All Content

Several factors contribute to the increased risk of prescribing errors for children, including weight-based dosing and drug formulation. This quality improvement project in one pediatric emergency department identified four key drivers and implemented four interventions to reduce errors. Prescribing errors were reduced across three plan-do-study-act cycles, and improvements were maintained six months after the final cycle.

Biases in decision support technologies precipitate racial inequities. This commentary discusses how algorithms in machine learning contribute to inaccuracies in the care of persons of color and the displaced. Legal actions to mitigate racial biases in decision making programs and implementation steps toward improvement are discussed.

Telehealth is increasingly used to connect rural hospitals with specialists in other areas and can improve patient outcomes. This study found that in situ simulation training in rural emergency departments resulted in small increases in the use of telemedicine for patients presenting with sepsis and led to improvements in sepsis process care outcomes.

Dashboards provide a way to display patient safety data (e.g., medication safety) in real time. This narrative review and SWOT (strengths, weaknesses, opportunities, threats) of existing patient safety dashboards can form the basis for developing future dashboards.

In 2016, the Centers for Disease Control (CDC) issued an opioid prescribing guideline intended to reverse the increasing death rate from opioid overdoses. This study describes the development, implementation, and effect of a safe prescribing strategy (TOWER) in an HIV-focused primary care setting. Providers using TOWER were more adherent to the CDC guidelines, with no worsening patient-reported outcome measures.

The advantages of automation can be safely achieved only when the technologies are implemented into processes that support their proper use in regular and urgent situations. This guideline outlines considerations for the safe use of computerized compounding devices to prepare parenteral nutrition admixtures with the broader application to other IV preparations in mind. Effective policy, training, system variation, and vendor partnerships are elements discussed.

Speech recognition software is a potential strategy to reduce documentation burden and burnout. This study compared the accuracy handwritten nursing notes versus online and offline speech recognition software. Findings indicate that the speech recognition software was accurate but created more errors than handwritten notes.

Patient portals allow patients and their caregivers to read clinical notes, view test results, and communicate with their provider, with the goal of improving patient safety. This scoping review found limited evidence of improved patient safety with the use of patient portals. Additionally, the authors found multiple naming conventions, such as patient portal, personal health record, and personal medical record.

Patient exposure to allergens healthcare settings, such as latex or certain medications, can lead to adverse outcomes. Based on data from an incident reporting system, researchers in this study developed a system for classifying allergy-related safety events. Classification categories include: (1) incomplete or inaccurate EHR documentation, (2) human factors, such as overridden allergy alerts, (3) alert limitation or malfunction, (4) data exchange and interoperability failures, and (5) issues with EHR system default options. This classification system can be used to support improvements at the individual, team, and systems levels. 

Diagnostic errors can be identified by measuring concordance of initial and final diagnosis, for instance admission and discharge diagnoses. In this study, researchers developed and tested an automated trigger system to determine concordance between pairs of diagnoses. In comparison to the reference standard (concordance determined by experts) the automated system performance was excellent.

Barcode medication administration (BCMA) processes are designed to prevent some types of medication administration errors. This article discusses how BCMA workflows support error prevention and how to identify workarounds that negate these error prevention mechanisms.

While artificial intelligence (AI) in healthcare may potentially improve some areas of patient care, its overall safety depends, in part, on the algorithms used to train it. One hospital developed four AI models at the start of the COVID-19 pandemic to predict risks such as hospitalization or ICU admission. Researchers found inconsistent instances of model-level bias and recommend a holistic approach to search for unrecognized bias in health AI.

Automated dispensing cabinets (ADCs) can reduce medication dispensing errors by requiring pharmacist verification. This study found that medication overrides (i.e., bypassing pharmacist review before administration) in one pediatric emergency department were frequently not due to an emergent situation requiring immediate medication administration and could have been avoided.

Clinician well-being is increasingly seen as a quality and safety issue. This commentary discusses how systemic efforts must be built to enhance occupational well-being among clinicians. This approach discussed should consider both human factors and organizational design strategies to reduce burnout, cognitive overload, process frustration, and technology use.

Transitions from one electronic health record (EHR) system to another can increase the risk of patient safety events. Using the principles of requisite imagination, this article outlines six recommendations for safe EHR transitions through proactive approaches, process improvement and support for healthcare workers.