Prior research has found that electronic health record (EHR) implementation has introduced risks to patient safety. Using data from one hospital’s EHR system, this study reviewed active allergy alerts in patient records and concluded that 37% of those records required reconciliation of allergy information across different areas of the EHR. These findings highlight the need for automated reconciliation algorithms and clinical decision support tools to help clinicians identify potential allergy discrepancies and avoid patient safety risks.

… team, and systems levels.  … Phadke NA, Wickner P, Wang L, et al. Allergy safety events in healthcare: …

Human factors (HF) is the interaction between workers, equipment, and the environment. The COVID-19 pandemic has accelerated the adoption of HF in intensive care units across the globe. This paper expands on the core concepts of HF and proposes the additional key concepts of agility, serendipity, innovation, and learning. Adoption of these HF concepts by leadership and staff can improve patient safety in intensive care units in future pandemics and other crisis situations.

… errors in nursing home residents. … Kane- Gill SL, Wong A, Culley CM, et al. Transforming the medication regimen …

Artificial intelligence can support improved patient safety outcomes. This study found that a deep learning model can accurately and efficiently identify allergic reactions in hospital safety reports and can potentially enable real-time event surveillance and system improvement.  

… events occurred per 100 overrides.  … Edrees H, Amato MG, Wong A, Seger DL, Bates DW. High-priority drug-drug …

Although clinical decision support is intended to improve safety, decision support alerts often result in alert fatigue and overrides. This prospective observational study examined overrides for exceeding the maximum dose of a medication in the intensive care unit. Researchers determined that insulin was the most frequent medication for which a maximum dosage alert was overridden. In almost 90% of cases, the overrides were deemed clinically appropriate. The authors conclude that more intelligent clinical decision support for medication dosing is needed to balance safety with alert fatigue in the intensive care unit. A past PSNet perspective discussed the challenges of implementing effective medication decision support systems.

Natural language processing (NLP) can efficiently analyze large narrative data sets to identify adverse events. Exploring the application of NLP to reduce medication errors, this AHRQ-funded review describes challenges associated with using NLP to extract information from clinical sources and highlights how engaging pharmacists in developing NLP systems can improve medication safety.

Computerized provider order entry and decision support alert providers to potential prescribing errors and drug–drug interactions that may cause harm. However, prior research has shown that inappropriate medication alert overrides occur frequently and place patients at risk for adverse drug events (ADEs). The cost of these preventable ADEs on a national level remains unknown. Extrapolating medication order data from a random sample of patients at a single academic medical center over a 1-year period, researchers estimate that in 2014 there were 29.7 million adult inpatient discharges generating anywhere from 1.02 billion to 1.07 billion medication orders and leading to between 75.1 million and 78.8 million medication alerts, depending on the model used. They conclude that around 5.5 million medication alerts may have been inappropriately overridden and estimate the associated cost of treating the preventable ADEs resulting from these overrides to be between $871 million and $1.8 billion. A past Annual Perspective discussed the role of computerized provider order entry in patient safety.

… emergency medicine and radiology notes . A  WebM&M commentary  discussed an adverse event that occurred due to a transcription error in a radiology study report. … Zhou L; Blackley SV; Kowalski L; Doan R; Acker WW; Landman AB; Kontrient E; Mack D; Meteer M; Bates DW; Goss FR. …

… process to prevent adverse medication events . A past WebM&M commentary discussed strategies to reduce errors associated …

Clinical decision support systems in electronic health records (EHRs) aim to avert adverse events, especially medication errors. However, alerts are pervasive and often irrelevant, leading patient safety experts to question whether their modest improvement in safety outweighs the harms of alert fatigue. This study assessed provider overrides of a commercial EHR's medication alerts in intensive care units at one institution. Providers overrode most alerts, and the majority of those overrides were appropriate. Inappropriate overrides occasionally led to medication errors and did so more frequently than appropriate overrides. A recent WebM&M commentary recommends employing human factors engineering to make clinical decision support more effective.

Medication-related clinical decision support is a ubiquitous component of computerized provider order entry (CPOE). Alerts are intended to reduce medication errors and may improve adherence to recommended treatments, but they have yet to improve clinical outcomes. This cross-sectional study examined how often inpatient providers overrode clinical alerts as well as whether those overrides were appropriate. Over 3 years, clinicians overrode nearly 340,000 alerts. While nearly all duplicate drug alert and drug allergy overrides were appropriate, most renal or age contraindication overrides were inappropriate. Although this single institution investigation of a homegrown, older CPOE system may not be generalizable to more common electronic health records, it does illustrate how alert fatigue compromises patient safety. A previous WebM&M commentary discussed the challenges of designing safe CPOE.

This retrospective study reviewed more than 47,000 overridden medication alerts and found that the vast majority of overrides were clinically appropriate and did not cause harm. From this sample, 7 adverse drug events were identified, and these events were more likely when the alerts were overridden in error. This study demonstrates the challenge of identifying clinically important alerts in a setting where alert fatigue is common.

The adoption of new technology in health care often produces unintended consequences, which can be mitigated by applying human factors engineering principles to user interface design. Due to efficiency gains, the use of speech recognition technology among physicians has grown in recent years. Investigators analyzed notes dictated by emergency medicine physicians and found that 71% of the notes contained errors. Given that 15% of the errors were considered critical, the authors suggest speech recognition technology may create miscommunication that could adversely affect patient care.

Alert fatigue is recognized as a barrier to patient safety and may particularly increase risks associated with medication prescribing. This study examined the frequency of manual overrides of alerts for medication allergies over a 10-year period. Clinicians were required to provide a reason for overriding the allergy alert. As with earlier studies, the rate of overrides was very high. Researchers determined that the alerts were irrelevant in more than half the cases. Providers also were more likely to override repeated alerts compared with new alerts. These results highlight the overuse of alerts in health care settings and the need to improve their use to effectively support patient safety.

Overdose of the commonly used over-the-counter analgesic acetaminophen can lead to serious liver toxicity. In the outpatient setting, unintentional acetaminophen overdose has been linked to poor health literacy and lack of standardization of medication instructions. This study of acetaminophen prescribing among inpatients at two academic medical centers found that 6.6% of patients received more than the recommended daily dose of the drug, with elderly patients and patients with preexisting liver disease being most vulnerable to dosing errors. The authors identified several other risk factors for errors, including use of scheduled around-the-clock dosing (instead of as-needed dosing) and prescribing of more than one acetaminophen-containing product. Notably, these errors occurred at hospitals with computerized provider order entry systems, and the authors note that existing clinical decision support systems lack the capability to warn clinicians when the maximum daily dose of a medication is about to be exceeded.