Lewandowska K, Weisbrot M, Cieloszyk A, et al. Int J Environ Res Public Health. 2020;17:8409.
Alarm fatigue, which can lead to desensitization and threaten patient safety, is particularly concerning in intensive care settings. This systematic review concluded that alarm fatigue may have serious consequences for both patients and nursing staff. Included studies reported that nurses considered alarms to be burdensome, too frequent, interfering with patient care, and resulted in distrust in the alarm system. These findings point to the need for a strategy for alarm management and measuring alarm fatigue.
Auditory warnings to flag patients at risk for sepsis can have unintended consequences, such as alert fatigue or distraction. Although heightened awareness of sepsis is crucial due to its potential for harm, the authors call for rigorous study and testing of these systems to reduce their negative effects. They highlight how recently published negative results illustrate the importance of designing sepsis alerting functions that are safe and effective. A WebM&M commentary discussed a case involving a misdiagnosis of sepsis.
Segal G, Segev A, Brom A, et al. J Am Med Inform Assoc. 2019;26:1560-1565.
Alerts designed to prevent inappropriate prescribing of medications are frequently overridden and contribute to alert fatigue. This study describes the use of machine learning to improve the clinical relevance of medication error alerts in the inpatient setting.
Hussain MI, Reynolds TL, Zheng K. J Am Med Inform Assoc. 2019;26:1141-1149.
This systematic review examined the override rates of several different clinical decision support approaches. Researchers conclude that role tailoring—the provision of different alerts to prescribers versus pharmacists—was the most successful method to reduce alert fatigue. They recommend redesigning decision support to reduce alert fatigue.
Dr. Saria is the John C. Malone Assistant Professor of computer science, statistics, and health policy at Johns Hopkins University. Her research focuses on developing next generation diagnostic, surveillance, and treatment planning tools to reduce adverse events and individualize health care for complex diseases. We spoke with her about artificial intelligence in health care.
Nuisance alarms, interruptions, and insufficient staff availability can hinder effective monitoring and response to acute patient deterioration. This news article reports on how hospital logistics centers are working toward utilizing artificial intelligence to improve clinician response to alarms by proactively identifying hospitalized patients at the highest risk for heart failure to trigger emergency response teams when their condition rapidly declines.
Implementing new information systems can have unintended consequences on processes. This commentary explores insights from a physician, both as a clinician and as the family member of a patient, regarding the impact of hard stops in electronic health records intended to prevent gaps in data entry prior to task progression. The author raises awareness of the potential for patient harm due to interruptions and diminishing student and clinician skill in asking questions to build effective patient histories.
Alarm fatigue contributes to distraction and can diminish care safety. This commentary reviews a single-center project that used smartphone technology to enhance cardiac monitoring. The authors describe the structure of the project, use of Plan-Do-Study-Act cycles to design the application, results of the pilot, and plans to expand the use of this technology to other units and broaden monitoring targets. A WebM&M commentary discussed harm associated with alarm fatigue.
Wong A, Rehr C, Seger DL, et al. Drug Saf. 2019;42:573-579.
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.
In this magazine article, Atul Gawande describes a range of frustrations physicians experience as digitization becomes more widespread in health care. He elaborates upon several elements of electronic health record use that can degrade care processes and create conditions for errors, such as burnout, lack of patient-centeredness, and alert fatigue.
Medication errors continue to be a worldwide patient safety challenge that requires both systems and individual practice strategies for improvement. This magazine article describes how pharmacists can address failures associated with processing, dosing, care transitions, and information sharing to prevent medication errors.
Buckley MS, Rasmussen JR, Bikin DS, et al. Ther Adv Drug Saf. 2018;9:207-217.
This retrospective study examined the performance of trigger alerts designed to predict drug-related hazardous conditions in both ICU and non-ICU patients. The authors conclude that the alerts were not effective in identifying drug-related hazardous conditions in either setting and suggest that poorly performing alerts may contribute to alert fatigue.
Shah T, Patel-Teague S, Kroupa L, et al. BMJ Qual Saf. 2018;28:10-14.
