Context: In response to national efforts to improve quality of care, policymakers and health care leaders have increasingly turned to quality improvement collaboratives (QICs) as an efficient approach to improving provider practices and patient outcomes through the dissemination of evidence-based practices. This article presents findings from a systematic review of the literature on QICs, focusing on the identification of common components of QICs in health care and exploring, when possible, relations between QIC components and outcomes at the patient or provider level.
Methods: A systematic search of five major health care databases generated 294 unique articles, twenty-four of which met our criteria for inclusion in our final analysis. These articles pertained to either randomized controlled trials or quasi-experimental studies with comparison groups, and they reported the findings from twenty different studies of QICs in health care. We coded the articles to identify the components reported for each collaborative.
Findings: We found fourteen crosscutting components as common ingredients in health care QICs (e.g., in-person learning sessions, phone meetings, data reporting, leadership involvement, and training in QI methods). The collaboratives reported included, on average, six to seven of these components. The most common were in-person learning sessions, plan-do-study-act (PDSA) cycles, multidisciplinary QI teams, and data collection for QI. The outcomes data from these studies indicate the greatest impact of QICs at the provider level; patient-level findings were less robust.
Conclusions: Reporting on specific components of the collaborative was imprecise across articles, rendering it impossible to identify active QIC ingredients linked to improved care. Although QICs appear to have some promise in improving the process of care, there is great need for further controlled research examining the core components of these collaboratives related to patient- and provider-level outcomes.
© 2013 Milbank Memorial Fund.