Using Implementation Science Frameworks in Genomics and Precision Medicine: We Can Do Better!Posted on by
A recent scoping review identified many structured approaches to the implementation of genomics and precision medicine and limited use of implementation science frameworks.
With continuous advances in genomics and accelerated translation from discovery into clinical practice, in our blog posts, we have repeatedly examined the crucial importance and emerging role of implementation science in the field (see list below for selected CDC posts).
The scoping review examined “structured approaches” to genomic medicine implementation. The authors defined structured approaches “as the evidence-informed theories, frameworks, and/or models guiding the implementation processes.” The authors reference P. Nilsen’s publication, Making sense of implementation theories, models and frameworks and assert that implementation through the use of these structured approaches is more likely to be successful.
The scoping review identified 34 reports that presented structured approaches to the implementation of genomic medicine. The reports were categorized into 4 groups that 1) describe/guide the process of translating research into practice; 2) understand and/or explain what influences implementation outcomes; 3) evaluate implementation; or 4) provide general discussion and exploration. Additionally, within these studies, the structured approaches were analyzed for the extent to which they were informed by evidence; how they were deployed in the genomic setting (i.e. using TSci Impact Framework to categorize translation functions of pre-adoption, adoption, implementation and sustainability); and how they measured intended outcomes. See Table 1 from the scoping review for complete categorization of structured approaches and studies referencing each approach.
From the 34 studies/reports, 30 unique structured approaches and 4 pre-existing approaches to implementation of clinical genomics were identified. In the analysis of the structured approaches used, most approaches were reported at the preadoption stage (n=21), followed by implementation (n=7), and adoption (n=7). None were reported at the sustainability stage. Interestingly, of the 30 existing structured approaches, only 5 were implementation science approaches. Most of the studies utilizing structured approaches were published within the last 5 years (n=23). Breaking down the type of approached used, the majority were ‘process’ (n=16). The type of outcome that the structured approaches aimed to address included service (n=14), health (n=12), and implementation outcomes (n=12). Genomic stakeholders were consulted in the development of structured approaches in 11 studies.
Results from the scoping review reinforced the findings of Roberts et al. (2017) that identified a paucity of implementation science principles applied to genomic medicine. The authors of the present review discuss possible reasons for the lack of theory-supported implementation in genomics, and highlighted efforts to improve on these issues. Stakeholder engagement in the implementation process is essential for transferable learning. Multidisciplinary involvement from the initiation of studies, including genomic experts and implementation scientists is critical. Articulating outcome measures and use of structured approaches for evaluation are important. Organizational and local level commitment to enable the time required by researchers and practitioners is needed. The authors suggest a common repository to share genomic implementation resources. In addition, they recommend that those responsible for the implementation of genomic medicine into clinical care utilize theory and expertise of implementation scientists to inform the design of implementation strategies.
In the past two decades, dozens of implementation science theories, models and frameworks have been developed and used in hundreds of studies focused on the dissemination of evidence and implementation of evidence-based interventions. Indeed, tools and resources have been developed (e.g. dissemination-implementation.org, T-CaST [Birken et al, 2018]) to support the selection of models for a range of different purposes, many of which could relate to applications of genomic medicine in clinical practice.
As the number of evidence-based genomic applications increases in the next decade, an important challenge facing us is how to build the national, state, community and health systems research and evaluation infrastructure for using principles of implementation science in genomic medicine. The ultimate success of genomic medicine in improving health will depend on it.
We welcome the input of our readers. Submit your comments here.