A novel data science and network analysis approach to quantifying facilitators and barriers of low tidal volume ventilation in an international consortium of medical centers

点击翻译按钮获取中文摘要

PROJECT SUMMARY/ABSTRACT This application, “A novel data science and network analysis approach to quantifying facilitators and barriers of low tidal volume ventilation in an international consortium of medical centers,” is in response to PAR-16-238, Dissemination and Implementation Research in Health (R01). Acute respiratory distress syndrome (ARDS) has high prevalence (10% of intensive care unit admissions) and mortality up to 46%. Low tidal volume ventilation (LTVV) is the most effective therapy for ARDS, lowering mortality by 20-25%, and is part of standard practice. However, use of LTVV is as low as 19% of ARDS patients. There is a poor understanding of the barriers to LTVV adoption: current approaches are deficient because they incorporate biases, lack consistency and comprehensiveness, ignore the influence of interpersonal network- or team- based factors, and do not address setting-specific variation. Our research team has previously identified some patient- and clinician-specific facilitators of and barriers to LTVV adoption. We have used two state-of-the-art data driven methods—data science and network analysis—to preliminarily quantify the impact of a diverse array of potential factors affecting LTVV adoption, including network- and team-based factors. The proposed research is guided by the Consolidated Framework for Implementation Research (CFIR) and Rogers' Diffusion of Innovations theory. The overall goals of the proposed research are to understand the differences in facilitators and barriers to LTVV adoption between academic and community settings through a definitive, systematic study in a large, diverse consortium of medical centers, and to advance implementation science by providing a model for how data science and network analysis can be applied to understand the adoption of a complex intervention. The overarching hypothesis is that there are different patient-, clinician-, network-, and team-based facilitators and barriers to LTVV adoption in academic and community settings. We will determine whether different patient- and clinician- (Aim 1 cohort study, clinician survey, and data science analysis), clinician interpersonal network- (Aim 2 network analysis), and team structure and dynamics-based (Aim 3 team construction and modeling) facilitators of and barriers to LTVV adoption exist between academic and community hospital settings. Successful completion of the proposed research will provide a comprehensive understanding of the differences in the facilitators of and barriers to LTVV adoption between academic and community settings, and will advance implementation science by serving as a model of how data science and network analysis can be applied to complex implementation problems. Implementation strategies that account for all these factors may be more likely to lead to significant practice change.