An epistemology of patient safety research: a framework for study design and interpretation

This series of four articles provides a valuable and timely analysis of methodology for evaluating interventions to improve patient safety and healthcare quality improvement. The first article emphasises the important role for strategic analysis of previous research to maximise the chances of success. It borrows from Donabedian to describe a “causal chain” that explores how an intervention might affect an organisation’s processes and health outcomes. The second article focuses on study design and makes the often-overlooked point that the salient intra-class correlations in cluster studies are smaller with before and after designs than with cross-sectional designs. The authors explore in depth the stepped wedge, a valuable controlled before-and-after design. The third article examines study end points. …