UCLA

Background

This paper identifies and describes measures of constructs relevant to the adoption or implementation of innovations (i.e., new policies, programs or practices) at the organizational-level. This work is intended to advance the field of dissemination and implementation research by aiding scientists in the identification of existing measures and highlighting methodological issues that require additional attention.

Methods

We searched for published studies (1973-2013) in 11 bibliographic databases for quantitative, empirical studies that presented outcome data related to adoption and/or implementation of an innovation. Included studies had to assess latent constructs related to the "inner setting" of the organization, as defined by the Consolidated Framework for Implementation Research.

Results

Of the 76 studies included, most (86%) were cross sectional and nearly half (49%) were conducted in health care settings. Nearly half (46%) involved implementation of evidence-based or "best practice" strategies; roughly a quarter (26%) examined use of new technologies. Primary outcomes most often assessed were innovation implementation (57%) and adoption (34%); while 4% of included studies assessed both outcomes. There was wide variability in conceptual and operational definitions of organizational constructs. The two most frequently assessed constructs included "organizational climate" and "readiness for implementation." More than half (55%) of the studies did not articulate an organizational theory or conceptual framework guiding the inquiry; about a third (34%) referenced Diffusion of Innovations theory. Overall, only 46% of articles reported psychometric properties of measures assessing latent organizational characteristics. Of these, 94% (33/35) described reliability and 71% (25/35) reported on validity.

Conclusions

The lack of clarity associated with construct definitions, inconsistent use of theory, absence of standardized reporting criteria for implementation research, and the fact that few measures have demonstrated reliability or validity were among the limitations highlighted in our review. Given these findings, we recommend that increased attention be devoted toward the development or refinement of measures using common psychometric standards. In addition, there is a need for measure development and testing across diverse settings, among diverse population samples, and for a variety of types of innovations.