by Anna Ver Hage, RN, MSN, CCRN, CNRN
A 60-year-old-man was admitted to the ICU after a traumatic brain injury. He was agitated and constantly removed his pulse oximetry device, which caused a shrilling alarm to ring continuously throughout the day. The patient received lorazepam (Ativan) to calm his restlessness, and a few minutes later his alarms began to sound, warning that the patient was experiencing a drop in oxygenation, tachycardia and tachypnea. Reports state that no one responded to these alarms, most likely thinking they were erroneous alarms as they had been all day. About one hour later, a high-alert alarm rang warning of respiratory arrest. A CT scan later showed an anoxic brain injury, and the family withdrew life support a few days later.1
Beeps, bells and soft, loud and ear-piercing alarms are all associated with patient care areas, but what if all this noise is not helping patients but actually harming them? The above scenario portrays a true clinical situation that has become more frequent across the nation. As hospitals continue to become more technologically advanced, the problems associated with technology grow. Alarm fatigue is among these problems and has led to sentinel events, unexpected events in healthcare that lead to death or serious physical or psychological injury or the risk of them. This sentinel event in 2010 prompted a nationwide focus on medical device alarm safety in hospitals, shedding light on the fatal consequences associated with alarm fatigue.2
Although medical device alarms are intended to promote patient safety, a growing concern exists that instead these alarms are contributing to patient harm. With more than 350 alarms ringing per patient in a 24-hour period, equating to thousands of alarms in a single critical care unit and tens of thousands alarms throughout the hospital in one day, there is little doubt that alarm fatigue is prevalent in busy hospital units throughout the United States.2,3 Many medical devices in hospitals have audible alarms and alerts. (Alarms indicate a clinical intervention may be necessary while alerts inform the user that an infusion is complete or the battery is low.) These devices include electronic blood pressure cuffs, pulse oximetry, telemetry, infusion pumps, mechanical ventilators, temperature probes and central station monitors. Over time, clinicians become accustomed to hearing the multitude of alarms and tend to become desensitized, which may lead to a lack of or a delay in response.2 This phenomenon has been termed “alarm fatigue.”2
Alarm fatigue is receiving national attention because of reports of sentinel events related to clinicians’ becoming desensitized by the high number of device alarms. Desensitization leads to delayed response times, and this practice can harm patients. Alarm fatigue has been associated with staff frustration, a delay in response to alarms and poor patient outcomes.2
In 2012, The Joint Commission (TJC) issued a sentinel event alert for 98 alarm-related incidents between January 2009 and June 2012.2,3 Of these reported incidents, death occurred in 80.2,3 Since reporting of most sentinel events to TJC is voluntary, the organization estimates these 98 alarm-related incidents account for just about 10% of the actual cases in the United States.2
TJC is not the only organization concerned with this patient safety issue. ECRI Institute, the Association for the Advancement of Medical Instrumentation (AAMI) and the U.S. Food and Drug Administration have also voiced concerns. The FDA’s Manufacturer and User Facility Device Experience reported 566 alarm-related patient deaths between January 2005 and June 2010.4
Each year ECRI, a nonprofit organization that uses research to establish best practices for improving safe patient care, releases a top 10 list of technology safety hazards. Over the past few years, ECRI has named “alarm hazards” the No. 1 health technology safety hazard, and it continues to remain at the top of the list for 2014.5 This is not a new phenomenon. For many years, studies have reported problems with excessive alarms, leading to both a delay in response time and inappropriate disabling of alarms.4
Monitoring devices are an essential part of providing optimal patient care. These devices provide the data to drive treatment and guide clinical decision making. An alarm is intended to get the attention of clinicians and alert them to a change in patient status, a life-threatening event, a potential malfunction or an unsafe situation.6 An alarm may be classified as a crisis alarm, warning alarm, system failure or advisory alarm and depending on the severity levels may be associated with a specific noise or volume. For example, a crisis alarm may be louder and higher pitched than an advisory alarm or alert indicating a low battery on an infusion pump. “Nuisance alarms” and “false alarms” are terms used to describe alarms that do not require an intervention. Nuisance alarms or alerts are not necessarily related to a patient condition but may indicate a low battery or a machine that needs to be plugged in. False alarms are related to a patient condition and would indicate a need for a response if they were real. Most false alarms are related to artifact, tight setting of parameters or a sensor that fell off.
