Artificial intelligence (AI) and machine learning (ML) are emerging as tools to improve patient care, but they are not without risks. This article proposes use of a safety checklist to determine readiness to launch AI technologies, prompting users to consider physical and mental health and economic and social risks and benefits.

Perioperative medication errors are common. This article highlights several interventions to reduce the risk of perioperative medication errors, including improved medication labeling, adoption of artificial intelligence for decision support and risk prediction, and the use of health information technology (IT), such as computerized physician order entry (CPOE), electronic medication administration records (eMAR), and barcode medication administration (BCMA).

Computerized provider order entry (CPOE), clinical decision support (CDS), and electronic medication management systems (EMMS) have increased efficiency and reduced prescribing errors, but poor design may introduce new safety hazards. Human factors and safety analysis methods can be used to increase the safety of new technologies, ideally before problems arise. This review identifies human factors and safety analysis methods applied to EMMS. Most methods focused on usability or design, and only one used a safety-oriented approach. Increased inclusion of human factors specialists could increase the use of safety-oriented methods of EMMS design.

The effective use of resources through stewardship initiatives can support error reduction through focusing actions of care. This commentary discusses how diagnostic stewardship can enhance diagnostic testing behaviors across the diagnostic process.

Poor usability of electronic health record (EHR)-based computerized provider order entry (CPOE) can lead to adverse events. Using a newly developed self-administered assessment tool, researchers identified several EHR usability and safety issues across medication, laboratory, and radiology CPOE functions.

Alert fatigue contributes to task burden and can threaten patient safety. In this study, researchers at one academic medical center found that machine learning techniques could enable intelligent filtering of medication alerts and reduce alert volume by 54%.

Computerized provider order entry (CPOE) can prevent prescribing errors, but patient safety threats persist. Based on qualitative interviews with multidisciplinary prescribers, the authors identified several issues related to CPOE interacting within a complex prescribing environment, including alert fatigue, remote prescribing, and default auto-population of dosages.

This study assessed the impact of electronic health record (EHR) interventions to standardize opioid prescribing practices across a large health system. Interventions included (1) deleting clinician preference lists, (2) default dose, frequency, and quantity, (3) standardizing formularies, and (4) dashboards with current opioid prescribing practices. In the 12 months after implementation, there was a decrease in the rate of opioid prescriptions overall, prescriptions exceeding three days, prescriptions exceeding prespecified morphine equivalent doses, and non-formulary prescriptions.

Using data from the Computerized Physician Order Entry (CPOE) Evaluation Tool, this study compared hospital performance against fatal orders and nuisance orders. From 2017 to 2018, overall performance increased and fatal order performance improved slightly; there was no significant change in nuisance order performance; however, these results indicate that fatal alerts are not being prioritized and that over-alerting in some cases may be contributing to alert fatigue.

Using root cause analysis data submitted to the Veterans Affairs (VA) National Center for Patient Safety from 2013 to 2018, this study analyzed health information technology (HIT)-related outpatient diagnostic delays to identify common safety concerns. The study identified five high-risk areas for diagnostic delays involving HIT: managing electronic health record inbox notifications and communications, clinicians gathering key diagnostic information, technical problems, data entry problems, and failure of a system to track test results.

This systematic review identified three critical points that can contribute to “failure to rescue” among inpatients with serious complications – (1) failure to recognize the complications; (2) failure to relay information regarding the complications to the care team, and; (3) failure to react in a timely and appropriate manner to the patient’s deterioration. Effective tools and interventions which can be implemented during each timepoint are discussed, including increased nurse staffing, rapid response teams, checklists, and early warning score systems.

The objective of this systematic review was to identify systemic defenses (such as barcode scanning) to confirm drug and patient identity, clinical decision systems, and smart infusion pumps) to prevent in-hospital intravenous (IV) medication errors. Of the 46 included studies, most discussed systemic defenses related to drug administration; fewer discussed defenses during prescribing, preparation, treatment monitoring and dispensing. Closed loop medication management and smart pumps were the most common systemic defenses examined in the included studies; the authors identify a need for further studies exploring the effectiveness of different combinations of systemic defenses.