Beyond To Err Is Human: From Error Events to Error Processes in Healthcare

Abstract

Twenty-five years after the publication of To Err Is Human, patient safety has successfully shifted attention from individual blame to system-level explanations of failure. Despite this progress, patient safety remains largely organized around an event-based view of error in which adverse events, near misses, and incident reports serve as the primary units of analysis. Errors are typically identified, investigated, and understood after they become visible. In this paper, I argue that this perspective is increasingly insufficient for understanding contemporary healthcare failures. Many errors do not emerge as isolated events but develop gradually through evolving interactions among individuals, teams, technologies, and organizational conditions. Rather than conceptualizing errors as discrete occurrences, I propose understanding them as processes that unfold over time and across organizational levels. Drawing on research on latent errors, organizational errors, multilevel dynamics, organizational drift, adaptation, and process theory, I develop a process perspective on error. Whereas systems thinking explains the conditions that make failure possible, a process perspective explains how those conditions become connected, amplified, and transformed into failure over time. The central contribution of the paper is to shift the object of explanation from adverse events to the trajectories through which adverse events emerge. The process perspective therefore shifts patient safety from explaining visible failures toward understanding the emergence of failure. This perspective is particularly relevant in contemporary healthcare, where digital technologies, artificial intelligence, and increasingly interconnected systems both create new pathways through which failures develop and provide new opportunities to observe those processes. By making error trajectories, weak signals, and emerging vulnerabilities increasingly visible, these technologies create new possibilities for anticipating and intervening in failure before adverse outcomes occur. I conclude that the next frontier of patient safety lies in complementing event-based approaches with a richer understanding of error trajectories and the mechanisms through which they emerge.