Unfortunately, with the many nuisance and false alarms daily, those high-priority alarms may elicit the same clinician response as a low-priority alarm. The next time you are at work, listen carefully to the alarms in your unit and see if you can identify the sound that each machine is making.
Attention must focus on how to minimize and optimally prevent alarm fatigue. Research has shown that most alarms have no clinical significance. Reports estimate as many as 85% to 99% of alarm signals are false or not clinically significant and, therefore, do not require any medical interventions.2 This means that as few as 1% of all alarms ringing daily actually require an intervention.
The ‘Cry Wolf Syndrome’
Many factors contribute to alarm fatigue, including alarm limits that are set too tight, false or nuisance alarms, default settings that are not adjusted for an individual patient or patient population and electrodes that have lost conductivity or electrodes that are inappropriately placed.2 Over time, these false alarms reduce the credibility of the alarm and increase response time, leading to a potentially catastrophic event. A majority of the reported events occurred in telemetry, critical care, general medicine and EDs.2 The main contributing factors in these areas include:2
• Absent or inadequate alarm systems
• Improper alarm settings
• Alarms signals not audible in the appropriate areas
• Alarm signals inappropriately turned off
TJC expects all its certified hospitals to address alarm safety and develop a plan to improve this patient safety issue. The 2014 TJC National Patient Safety Goal (NPSG) on alarm management is in a two-phase rollout with Phase 1 having begun in January 2014 and Phase 2 beginning in January 2016.7
Phase 1 requires hospitals to make alarm management a priority, look at their alarm-related incidents and determine alarm management needs. Phase 1 involves gathering data and asking questions of the clinical staff. Discussions should include the alarms that are clinically necessary, the alarms that may be contributing to fatigue among clinicians and an assessment of the potential for patient harm.7 Questions to discuss include whether patients are at risk from erroneous alarms contributing to a delay in response times. This institutionwide approach is best done by creating an interprofessional taskforce representing major stakeholders, such as management, nursing and device manufacturers.
During Phase 2, hospitals assess the data collected in Phase 1 and based on the institutions’ individual patient safety concerns develop and implement policies and procedures that address alarm management.7 These concerns may include the management of inappropriate settings based on the individual patient or patient populations.7 This may also require defining clinical situations in which it is appropriate to disable alarms and who is qualified to disable or adjust parameters.7 Policies should address recommendations for checking that accurate settings have been set and how and when to monitor the device alarms to ensure proper functioning.2,7
A key to the success of Phase 2 involves educating healthcare clinicians about the monitoring devices and alarms in their clinical arenas and setting expectations for the management of alarms for which they are responsible.7 A task force should be formed to carry out solutions to the issues that were collected in Phase 1. In addition to education and policy revision, solutions should include updating outdated or malfunctioning equipment.7
Roundtable discussions, such as the one held at the 2013 AAMI annual meeting, focus on what actions should be taken to minimize insignificant alarms and improve response time to the clinically significant alarms. The American Association of Critical-Care Nurses’ Practice Alert on Alarm Management was an essential resource during this roundtable discussion as AACN has offered strategies to approach alarm management. AACN emphasizes the need for staff education on ECG monitoring, customization of alarm parameters and a determination of who meets eligibility criteria for monitoring and recommends these strategies be implemented through an interdisciplinary approach.8
TJC, AAMI, ECRI and AACN offer the following recommendations to minimize alarm fatigue:
The epidermis is a poor conductor of electricity; therefore, the evidence supports performing a skin preparation before placing the ECG electrodes.8 Simple interventions that will enhance electrode conductivity include washing the area with soap and water to rid the skin of oils that may interfere with conductivity, roughening the epidermal surface by wiping with an abrasive surface (gauze or a dry washcloth) and cutting excess body hair.8,9 These interventions may take only a few seconds, but evidence suggests that by preparing the skin, artifact is reduced, equating to less time spent on managing false alarms.8
• Daily maintenance: Evidence also suggests that changing electrodes daily in addition to when they appear damaged or are no longer intact optimizes conductivity and enhances electrode contact with the skin.8 The outer layer of the epidermis, the stratum corneum, can regenerate within 24 hours, so best practice is to change and repeat skin preparation daily.8,9
• Get it right the first time: Proper placement of ECG electrodes is essential for an accurate signal. Placing electrodes on the fleshy parts of the arms and legs and avoiding bony areas or muscle groups minimize the artifact that commonly occurs during movement.9 By pressing firmly around the outer area of the electrode during application, the gel center remains intact. This practice will reduce the pockets of air that can develop if the gel center is disrupted and could cause artifact.9
One quality-improvement project in a medical progressive care unit made slight adjustments in the alarm parameter settings so that when an alarm sounded, it was because action was required or the alarm was clinically significant.10 High heart rate alarms normally set at 120 were increased to 150 bpm and low heart rate parameters were lowered from 60 bpm to 50 bpm.10 Oxygen saturation percentage parameters were adjusted from a low SpO2 of 90% to 88%.10 Education was provided to the nursing staff. Nurses were taught to be proactive when setting alarm parameters instead of reacting to the alarms’ being triggered. Rationales for individualizing patient alarms were provided.10 These simple changes resulted in a 43% reduction in critical alarms over an 18-day period.10 After the interventions were put into place, the number of critical alarms was 9,647, significantly less than the 16,953 alarms preintervention.10 The low oxygen saturation alarm sounded 1,685 times before the inventions, and 623 afterward.10
Recommendations from this quality-improvement project included:10
• Staff should be trained to appropriately set alarm parameters based on their patients’ needs.
