Implementation Science is a field of research that specifically addresses program implementation and its link to outcomes.  However, it is challenging to meaningfully applying existing implementation frameworks to federally funded programs. Established implementation science frameworks are geared to well-defined, evidence-based interventions (CERAG, 2009; NIRN, 2011; Barbeea, et. al., 2011, Proctor, et. al., 2009). Federal agencies we work with often have discretionary grantees that are generally engaged in developing innovations that might become evidence-based models and practices, rather than implementing established evidence-based models.

This report describes research and observations that suggest organizational characteristics, activities, and processes (known as “implementation drivers”) facilitate the successful implementation of innovative program models and practices.  This information was obtained by means of a qualitative research process with two separate groups of federally funded programs from 2010 -2011 involving site visits and conference calls with local community programs that demonstrated initial evidence of successful program implementation. In-depth discussions with grantee staff members and other stakeholders were used to explore the relevance of more than 115 implementation drivers and concepts adapted from NIRN, Getting to Outcomes (GTO), and other implementation science models to federally funded grant projects.

This research demonstrated that implementation science models such as National Implementation Research Network’s (NIRN) that were designed specifically to promote the successful replication of evidence-based programs and practices in community settings are not always applicable to federally funded grantees and may not account for all the factors that facilitate their successful implementation. First, established implementation science models are generally geared toward existing well-defined, evidence-based interventions; however, federal discretionary grantees most often focus on developing innovations that might become evidence-based models or practices in the future rather than on implementing established evidence-based models.  In addition, existing implementation science models do not take into consideration the realities and constraints imposed by the Federal grant award and management process. For example, organizations that seek Federal grants must respond to Funding Opportunity Announcements within short timeframes and adapt their proposals to be responsive to problems or issues that have already been identified by the Federal government. The implementation timeframes of discretionary grant projects are further constrained by the limited duration of Federal grants, which typically last between three and five years. Existing implementation science models generally do not account for this narrow window for Federal funding, which affects the speed and trajectory of the entire implementation process from start-up to long-term program adaptation and sustainability.   

Despite the limitations of applying existing implementation science models to federal discretionary grantees, this research determined that many of the implementation drivers and concepts described by NIRN, GTO, and others models are useful and relevant to understanding the successful implementation of time-limited child welfare projects that are not necessarily based on an evidence-based program or practice. Further this research suggests that the application of many concepts from established models may be useful in understanding the drivers that contribute to the successful implementation of program models.