Despite past efforts to improve patient safety—and there has been effort and activity aplenty—routine safety processes continue to fail routinely. Poor hand hygiene remains a major vector of health care–associated infection. Medication errors lead to adverse drug events with alarming frequency. Transitions in care often introduce new opportunities to harm patients. Less common but entirely preventable events such as wrong site surgery and retained foreign objects following surgery, to name just two, have not abated despite major efforts aimed at eliminating them. This lackluster history of combating preventable harm to patients may not be due to the fact that we have the wrong solutions. Rather, we have not fully understood the problems that the solutions are meant to address.
We have tended to rely on best practices, toolkits, protocols, checklists, and solution bundles as if they are magic bullets. These generalized approaches are most successful when a process, such as insertion of central lines in ICU patients, varies little from place to place and the causes of failure are few and common. Such circumstances are rare. Most persistent safety problems are laden with greater complexity and variation. Addressing these requires an approach that can uncover the many underlying causes of the same problem, recognize the variability in underlying causes from place to place, and customize solutions to directly target the factors in play.