Abstract
Uptake of evidence-based practices (EBPs) in pulmonary and critical care medicine is frequently incomplete. To address these gaps, implementation scientists seek to understand the clinical and societal contexts in which innovations and EBPs are introduced. They also design and evaluate complex interventions to facilitate the adoption of an EBP in those contexts. We propose that well-established methods for analyzing hierarchical, observational data can complement and strengthen this process by identifying sources of practice variability. This paper reviews the dominant framework used to understand the clinical context of implementation programs, describes how measuring practice variability could help streamline this approach, and tests an assumption of the proposed combined methodology using observational data from a national study of patients on mechanical ventilation conducted by the Prevention and Early Treatment of Acute Lung Injury Network. We discuss how the combined approach can be used (1) to focus the search for determinants of practice, (2) to quantify the impact of evidence generation and evaluate the success of implementation projects, and (3) to facilitate comparisons between implementation strategies when multiple approaches are trialed simultaneously.