In 2010, a 60-year-old man died in the intensive care unit of Massachusetts General Hospital. He died not from the injury to his head from a fallen tree branch but from a delayed response to alarms indicating a drastic change in his condition. Around 5:00 PM that day, alarms began sounding that warned clinicians of tachycardia, tachypnea, and low blood oxygen levels. It was not until 6:00 PM that the critical apnea alarm sounded, indicating the patient had stopped breathing. Emergency personnel were called to the scene, but the damage had been done: The patient was unresponsive and a computed tomography scan later showed an anoxic brain injury. The family withdrew life support several days later after no improvement in condition.1 This sentinel event presented tangible evidence of a problem with alarm use.
Alarms are especially prevalent in critical care today, as the use of technology becomes ubiquitous. Monitoring devices warn staff of unsafe physiologic conditions. For units with telemetry monitoring, the average number of alarms per patient day can average 946.2 This translates to thousands of alarm signals on every unit and tens of thousands throughout the hospital.3 Experts estimate that between 85% and 99% of alarm signals do not warrant clinical intervention. This can happen as a result of parameters being set too tight, default settings not being modified for the individual patient, electrocardiographic (ECG) electrodes drying out, or sensors being misplaced.4 The problem with an increasing reliance on physiological monitoring technology in hospitals and especially critical care is that the sheer number of alarms can overwhelm clinicians and nurses, causing them to ignore or disable them. These "nuisance" alarms do not elicit an actionable response and therefore create an environment where important alarms are either silenced or ignored. This concept is known as alarm fatigue.
To prevent further complications with alarm management, hospitals and staff need to work to reduce the total number of alarms. Fortunately, there are unit-based interventions and systems-based policies that are available to combat alarm fatigue. This article will review current recommendations of The Joint Commission and the Emergency Care Research Institute (ECRI), while also assessing the research and interdisciplinary collaboration being done to mitigate the effects of excessive alarms. The goal is to inform staff nurses and management about the issue and to provide tools to intervene.
PREVALENCE OF ALARMS
Gone are the days when nurses have only their senses and a stethoscope to assess a patient. In critical care today, technology enables a multitude of assessment tools in real time. Use of these medical devices has grown rapidly in the last 3 decades, bringing the number of different alarms from 6 in 1983 to more than 40 in 2011.5,6Table 1 features common alarms that contribute to alarm overload in hospitals. To complicate matters, there is no standardization of alarm sounds and their decibel levels. A nurse floating to a different floor or switching hospitals must relearn the sounds or react according to their perceived importance. Monitor alarms were implemented with patient safety in mind but now contribute to an increasingly noisy environment and a false sense of security for hospital personnel.
A lack of policies and standardization is missing from hospitals, but to understand alarm overload, the root causes need to be identified. The real problem lies in the cause of excess alarms and the nurse's response to the cacophony of beeps. The habituation of alarm signaling occurs because the number of alarms can reach an inappropriate and ineffective level. There are 2 types of alarms at fault for this barrage of noise: false alarms and nonactionable alarms. False alarms are usually generated by patient motion, poor sensor placement, and limitations in the device alarm detection algorithm. In an attempt to cut costs, hospitals purchase reprocessed sensors, which may result in a higher incidence of false alarms.4 The other dangerous type of alarm is that which does not require clinical intervention. These nonactionable alarms distract clinicians' attention needlessly and add to an environment of constant stimulation. Two very common examples of these short-term alarms are low oxygen saturation and tachycardia. Both of these alarms occur frequently and are self-correcting as the alarming vital sign returns to the normal range.7
In an environment where such nuisance alarms are constantly sounding, nurses respond to these alarms in accordance with what they deem urgent. The American College of Clinical Engineers Healthcare Technology Foundation conducted a national online survey of clinicians, engineers, technical staff, and managers regarding the effects of alarms. Most respondents agreed or strongly agreed that alarms activate too frequently (81%), disrupt patient care (77%), and cause health care personnel to distrust their significance and disable them (78%).8 Bitan and associates9 found that nurses respond to alarms based on their "probability match." This means that if a nurse sees an alarm as 10% reliable, the response rate will be about 10%. They also found that nurses tend to react more to alarms if they are considered rare or exist for a longer duration.9 Further complicating alarm management, nurses are inadvertently affected by increases in workload, patient acuity, and task complexity.10
IMPLICATIONS FOR THE PATIENT
The aforementioned cause-and-effect relationship between excessive alarms and the nurses' response contributes to the concept known as alarm fatigue. This is a multifaceted problem that has real consequences for both nurses and patients. The eventual result of alarm desensitization and habituation are unsafe practices that affect patient safety. When a nurse deems an alarm no longer effective or aggravating, the natural reaction is to silence, delay, change the parameters, or turn it off entirely.8 A "cry wolf" scenario is instigated and alarms go unnoticed or even ignored because of their perceived importance. As evidenced by recent sentinel events in health care, the end result costs lives. The ECRI has placed alarm fatigue atop their 10 health technology hazards for 2012 and rates it as a national problem.11 The number of sentinel events has become a growing concern, as The Joint Commission's Sentinel Event database includes reports of 98 events related to alarms between January 2009 and June 2012. Eighty of those reports resulted in death (2013).3
Alarms do affect patient safety, but even more visible is their affect on patient satisfaction. Most alarms per patient day come from the cardiac monitor in each patient's room. These alarms, although sometimes heard at a central monitoring station or a nurse's station, also sound in a patient's room. Cardiac monitors usually feature 4 different alert tones based on the importance of the issue. These alarm levels are, in order of increasing importance, message, advisory, warning, and crisis.2 Whether it is a simple message notifying staff that a lead is off or a crisis alarm warning of a rhythm that has converted to ventricular tachycardia, the patient usually cannot distinguish between the two. Furthermore, physiological alarms exist to inform staff of their patient's overall health. Without medical training, these alarms in patient rooms affect sleep patterns and cause undue stress and anxiety for patients. This in turn thwarts recovery and leads to extended patient stay. A study by Hweidi12 highlights the effects of alarms on patient satisfaction. These alarms are stressors and inhibit patient's psychological health and well-being.12
