Implementation Practice Models for Development in Low- and Middle-Income Countries: Systematic Review of Peer-Reviewed Literature

"[P]ractitioners, especially in development contexts in LMICs, often face resource and other constraints and must prioritize program implementation. At the same time, design of effective programs requires use of theory. Thus, implementation practice models offer a practical approach to the use of theory in program design in development settings."

Implementation practice models are theoretical models that take a practical and practitioner-focused approach to behaviour change (e.g., changes in health promoting and risk behaviours in a priority population). This paper describes a systematic review of peer-reviewed literature on the development, application, and evaluation of implementation practice models for health behaviour change, with a focus on development programmes and research (e.g., efforts to empower people to improve their well-being and address the causes and effects of poverty) in low- and middle-income countries (LMICs).

First, the paper reviews the concept of a continuum of behavioural theory - from theory design and testing to application of implementation practice models - in development programmes. At the theory design end of the continuum, prominent frameworks include the health belief model, transtheoretical model, social cognitive theory, and social-ecological model, all of which involve multi-dimensional constructs such as perceived vulnerability, social norms, self-efficacy, response efficacy, decisional balancing, and context-specific circumstances that can promote or hinder desired behavioural change. However, in the context of implementation, particularly within development programmes in LMICs, theoretical complexity becomes a barrier. Practitioners need theories that are pragmatic and can be applied despite resource constraints and other implementation barriers. Implementation practice models, as described in this paper, attempt to demystify theory and isolate essential variables that can be addressed in a development context.

There are a number of widely known examples of implementation practice models, which this paper reviews. Examples include the Fogg Behavioral Model (FBM), COM-B (capability, opportunity, and motivation), and EAST (easy, attractive, social, and timely). One common characteristic of these models is their attention not only to individual characteristics (e.g., attitudes, beliefs, and other personal factors), but also to the intersecting environmental factors that influence behaviour.

To understand the extent and nature of evidence on implementation practice models, the researchers conducted a systematic search of published, peer-reviewed articles in the English-language health, social science, and business literature that apply implementation practice models, that are located in an LMIC, and that have a behaviour change objective. They located 1,078 articles and ultimately included 25 of them.

• Table 1 in the paper provides a summary of basic information gleaned from each implementation model in the articles reviewed. The articles deal with interventions relating to a wide variety of health issues and behaviours, including maternal and child health, sexual health, family planning, and nutrition.
• As shown in Table 2, the interventions used a wide range of intervention approaches and strategies, including mass media (radio, TV), interpersonal communication (IPC) through community outreach, and visits to households by health workers. High levels of awareness of the promoted health messages were reported.
• Table 3 provides a summary of the study design and outcomes in the articles reviewed. Nine of the studies reported on intervention awareness/reactions as a measured outcome, and 5/9 report positive statistically significant effects on that outcome. Of the studies that measured pre-behavioural outcomes, such as attitudes, beliefs, and social norms, all showed a positive statistically significant effect on those outcomes. Finally, 23 studies reported on behaviour change as a measured outcome, and all showed a positive statistically significant effect on the targeted behaviour(s). However, each of these studies used self-report measures of behaviour.
• The models found in this review, such as PERForM, Fogg Behavior Model, COM-B, Behavior Change Wheel, and SATURATION +, share common characteristics of focusing on the three tenets of OAM: opportunity (i.e., situational conditions), ability (i.e., task knowledge), and motivation (i.e., attitudes, beliefs, norms). This shared focus on OAM represents the simplification of behavioural theory and, as the researchers argue, has the potential to maximise its utility in application.

According to the researchers, this review has implications for both research and programming in LMICs. For example, while the peer-reviewed literature shows that implementation practice models are generally effective in promoting behaviour change, there are relatively few rigorously controlled studies. Future research could also examine whether implementation practice models are effective when dealing with multidimensional behaviours requiring potentially complex decision making, as in the case of vaccine hesitancy. Furthermore, the researchers call for increased focus on valid and reliable measurement of OAM variables and development of standardised metrics in the field.

In conclusion, "implementation practice models for development are a promising and growing approach to behavior change in LMICs. Intervention practice models research should be expanded and applied in new domains, such as vaccination." Increasing the use of implementation practice models in interventions in LMICs has the potential to "improve quality, increase the rigor of evaluations, and thus improve the evidence base on the effectiveness of such programs over time."