Bridging The Gap Between Implementation Science and Parenting Intervention

Kaminski et al.¹ recently reported selected findings from a study evaluating two group-based parenting interventions designed to promote “more positive socioemotional, behavioral, cognitive, and health outcomes” for children in low-income families (defined as < 200% of the federal poverty line). The authors provide a laudably clear and detailed flow diagram of study enrollment and follow-up assessment rates. However, specific and meaningful information about intervention exposure, participation rates, and other fundamental implementation considerations are not provided alongside the reporting of outcomes. The omission of core implementation data brings into sharp focus two interrelated and chronic limitations in the intervention research literature, in general, and particularly in the knowledge base on parenting interventions. These gaps, described below, heavily constrain the public health impact of programming for vulnerable families and children.

This first limitation is inattention to or inaccuracy in describing where a study falls on the research-to-practice continuum,^2,3 obfuscating the reader’s ability to evaluate the generalizability, applicability, and transferability of a study’s findings. For example, Kaminski et al. use the term “effectiveness” in reference to the evidence generated from what are clearly efficacy trials. The authors also characterize two different approaches, each with “its own curriculum” and simultaneously tested in initial efficacy trials, as “implementation adaptations” of a single “public health model” (based on a shared “philosophy” and “core set of goals.”). Accuracy matters, especially when presenting findings to a broad public health audience that includes policymakers and practitioners with neither the time nor expertise to disentangle misapplied terms and concepts.

The second limitation pertains to the need for researchers to provide greater information on external validity in both efficacy and effectiveness studies. A decade ago, Glasgow et al.⁴ put forth a set of recommendations for specific changes that researchers, journal editors, and funding organizations can make to meet this need. These changes include “increased attention to moderating factors … such as participant characteristics (reach) and setting characteristics (adoption)”^4(p1264) on the part of researchers. In 2008, an editorial by Steckler and McLeroy⁵ announced the new emphasis of the American Journal of Public Health on external validity and identified specific categories of external validity information that should be reported when appropriate, noting that:

It has been frequently argued that internal validity is the priority for research. However, in an applied discipline, the purpose of which includes working to improve the health of the public, it is also important that external validity be emphasized and strengthened.^5(p9)

The call for improved reporting on external validity “to guide decision makers in their selection of research-tested interventions” vigorously continues today.⁶

Perou et al.⁷ report that extensive implementation data on both process and cost were collected as part of the Legacy study, completed in 2009. Participation rates and cost data for the Legacy interventions are highly salient—arguably critical—to interpreting intention-to-treat outcomes, given the serious challenges to reach and sustainability documented in the literature on preventive parenting interventions.^8,9 Kaminski et al. describe a set of closed parenting groups, designed to be three or five years in duration and comprising 101 or 250 weekly sessions depending on approach. The literature is well established regarding the challenges of low enrollment, poor attendance, and high attrition in group-based parent-mediated interventions.¹⁰ This finding is consistent even for brief approaches.¹¹ The issue here is not whether intensive and long-term interventions are necessary. The crux of the matter is the need for greater specificity about the magnitude, timing, duration, cointerventions, and population characteristics associated with enduring positive outcomes.¹²

To make judicious choices in a time of extremely limited resources for maternal and child health programs, decision-makers need confidence that a highly intensive approach can be implemented (applicability) and outcomes replicated (transferability) within feasible cost parameters of community settings. In a recent article elaborating on themes set forth in the 2009 landmark report on prevention science by the National Research Council and Institute of Medicine,¹³ Yoshikawa et al.¹¹ underscore the urgent need for cost-effectiveness studies on mental health prevention and promotion strategies. Analysis of the interplay of dosage, participant characteristics, and service delivery factors is essential to effective programming and policies, and it demands the same primacy and immediacy of consideration as the results of an intention-to-treat analysis. This is a crucial direction for future research on both selective and indicated¹⁴ intervention efforts addressing children’s mental, emotional, and behavioral health.