Modeling the Scaling-up of Early Crime Prevention: Simulating Obstacles to Identify Opportunities

Although research on the effectiveness of developmental prevention strategies has grown considerably in recent decades, the translation of scientific knowledge into evidence-based policy and practice in this area remains challenging.  Building on recent literature reviews and insights from effective interventions, this paper considers the obstacles associated with scaling-up evidence-based programs and proposes an approach for evaluating their possible effects on later outcomes.  Previous work has highlighted a number of factors that may affect the prospects of scaling-up evidence-based programs, including (1) heterogeneous and divergent target populations, (2) fidelity to the treatment model, and (3) implementation context.  These factors have not been enumerated or estimated with much degree of precision to date.  In turn, this has necessitated the use of arbitrary and general “penalties” when evaluating prospects for program dissemination.  Data from the series of randomized controlled experiments of the Nurse-Family Partnership (NFP) are used to develop a computer simulation model that varies some representative inputs under each of the three factors.  The possible impacts on the outcome(s) of the process (e.g., intervention effects on later criminal behavior) are then examined.  Specifically, we use data and findings from NFP studies to seed the simulation model to (1) generate an ideal, control system that preserves effects (or counters some expected attenuation of effects) upon scaling-up the program and (2) systematically alter particular components to “experimentally” assess the impact on the model as it represents the decay/stability of the effect across the scale-up process.  The analysis then considers the usefulness of this model in terms of its correspondence with actual results observed in the process of scaling-up NFP.  From there, simulation is considered as a possible tool for enhancing the formative understanding of the possible impacts of interventions in situations where the ability to manipulate “real world” conditions is otherwise limited.  In this sense, computer simulation will be investigated as a tool to potentially help in intervention planning.  The paper’s discussion considers the implications of the simulation study for enhancing knowledge on scaling-up evidence-based practices and also assesses the possible value of the method to implementation science in prevention programs more broadly.    It is crucial that these questions are considered further as the success of translational research depends not only on the evidence that comes from rigorous research on effective programs, but also necessarily requires some understanding of the process of subsequently taking those programs to scale.