Elsevier

Seminars in Perinatology

Volume 39, Issue 5, August 2015, Pages 393-415
Seminars in Perinatology

Enculturating science: Community-centric design of behavior change interactions for accelerating health impact

https://doi.org/10.1053/j.semperi.2015.06.010Get rights and content

Abstract

Despite significant advancements in the scientific evidence base of interventions to improve newborn survival, we have not yet been able to “bend the curve” to markedly accelerate global rates of reduction in newborn mortality. The ever-widening gap between discovery of scientific best practices and their mass adoption by families (the evidence-practice gap) is not just a matter of improving the coverage of health worker-community interactions. The design of the interactions themselves must be guided by sound behavioral science approaches such that they lead to mass adoption and impact at a large scale. The main barrier to the application of scientific approaches to behavior change is our inability to “unbox” the “black box” of family health behaviors in community settings. The authors argue that these are not black boxes, but in fact thoughtfully designed community systems that have been designed and upheld, and have evolved over many years keeping in mind a certain worldview and a common social purpose. An empathetic understanding of these community systems allows us to deconstruct the causal pathways of existing behaviors, and re-engineer them to achieve desired outcomes. One of the key reasons for the failure of interactions to translate into behavior change is our failure to recognize that the content, context, and process of interactions need to be designed keeping in mind an organized community system with a very different worldview and beliefs. In order to improve the adoption of scientific best practices by communities, we need to adapt them to their culture by leveraging existing beliefs, practices, people, context, and skills. The authors present a systems approach for community-centric design of interactions, highlighting key principles for achieving intrinsically motivated, sustained change in social norms and family health behaviors, elucidated with progressive theories from systems thinking, management sciences, cross-cultural psychology, learning and social cognition, and the behavioral sciences. These are illustrated through a case study of designing effective interactions in Shivgarh, India, that led to rapid and substantial changes in newborn health behaviors and reduction in NMR by half over a span of 16 months.

Introduction

Major scientific advances have been made in our understanding of neonatal mortality in low resource settings over the last two decades. A host of proven interventions have been identified, many of which have substantial efficacy for reducing neonatal mortality rate (NMR). When delivered together at high coverage, these interventions can potentially avert up to an estimated 71% of newborn deaths.1 Yet, observed reductions in NMR globally have only averaged about 2–3% annually.2 A decade since the 2005 Lancet Neonatal Survival series, 2.9 million newborn deaths—most of them preventable—still continue to occur each year. It is obvious that there are gaps in our current approach to reducing neonatal mortality that are leading to a significantly lower impact than that is possible. Until we recognize and address these gaps, we will continue to miss global targets, and more importantly, millions of babies will continue to die needlessly. It is therefore a moral and ethical imperative that we pause to critically analyze reasons for our suboptimal impact and take corrective actions.

There has been an ever-widening gulf between the discovery of best scientific practices for improving newborn health and survival, and their translation into what families actually practice at home. Even the most basic technology-agnostic life-saving behaviors such as immediate and exclusive breastfeeding have met with limited success. A focus on discovery of life-saving interventions alone is therefore insufficient. Fundamental breakthroughs are needed in methodological approaches to ensure mass adoption of these discoveries.

In recent years there has been an increasing emphasis and growing body of work on applying implementation science methods within the ambit of health systems research to address the evidence-implementation gap—the gap between availability of evidence-based interventions and their effective delivery to large populations to achieve health impact.3, 4 However, it is not the interventions per se that are delivered to communities. Rather, it is a set of interactions between health workers, and families and communities that must translate into the intervention being provided to mothers and their newborn babies. For example, while “immediate and exclusive breastfeeding” is an evidence-based intervention from an epidemiological perspective, it is in effect the desired behavioral outcome of a set of interactions between health workers and mothers. While implementation research can guide the process of ensuring more equitable and reliable reach of such health worker-community interactions, the content and process of the interactions themselves must be designed based on sound behavioral science approaches in order to achieve effective coverage and impact.

When compared to the rigorous methods of epidemiology, behavior change comes across as a soft science with vague methods and lacking in systematic application to achieve desired outcomes, particularly at scale. Not surprisingly, while there is consensus among epidemiologists around the interventions and coverage rates needed to achieve desired reductions in mortality,1 there is a gap in understanding how these interventions should be delivered to communities to achieve desired behavioral outcomes. The set of evidence-based interventions per se is accessible to country-level policy makers through published systematic reviews and global technical recommendations, and many of these interventions make their way into national guidelines. However, the translation of these interventions ultimately into the design of interactions between health workers and families is left to program implementers. Without the knowledge of scientific methods and guiding principles for designing effective interactions, program implementers tend to adopt the same traditional methods that have failed time and again in achieving desired behavioral outcomes. The net results of low annual rates of mortality reduction, therefore, are hardly surprising.

There has been an extensive body of research on behavior change over the last several decades, resulting in theories and models that help explain behavior as well as suggest ways to influence and change behavior.5 However, most existing theories, such as the social cognitive theory6 and the trans-theoretical model,7 are aimed at modifying predominantly habit-forming behaviors in a largely western cultural context. Consequently, these theories are biased toward western perspectives of communication that involve rationality-driven explicit messages, as compared to traditional cultural preferences in communities towards intuition, experience, and trust.8 The resulting programmatic approaches stemming from these theories are therefore not fully compatible with the cultural context of the communities for which they are developed, thus limiting their effectiveness.

