Assessing Antibiotic Stewardship Program (ASP) Implementation: Psychometric Properties of Survey measures based on the Consolidated Framework for Implementation Research (CFIR)

Background: Antibiotic stewardship programs (ASPs) are crucial to prevent the emergence of antibiotic resistance and to improve outcomes for patients. Few measurement instruments examine ASP implementation. A validated instrument rooted in implementation science theory will increase our understanding of ASP implementation and enable comparisons across studies. Methods: We used the Consolidated Framework for Implementation Research (CFIR) to guide development of ASP implementation measures. Survey questions were developed by infectious disease and survey design experts to assess 22 potential determinants of implementation across five domains of the CFIR. Survey items were reviewed by CFIR experts for face and content validity. Antibiotic stewards (infectious disease pharmacists and physicians) were recruited from 154 Veterans Affairs (VA) hospital and Intermountain Healthcare (IH) sites to complete the survey. Survey responses were aggregated by site, and results were determined at the site level. We conducted confirmatory factor analyses (CFA) to assess structural validity of measures and evaluated inter-item consistency. Results: A total of 182 stewards completed the survey (155 VA and 27 IH responses). At least one response was obtained from 126 hospitals (81.8%). CFAs for most CFIR constructs exhibited good fit (CFI > .90; TLI > .90; RMSEA< .05, SRMR < .08). Scale reliabilities ranged from .54-.96 indicating low to strong inter-item consistency. Determinants that were rated highly present at the sites (across site means ³ 4.0 or above) included Self-Efficacy, Engaging, Evidence Strength and Quality and Relative Advantage, indicating stewards found ASP evidence compelling and felt their personal involvement was effective in engendering positive results for the ASP. Conclusions: Psychometric properties measured via CFA indicate validity of our CFIR-based measure of determinants for ASP implementation outcomes. This is the first validated set of measures based on the CFIR for ASP implementation and the first to include measures for 4 out of five CFIR domains. These validated measures will help to provide much needed understanding of barriers and facilitators of ASP implementation. Research and quality improvement teams can use these measures to identify contextual determinants of ASP implementation and use this information to guide selection of strategies and compare results across multiple studies.