A scoping review of implementation science theories, models, and frameworks - an appraisal of purpose, characteristics, usability, applicability, and testability.

BACKGROUND

A proliferation of theories, models, and frameworks (TMFs) have been developed in the implementation science field to facilitate the implementation process. The basic features of these TMFs have been identified by several reviews. However, systematic appraisals on the quality of these TMFs are inadequate. To fill this gap, this study aimed to assess the usability, applicability, and testability of the current TMFs in a structured way.

METHODS

A scoping review method was employed. Electronic databases were searched to locate English and Chinese articles published between January 2000 and April 2022. Search terms were specific to implementation science. Additionally, hand searches were administered to identify articles from related reviews. Purpose and characteristics such as the type of TMF, analytical level, and observation unit were extracted. Structured appraisal criteria were adapted from Birken et al.'s Theory Comparison and Selection Tool (T-CaST) to conduct an in-depth analysis of the TMFs' usability, applicability, and testability.

RESULTS

A total of 143 TMFs were included in this analysis. Among them, the most common purpose was to identify barriers and facilitators. Most TMFs applied the descriptive method to summarize the included constructs or the prescriptive method to propose courses of implementation actions. TMFs were mainly mid-range theories built on existing conceptual frameworks or demonstrated grand theories. The usability of the TMFs needs to be improved in terms of the provision of conceptually matched strategies to barriers and facilitators and instructions on the TMFs usage. Regarding the applicability, little attention was paid to the constructs of macro-level context, stages of scale-up and sustainability, and implementation outcomes like feasibility, cost, and penetration. Also, fewer TMFs could propose recommended research and measurement methods to apply the TMFs. Lastly, explicit hypotheses or propositions were lacking in most of the TMFs, and empirical evidence was lacking to support the claimed mechanisms between framework elements in testability.

CONCLUSIONS

Common limitations were found in the usability, application, and testability of the current TMFs. The findings of this review could provide insights for developers of TMFs for future theoretical advancements.