Fundamentally, newborn care practices in traditional societies are deeply entrenched within a culture transmitted over generations, and are manifestations of a worldview predominantly shaped by intuitive assumptions of cause and effect based on observations and experiences in the absence of access to modern sources of knowledge (Table 1, Table 2). To the educated mind, many of these practices appear bizarre, irrational, and incomprehensible at the surface. They are labeled as a “black box” that can neither be understood, nor is there a perceived need to further understand it. This leads to behavior change approaches that are essentially “blind” to existing traditional practices and their underlying beliefs. These include information-centric approaches that provide unidirectional communication regarding desired behaviors (“do’s and don’ts”), sometimes complemented with rationale presupposing a reasoning framework that resonates with the interaction designers but not necessarily the users (Table 1, Table 2). While we do not perceive the need to unpack the community’s “black box” of practices and beliefs, paradoxically, we expect them to understand scientific rationale and adopt scientific practices, not realizing that these could appear as incomprehensible black boxes to them as well. Moreover, the typically one-size-fits-all nature of interactions, while unavoidable in a mass communication format, does injustice to interpersonal and group communication, the dominant communication paradigms in traditional communities. Current approaches to interaction design do not take into account the different characteristics, understanding, and values of users that are critical for effectiveness.9, 10 The persistent failure of such blind approaches in changing community behaviors has led to disillusionment among policy makers and program implementers with the possibility of intrinsically motivated behavior change. A growing impatience for impact has led to resorting to financial incentives as a dominant strategy for changing behaviors which—while potent in some contexts—potentially undermines intrinsic motivation, lowers self-esteem, and can adversely affect non-incentivized behaviors with far-reaching social, economic, and health consequences.11, 12

Previously, we laid out the Behavior Change Management (BCM) framework, a multi-level ecological model for theoretically and empirically sound development of epidemiologically targeted behavior change interventions to reduce neonatal mortality with greater precision, predictability, and pace. It was applied to guide the design of a highly effective behavior change intervention to improve domiciliary newborn care practices in Shivgarh, India13 that led to an NMR reduction of 54% and unanticipated improvements in maternal health over a short span of 16 months.14, 15 Unlike some other community-based trials that included active case management or health facility strengthening,16, 17, 18 the mortality impact of the Shivgarh intervention was solely due to rapid and substantial behavior change across a range of preventive maternal and newborn care behaviors, principally skin-to-skin contact and immediate breastfeeding, catalyzed through efficacious health worker-community interactions.

In order to design effective approaches to behavior change, it is essential to unpack the black box of community behaviors, and create a path of least resistance for families to shift from their existing practices to evidence-based practices. In this article, we explore the nature of traditional family health behaviors and delve deeper into the cultural underpinnings and causal mechanisms that shape these behaviors. This will aid in understanding the reasons for the limited success of usual approaches to behavior change, as well as guide the development of alternative approaches that are more effective. We will interrogate the process of interaction design in Shivgarh as a case study using progressive theories from diverse disciplines, including systems thinking, cross-cultural psychology, learning and social cognition, and the behavioral sciences. We will thus derive some generic principles of community-centric design of interactions that can lead to intrinsically motivated, sustained behavior change.

Section snippets

Interactions lie at the interface of two distinct sociocognitive systems

East is East and West is West, and never the twain shall meet.

—Rudyard Kipling

We hypothesize that one of the key reasons for the failure of health worker-family interactions to lead to behavior change is our failure to recognize that the content, context, and process of interactions need to be designed keeping in mind a very different worldview.

Interactions are typically designed keeping two individuals in mind, for example, a health worker and a mother. However, it must be recognized that the

Community-centric interaction design: The need for a systems approach

Given that interactions occur between two different sociocognitive systems not just individuals, they need to be designed keeping in mind a systems perspective, not just an individual user perspective.

The typical approach to interaction design at a health system level is generalized as follows. Large datasets from demographic surveillance and verbal autopsies provide a reasonably accurate picture of the burden and causes of neonatal deaths at global, regional, and country levels.2 Owing to the

Conclusion and key messages

The unfortunate reality of global child health is that we have been struggling to create adoption of even natural, universally available, low-cost practices, such as breastfeeding and skin-to-skin care. Surely, our usual approaches to health delivery are not delivering results to the extent possible, given, for example, the identification of 59 proven interventions in the most recent Lancet Every Newborn series.1 The authors believe that paradoxically, it is our paradigm of viewing ourselves as

References (32)

  • K. Glanz et al.

    The role of behavioral science theory in development and implementation of public health interventions

    Ann Rev Pub Health

    (2010)
  • A. Bandura
  • J.O. Prochaska et al.

    The transtheoretical model of health behavior change

    Am J Health Promot

    (1997)
  • Y. Miike

    An Asiacentric reflection on Eurocentric bias in communication theory

    Commun Monogr

    (2007)
  • D.L. Medin et al.

    The cultural side of science communication

    Proc Natl Acad Sci U S A

    (2014)
  • D. Kahan

    Fixing the communications failure

    Nature

    (2010)
  • Cited by (32)

    • Investing in the foundation of sustainable development: pathways to scale up for early childhood development

      2017, The Lancet
      Citation Excerpt :

      Services need to be adapted to local context, address existing beliefs and practices, and be delivered through channels that are acceptable and feasible. Findings from multiple studies103 have shown the importance of engaging community members at an early stage to create understanding, build ownership, and make optimal use of local resources. Formative research is needed, as a principle, to complete a rigorous process of adaptation design and testing.104

    View all citing articles on Scopus
    View full text