This is a book about implementation research. It introduces the concept of implementation research and a tool for evaluating it.
The CFIR Index Manual, Version 3.1, from the Stanford University School of Medicine explains how to administer and score the Consolidated Framework for Implementation Research (CFIR) Index, a quantitative measure of the contextual barriers and facilitators that influence an implementation process. The index contains 29 items across four dimensions, rated on a 5-point Likert-type scale that ranges from -2 (Barrier) to 0 (Neutral) to +2 (Facilitator), using information gathered by trained raters during a half-day site visit that includes key informant interviews, document review and ethnographic observation.
The manual covers background on implementation science, methodology and data sources, site visit procedures, scoring and interpretation, feedback to programs, and a detailed description of every dimension and item with sample questions and response coding. It also reports psychometric results, including reliability and validity findings from the two National Institute on Drug Abuse clinical trials conducted in Vermont and New Hampshire from 2014 to 2016, and ends with sample questions and references.
| Number of items and dimensions | 29 items over four dimensions |
|---|---|
| Rating scale | 5-point Likert-type scale with specific anchors |
| Scale range | -2 (Barrier) to 0 (Neutral) to +2 (Facilitator) |
| Inter-rater reliability | ICC of .86 (p< 0.001) |
| Data collection format | Half-day site visit by independent trained raters |
| Version and date | Version 3.1, May 6, 2019 |
The measure consists of 29 items over four dimensions. The fifth CFIR dimension, Process, and the Design Quality and Packaging item were dropped.
Ratings are made on a 5-point Likert-type scale with specific anchors, ranging from -2 (Barrier) to 0 (Neutral) to +2 (Facilitator).
Independent and trained raters gather information during a half-day site visit through key informant interviews, document review and ethnographic observation.
Assefa and McGovern found an intra-class correlation of .86 (p< 0.001) across all CFIR Index items at two sites and two raters.