Barcode medication administration has been shown to improve medication safety. This pre–post study analyzed voluntary error reports at a single academic medical center and found that adverse drug events decreased after barcode medication administration was implemented, as seen in prior studies.

This article uses data from a large database of voluntarily reported errors to characterize errors in laboratory medicine. Most errors occurred at the preanalytic phase (before the specimen arrived in the laboratory), with many errors arising from misidentification of specimens. The vast majority of the more than 30,000 errors analyzed did not lead to patient harm. A pathology misidentification error that resulted in disclosure of an incorrect diagnosis to a patient is discussed in this AHRQ WebM&M commentary.

Hospital incident reporting systems are ubiquitous but limited, as their voluntary nature results in many events going unreported. Prior research has documented that physicians, in particular, do not file incident reports. This evaluation of more than 260,000 incident reports from a broad cross-section of hospitals examined links between the severity of the incident and who reported the incident. Physicians reported only 1.1% of all events, similar to a prior study using the same incident reporting system, but physicians did tend to report incidents that caused more harm to patients. A successful intervention to improve physician incident reporting was described in a prior study.

This descriptive study analyzed nearly 100,000 reports from 26 acute care hospitals with investigators discovering wide variations in reporting rates across sites. The most common classification included medication-related events, and more than half of all events affected a patient before being caught. The authors report that nurses were the most frequent users of the electronic reporting systems, whereas physicians accounted for an overwhelming minority. A past study found similar underuse of reporting systems by physicians and recommended alternative methods for capturing physician-based information about adverse events.

This editorial builds on the discussion from a study suggesting that overall improvement in the adoption and implementation of patient safety systems is slow. The authors offer a series of explanations for these delays in important improvements and apply the concept called the "theory of constraints." This theory asks the question of what should change, to what should it change, and how should change occur. Responses are framed with discussion of six thought processes that must occur at an organization for change to become possible. These include agreement that a problem exists, agreement that a proposed solution actually solves the problem, and identifying obstacles and how they can be overcome. The authors argue that sustained change occurs only when these root causes receive appropriate exploration and direct action in fostering improved safety systems.