are both exercises that can benefit research and practice through re-focusing intervention efforts on specific populations and attending to the social–cultural context in which adherence is negotiated. The need for better articulation of the phenomenon of interest has been identified for some time (see for example the article by Bain-Brickley et al), but there continues to be a general lack of targeted efforts to evaluate the general strength and accuracy of the behavioral models that underlie intervention efforts.
Source: ...