GUIDELINES AND IMPLICATIONS FOR NURSING
Suffice to say that alarm fatigue is very prevalent in health care today. The increase in technology has led to the sacrifice of specificity for sensitivity. Fortunately, there are interventions to reduce alarm overload that nurses and hospital staff can implement. The central tenet of alarm fatigue reduction is alarm management. Proper management of alarms will lead to an overall reduction in alarms, which will, in turn, mitigate the effects of alarm overload on nurses and patients. This should be done using evidence-based practice and requires input from clinicians, the medical devices industry, and hospitals.9 Following are the highlights from current research and recommendations from organizations such as the ECRI and The Joint Commission. Table 2 suggests interventions for each of the problems associated with alarm fatigue based on these sources. The table echoes the Practice Alert put out by the American Association of Critical Care Nurses in April 2013.13
The first facet of alarm fatigue to tackle is the prevalence of nonactionable alarms. These alarms include all alert messages that do not require action and are therefore unnecessary. In a quality improvement project done by Kelly Creighton Graham and Maria Cvach, simple adjustments to the physiological monitors in every patient room decreased overall alarms per patient day by 43%. The most influential intervention they performed was changing some of the default alarm settings. The low limit for heart rate was changed from 60 to 50 bpm, and the high limit, from 120 to 150 bpm. The oxygen saturation low limit was then changed from 90% to 88%. This change was based on a study by James Welch in 2011 that found low limit parameter changes of 90% to 88% to reduce pulse oximetry alarms by 70%.8 Finally, the number of premature ventricular contractions high limit was changed from 6 to 10. The intent of these changes was to make the alarms sound when nursing staff would typically intervene.
Another key aspect of the project was the attention to nurse education initially and ongoing throughout the implementation process. The focus should be put on teaching nurses about alarm management and more specifically the customization of alarm parameters to meet the physiological needs of the patient. Engineering departments and representatives from monitor vendors should be consulted for the best ways to make changes.4 Through education, nurses will be equipped with the necessary tools to modify parameters safely and effectively, again reducing nonactionable alarms.
In addition to interventions to reduce nonactionable alarms, there are methods for minimizing the amount of false alarms. These alarms are produced incorrectly and do not accurately represent the physiological state of patients. It is while receiving care or during activity that incorrect rhythms appear because of artifact. The resulting alarms are false and therefore warrant a pause in alarm function during certain activities. To ensure good conductivity for ECG electrodes, it is recommended that the skin be prepared with soap and water and wiped with a rough washcloth or gauze.14 Alcohol will inadvertently dry out the skin. Changing ECG electrodes daily has also been found to reduce false alarms.15 Biomedical engineering departments can help by routinely testing and inspecting medical devices.4 Pulse oximetry monitoring is a major contributor to false alarms, as blood oxygen saturation fluctuates in states of low perfusion and increased motion. Welch recommends using the newest technology to capture SpO2 levels accurately. He also endorses the customization of delays and parameters to reduce false alarms.
Finally, to promote alarm management on a systems level, an emphasis needs to be put on the development of policies and procedures. Interprofessional teams with members of the clinical, technical, and administrative communities will address policy development and direction for nurses and other staff. Graham and Cvach2 used an Alarm Management Task Force to implement default changes, unit-based education, and new policies. In June, The Joint Commission put out a National Patient Safety Goal on Alarm Management that includes several elements of performance. It challenges hospitals to make alarm safety a priority and to establish policies and procedures for managing alarms. They recommend that hospitals first inventory all alarms used and then to identify the most important to manage. For these alarms, the policies and procedures must address when alarms can be suspended, when they can be changed, and who has the authority to make the changes.15 Alarm management at the systems level will ultimately provide the foundation for combating alarm fatigue.
CONCLUSION
With an average of 946 alarms per patient day on some units in the hospital, nurses cannot possibly respond to them all.2 In fact, nurses are not expected to intervene for every single alarm warning of changing physiological parameters. With so much noise, a clinician's ability to distinguish between the actionable and the nonactionable or false alarms is hindered. Excessive alarms undermine the original purpose of monitoring technology to increase patient safety. Although nurses are no doubt affected by alarms in practice, patients are exposed to a whole host of alarms in their own rooms. This creates an atmosphere of stress through constant stimulation and thus impedes the healing process. It is mutually beneficial for both patients and staff to reduce alarms.
Input from all parties involved is necessary to combat the multifaceted problem of alarm fatigue. At a systems level, management and other administrative members need to devise strategies for developing policies and procedures that take into account the needs of nurses and the knowledge of vendor representatives. These regulations should be appropriate for each individual area of the hospital and likewise focus on the most prevalent and worrisome alarms. To foster awareness and to promote solutions, nurse education will play a vital role. Nurses will learn many alarm management tools based on current research and clinical guidelines.
Although many evidence-based recommendations exist, there is still a lack of experimental research and clinical studies pertaining to alarms. Reports like that of Graham and Cvach2 prove that it is possible to drastically reduce alarm numbers by enforcing small changes, but the amount of clinical data based on patient outcomes is insufficient. Also, much of the information presented in Table 2 is theory-based evidence from expert opinions or peer-reviewed professional organizational standards.13 Alarm management is moving in the right direction, but still more is needed to recognize alarm fatigue and standardize its reduction.
References