Open Access Highly Accessed Open Badges Research

Measurement of a model of implementation for health care: toward a testable theory

Joan M Cook12*, Casey O’Donnell1, Stephanie Dinnen1, James C Coyne34, Josef I Ruzek25 and Paula P Schnurr26

Author Affiliations

1 Department of Psychiatry, Yale School of Medicine, 950 Campbell Avenue, NEPEC/182, West Haven, CT, 06516, USA

2 National Center for PTSD, 215 North Main Street, White River Junction, VT, 05009, USA

3 Department of Psychiatry, Perelman School of Medicine of the University of Pennsylvania, 423 Guardian Drive, Philadelphia, PA, 19104, USA

4 Department of Psychology, University of Groningen, the Netherlands, P.O. Box 660, 9700 AR, Groningen, The Netherlands

5 Department of Psychiatry, Stanford University-Menlo Park Division, 795 Willow Road, Menlo Park, CA, 94025, USA

6 Department of Psychiatry, Geisel School of Medicine, Dartmouth College, 1 Rope Ferry Road, Hanover, NH, 03755, USA

For all author emails, please log on.

Implementation Science 2012, 7:59  doi:10.1186/1748-5908-7-59

Published: 3 July 2012



Greenhalgh et al. used a considerable evidence-base to develop a comprehensive model of implementation of innovations in healthcare organizations [1]. However, these authors did not fully operationalize their model, making it difficult to test formally. The present paper represents a first step in operationalizing Greenhalgh et al.’s model by providing background, rationale, working definitions, and measurement of key constructs.


A systematic review of the literature was conducted for key words representing 53 separate sub-constructs from six of the model’s broad constructs. Using an iterative process, we reviewed existing measures and utilized or adapted items. Where no one measure was deemed appropriate, we developed other items to measure the constructs through consensus.


The review and iterative process of team consensus identified three types of data that can been used to operationalize the constructs in the model: survey items, interview questions, and administrative data. Specific examples of each of these are reported.


Despite limitations, the mixed-methods approach to measurement using the survey, interview measure, and administrative data can facilitate research on implementation by providing investigators with a measurement tool that captures most of the constructs identified by the Greenhalgh model. These measures are currently being used to collect data concerning the implementation of two evidence-based psychotherapies disseminated nationally within Department of Veterans Affairs. Testing of psychometric properties and subsequent refinement should enhance the utility of the measures.