Evaluating and optimizing the consolidated framework for implementation research (CFIR) for use in low- and middle-income countries: a systematic review

Abstract

Background: The Consolidated Framework for Implementation Research (CFIR) is a determinants framework that

may require adaptation or contextualization to fit the needs of implementation scientists in low- and middleincome countries (LMICs). The purpose of this review is to characterize how the CFIR has been applied in LMIC

contexts, to evaluate the utility of specific constructs to global implementation science research, and to identify

opportunities to refine the CFIR to optimize utility in LMIC settings.

Methods: A systematic literature review was performed to evaluate the use of the CFIR in LMICs. Citation searches were

conducted in Medline, CINAHL, PsycINFO, CINAHL, SCOPUS, and Web of Science. Data abstraction included study

location, study design, phase of implementation, manner of implementation (ex., data analysis), domains and constructs

used, and justifications for use, among other variables. A standardized questionnaire was sent to the corresponding

authors of included studies to determine which CFIR domains and constructs authors found to be compatible with use in

LMICs and to solicit feedback regarding ways in which CFIR performance could be improved for use in LMICs.

Results: Our database search yielded 504 articles, of which 34 met final inclusion criteria. The studies took place across 21

countries and focused on 18 different health topics. The studies primarily used qualitative study designs (68%). Over half

(59%) of the studies applied the CFIR at study endline, primarily to guide data analysis or to contextualize study findings.

Nineteen (59%) of the contacted authors participated in the survey. Authors unanimously identified culture and engaging

as compatible with use in global implementation research. Only two constructs, patient needs and resources and individual

stages of change were commonly identified as incompatible with use. Author feedback centered on team level influences

on implementation, as well as systems characteristics, such as health system architecture. We propose a “Characteristics of

Systems” domain and eleven novel constructs be added to the CFIR to increase its compatibility for use in LMICs.

Conclusions: These additions provide global implementation science practitioners opportunities to account for systemslevel determinants operating independently of the implementing organization. Newly proposed constructs require further

reliability and validity assessments.

Trial registration: PROSPERO, CRD42018095762