The Atlas Context Data Repository: A Feasible, Acceptable, and Useful Prototype for Context Data Collection and Future Predictive Analysis

Abstract

Background: Evidence from health care and other fields indicates that context matters for successful implementation of quality improvement initiatives. Most context assessment frameworks and tools are evidence-informed; that is, based on literature and expert opinion/consensus. Quantitative evidence is lacking on which aspects of context most strongly predict implementation success. Implementers need this evidence to prioritize contextual factors for assessment and action.

Methods: The Atlas Initiative developed a prototype of surveys and data collection processes for populating the Atlas Context Data Repository, a large database that will enable statistical analyses of the relationship between context and implementation success. The prototype was piloted with eight US health care sites to assess its feasibility, acceptability, and utility for individual implementers and for use at scale. The researchers collected and analyzed quantitative data from the repository surveys, shared context assessment reports with sites, interviewed sites about their experiences, and held internal team debriefs.

Results: Sites found the Atlas prototype feasible and acceptable to implement and the context assessment reports useful for understanding their strengths and challenges. The Atlas Initiative successfully recruited repository participants and supported survey administration. The sample data, although too small to make predictive inferences, had reasonable distributions and low missingness.

Conclusion: The results indicate that the Atlas Context Data Repository can be feasibly implemented at scale with minor modifications to the prototype, that the repository database can support future statistical analyses when it has more data, and that participation in the repository is valuable and worthwhile for sites.