Alert fatigue associated with electronic health records (EHRs) contributes to primary care physician burnout and can increase medication errors. The phenomenon is especially well-described in the Veterans Affairs (VA) system, where providers receive more than 100 alerts per day, which require an average of 85 seconds to address. This study describes a nationwide VA initiative to reduce EHR alerts in primary care and teach providers to process alerts more efficiently. Alerts decreased by a small but significant amount—from an average of 128 per day to an average of 116 per day. Providers who received the most alerts before the initiative experienced the largest alert reduction. A PSNet perspective described a way forward in improving EHR safety.
Wright A, Ai A, Ash J, et al. J Am Med Inform Assoc. 2018;25:496-506.
Clinical decision support (CDS) includes electronic alerts that can prevent errors. Excessive or erroneous alerts may lead to alert fatigue or other unintended consequences. Researchers used a blend of qualitative methods such as interviews and quantitative data like alert rates to develop a taxonomy of CDS alert errors. The taxonomy includes the origin of the error, which most commonly occurred with introduction of a new decision support rule, and describes the underlying reason for the error, such as problems with new terms, conceptualization, and building the rule as intended. Errors could cause an alert to fail to appear for a relevant situation or could cause an irrelevant or erroneous alert to appear. Most errors came to light through reports from users. The authors recommend classifying CDS alert errors using this taxonomy so that safety efforts will be consistent and actionable.
Cutrona SL, Fouayzi H, Burns L, et al. J Gen Intern Med. 2017;32:1210-1219.
Electronic health record alerts contribute to alert fatigue and increase provider workload. Some alerts are more time-sensitive than others and a delayed response can adversely impact patient safety. This study found that time-sensitive alerts were less likely to be opened by primary care providers within 24 hours if the provider's InBasket had a high number of notifications at the time of alert delivery or if the alert was sent on the weekend.
Lester PE, Rios-Rojas L, Islam S, et al. Drugs Aging. 2015;32:227-33.
Older patients are particularly vulnerable to medication errors, with certain high-risk medications accounting for a large proportion of adverse drug events in these patients. This study evaluated the effect of warnings within a computerized provider order entry (CPOE) system targeting prescribing of unsafe medications to patients aged 65 years and older. The warnings resulted in a significant decrease in prescribing of two of the three medications targeted over a 3-year period. The authors note that there were readily available, safer alternatives for those medications, but not for the drug which continued to be prescribed. Also, prescription rates of all three medications were unchanged in younger patients, indicating that the tailored nature of the alerts played a role in their effectiveness. While clinical decision support within CPOE does have some effect on safe prescribing, the use of computerized warnings of this type must be balanced against the very real possibility that alert fatigue may develop as a result.
Ahn EK, Cho S-Y, Shin D, et al. Healthc Inform Res. 2014;20:280-7.
Alert fatigue is a well-described limitation of clinical decision support systems. This qualitative study found alert overrides occurred most frequently in the emergency department, and the most common reason reported was that the alert was clinically irrelevant, emphasizing the need to tailor alerting systems for different clinical settings.
This before-and-after study found that introduction of a tiered alert system for medication dosages in pediatric patients led to an increase in alerts, but also resulted in fewer overridden alerts and more medication order revisions. This work emphasizes the need to improve electronic medication alerts to make them more actionable and reduce alert fatigue.
Clinicians use thousands of prescription medications during routine care, and new medications are regularly incorporated into practice. Confusion between medications with names that appear or sound similar is a common cause of medication errors. This observational study sought to determine whether a computerized provider order entry system—with an alert that prompted providers to enter the indication when certain medications were ordered and required users to click "OK" to ignore the alert, to add the drug to a problem list, or to cancel the order—identified drug name confusion errors. These alerts intercepted 1.4 drug name confusion errors per 1000 alerts. While authors recommend that these alerts be implemented to decrease medication errors, they suggest narrowing the number of medications selected to prompt alerts to reduce risk of alert fatigue. A previous AHRQ WebM&M commentary describes an incident involving a look-alike drug error and reviews strategies to enhance safety of medication selection.
Please select your preferred way to submit a case. Note that even if you have an account, you can still choose to submit a case as a guest. And if you do choose to submit as a logged-in user, your name will not be publicly associated with the case. Learn more information here.