• Duplicate alarms should be avoided, such as setting both a high heart rate and a tachycardia alarm.
• Alarm parameters should be set to limits that require a clinical intervention.
To each his own: Each patient is unique with specific medical needs depending on a variety of factors, including diagnosis, medical history and plan of care. Default parameters may not be appropriate for each patient and, in fact, may vary greatly from one patient to the next. Alarms need to be customized to meet the individual needs of the patient. AACN recommends customizing the alarms within the first hour of taking care of a patient and then individualizing these alarms as needed and when changes occur.8 Tailoring alarm parameters to the individual patient or to the specific patient population will minimize alarms.
Turn it down … or up: Alarm noises must be audible, and monitors should be seen throughout the patient care areas.8 Alarms must be heard over the other noises on the hospital unit, and monitor screens must be accessible to staff when staff are away from the patient’s room. This encompasses all high-alert alarms, including the mechanical ventilator and fall-risk alarms.
Studies have demonstrated that most alarms are related to pulse oximetry (Sp02) monitoring.8 SpO2 sensors are typically placed on the fingertips, so patients who are moving can create artifact and noise. Using adhesive sensors will minimize some of this artifact; they are considered more accurate in patients with decreased perfusion or who are moving.8 AACN recommends using disposable adhesive sensors and replacing them as needed.8
Turning the high SpO2 alarm off is clinically indicated in most patients as an oxygen saturation of 100% does not require an intervention. Customizing the delay and threshold settings to the patient can significantly reduce erroneous alarms.8 This may require communicating with biomedical engineering to help make these changes.
A Plan for Action
Implementing interventions to reduce alarm fatigue requires proper education about the monitors. Training should include the organization’s own policies and procedures for managing alarms and should be performed with all new staff and then on a regular basis.2 AACN recommends educating nurses on the importance of suspending alarms before performing care, such as drawing labs from an arterial line, suctioning, performing chest physiotherapy or turning a patient.8 Patients may trigger an irregular heart rate alarm just by moving around in bed. Movements such as scratching an electrode can trigger a ventricular tachycardia alarm. Therefore, education is necessary to teach nurses how to differentiate true alarms from false alarms.
One retrospective study reviewed data stored in the central monitor data base to analyze what factors contributed to the high number of false and nuisance alarms. The pulse oximetry alarm appeared to be a main offender. Reducing the low SpO2 alarm from 90% to 88% and instituting a 15-second alarm delay led to a decrease in the number of low SpO2 alarms by more than 80%.11
Really? Is it Necessary?
Patients for whom monitoring is indicated should be monitored. If no clinical indication exists for monitoring, monitoring is unnecessary and would only contribute to the number of alarm noises.8 For example, a patient with bradycardia at baseline should not have the low heart rate alarm limit set to 60 beats per minute if the baseline heart rate is usually less than 60 bpm. If the patient has an irregular heart rate due to atrial fibrillation, neither the irregular heart rate nor the atrial fibrillation alarm should be enabled as it will not require any clinical intervention. If the patient is having frequent premature ventricular contractions at baseline, increasing the number of PVCs per minute will be an effective alarm reducer. Monitor only what needs monitoring and don’t turn off alarms that are clinically necessary. Patient safety is the priority.
Policies and procedures should be developed to address alarm management, establish guidelines for alarm settings and offer suggestions for how to tailor alarms to individual patient needs.2 In addition, an institution’s expectations for performing maintenance on device monitors should be addressed.2,8 Policies should include:2
• Proper skin preparation before electrode placement
• Proper placement of electrodes
• Recommendations for changing electrodes
• Individualization of patient alarms each shift
• Assessment of alarm audibility
• Proper documentation of alarm settings
• Maintenance of the device monitors
The American Heart Association (AHA) 2004 Practice Standards for ECG Monitoring in Hospital Settings offers indications, clinical situations and time frames for when ECG monitoring is necessary and outlines best practices for hospital-based ECG monitoring.12 Recommendations include strategies to improve diagnostic accuracy when monitoring for cardiac dysrhythmia, ischemia and QT intervals.12 Hospitals should use these practice standards when developing policies and procedures on alarm management.
Back to the Evidence
High-quality control trials continue to be needed on alarm fatigue. At this time, randomized control trials and meta-analysis are lacking. The recommendations available are based mostly on expert opinion; quality-improvement initiatives; recommendations for best practice from safety organizations, such as TJC, ECRI and AAMI; and a few small-scale observational studies in critical care areas and progressive care units. Although there is a lack of quality data to support the recommended interventions, the literature does support the need to have recommendations in place and to further evaluate practice and standards for care.
The Healthcare Technology Foundation (HTF) conducted a national survey on clinical alarm issues, first in 2005 and then in 2011, to discover changes in perceptions and reassess the problem after improvements were made. There was little difference between survey results from 2011 and 2005. In 2005, 95% of respondents agreed that alarm sounds and visual displays should differentiate the priority of an alarm while in 2011 96% agreed.13 In both surveys, 78% of respondents felt that nuisance alarms reduce trust in alarms and cause caregivers to inappropriately turn alarms off.13 A high percentage reported that nuisance alarms disrupt patient care.13 Twenty percent of respondents reported adverse events related to alarm issues.13 (These events may or may not have been reported to the FDA, which stresses the importance of alarm management reform.13)
One recommendation from the study focused on the need for improved central alarm management, with many hospitals reporting they have designated monitor watchers.13 The HTF recommends hospitals consider using monitor watchers for managing alarm fatigue.13
The HTF surveys have many recommendations for alarm management. One implication from the studies revealed the need to use a systems approach: opening lines of communication and developing solutions related to alarm management issues.13 Other strategies focus on the need for placing a high priority on ways to reduce nuisance alarms so desensitization to the noise will be minimized, improving response times to alarms and correcting the practice of ignoring or inappropriately disabling alarms.13
The Practical Use of the Latest Standards for Electrocardiography (PULSE) Trial is a five-year multisite randomized clinical trial under way to evaluate nurses’ knowledge about ECG monitoring and the effect of implementing AHA Practice Standards for ECG Monitoring.14 The focus is to improve the training of nurses who work with ECG monitoring and improve quality of care and patient outcomes when detecting and diagnosing dysrhythmias, ischemia and QT analysis.14
The first phase of the PULSE Trial assessed patients in cardiac units at 17 hospitals with 4,678 patients enrolled. Of those enrolled, only 295 patients were not on ECG monitors, and 26% of the patients being monitored lacked any indication to support the need for ECG monitoring.14 The preliminary data gathered indicated there was a lack of knowledge among nurses and demonstrated a need for further education on ECG monitoring, specifically ischemia monitoring.14 Of patients with an indication for ischemia monitoring, only 35% were being monitored.14 The PULSE Trial also will assess the benefits of using an online ECG monitoring education program.14
Continued research and the use of evidence-based practices to develop standards of care are main priorities in finding solutions to this crucial problem. Individual organizations need to reflect on their own issues and develop practical solutions to deal with their needs.
Alarm fatigue is receiving national attention. This multifactorial problem with numerous dimensions has a huge impact on patient safety and patient outcomes. Nurses are the key to suggesting strategies, conducting trials of interventions and discovering solutions to the problem, all to help reduce the negative consequences associated with alarm fatigue and provide safe, quality patient care.
Anna Ver Hage, RN, MSN, CCRN, CNRN, has authored numerous articles along with coauthoring national guidelines on the care of neuroscience patients. She is an RN case manager and a clinical nurse in the neurosurgical critical care unit at University of Colorado Hospital in Aurora.