This issue's ECG Challenges column is devoted to a topic that affects every nurse who works in an acute or critical care unit. Hospital alarms can come from telephones, ventilators, heart monitors, intravenous pumps, antiembolism devices, sleep apnea devices, smoke alarms, patient-controlled analgesia pumps, pagers, cell phones, patient call systems, oxygen saturation devices, and fall prevention devices. Physiologic clinical alarms play a vital role in the information systems that acute and critical care nurses depend on for the daily care of their patients.

Although alarms serve as an indispensible part of modern hospital care, they can also be a detriment to patient comfort and affect ergonomic issues for nurses. Physiologic alarms are listed second among the top 10 technology hazards for 2011 by the ERCI Institute, a Pennsylvania patient safety organization.1 Alarm fatigue and misuse can lead to unintended consequences for patients and users. This column will review the use of clinical alarms and examine issues related to their effectiveness and safety.

Historical Considerations

The development of clinical alarms closely followed advances in the use of technology in hospitals. The early intensive care units of the 1950s used the relatively simple technology of the day (chest tubes, oxygen, metal tracheostomy tubes, sphygmomanometers), but electronic monitoring was not available.2 Hughes Day's3 original attempts in 1960 to monitor heart rhythms on general medical floors at Bethany Hospital in Kansas City, Kansas, included an alarm to notify the staff of cardiac arrest. Unfortunately, resuscitation outcomes were poor, because the staff did not know how to respond appropriately to the alarms. Day's4 original dedicated cardiac care unit (opened in 1962) used oscilloscopes and a central alarm box, with subsequent improved outcomes.

Fairman and Kagan5 described an incident that occurred at Hospital of the University of Pennsylvania in the mid-1960s that underscores the early importance of reliable alarm systems when using technology:

In November 1964, an unattended postoperative patient attached to a Bird respirator died. The patient became disconnected from the respirator, and because the machines lacked alarms [emphasis added] and the patient was not readily observed, the nurses were unaware of the patient's crisis and responded too late.5^(p71)

Such incidents spurred the development of complex alert systems to notify staff of equipment malfunctions or changes in patient conditions.

Components and Functions of Alarm Systems

Physiologic alarms are components of devices that monitor some aspect of human physiology. The 3 most common physiologic monitors are cardiac, ventilator, and pulse oxygenation. Physiologic alarms are defined as "any alarm that is designed to protect the individual receiving care or alert the staff that the individual is at increased risk and needs immediate assistance."6^{(p 317)} An alarm system is composed of the individual or parameter monitored, the monitoring device, and the practitioner who has responsibility for responding to the alarm.

The ECRI Institute has compiled a list of specific criteria that should be present when initiating devices that use clinical alarms.7 First, alarms should be initialized as soon as the device is connected to the patient or after a brief warm-up period. For identification purposes, alarms that indicate a significant risk to the patient should be both audible and visible. In addition, the cause of the alarms must be easily identifiable.

Prioritization of alarms requires that those indicating immediate life-threatening events override less urgent alarms. An example of this is a noncritical "lead off" alert in a 5 lead cardiac monitoring system. A ventricular tachycardia alarm should override the noncritical alert. Visible and audible signals for critical alarms must be easy to differentiate from less urgent alerts. Active tiered alarm systems that provide escalating alerts (when initial alerts remain unanswered) are preferred over passive alarm systems.

Finally, there should be specific criteria for disabling alarms. Critical alarms, such as those for ventricular fibrillation, should not have the capability of being disabled. Audibility features of high-risk alarms may be disabled, but the visible alerts should remain active until the problem has been resolved. If any alarm is disabled, there should be a prominent visible alert present so that the practitioner is aware of the situation.

Human Oversight

Most alarm systems present signals at a central monitoring station. Either these monitoring stations are manned with medical technicians, or they can be located in an area where any staff member can hear the signal, such as a central nursing station. The limited research on the efficacy of manned monitoring stations is inconclusive and there are pros and cons to the use of monitor technicians.8 Automatic pager alert systems and slave monitors at strategic locations on a unit serve as other methods to ensure audibility of alarms in geographically remote areas of the hospital or care area.

Alarm enhancement technologies can improve the effectiveness of alarms by increasing the audibility or visibility of the alarm interface with the practitioner.9 These technologies include auxiliary displays, bed-to-bed notification systems, individual notification devices, and broadcasting devices. Issues related to these auxiliary devices include notification of the correct responsible clinician, facilitation of feedback to confirm alarms are addressed, cost issues, and the noise level in the unit.

Alarm Fatigue

Alarm fatigue is a type of human error that occurs when a practitioner is desensitized to alarms and alerts.10 According to a landmark Institute of Medicine report in 1999, human error is a major cause of hospital deaths.11 The report recommended that systems be used to help prevent the effects of human error. Alarms themselves are systems; therefore, alarm fatigue is an unintended consequence of clinical alert systems. Factors that contribute to alarm fatigue include a high number of false alarms or nuisance alarms, practitioner inattention, and intentional or inadvertent deactivation of engineering safeguards.12

The dangers of alarm fatigue have been known for some time. The Joint Commission included clinical alarms on its list of active national patient safety goals in 2004.6 According to the US Food and Drug Administration, 237 deaths related to clinical alarms were reported between 2002 and 2004.9 Half of the deaths were related to operator error; these errors involved education and training, workplace problems, and personal problems.

The issue has taken on a new urgency since the Boston Globe published a front-page article about an incident that occurred at a Massachusetts hospital in February 2010.13 A patient showed warning signs of extremely low heart rate, but none of the staff members reported hearing alarms. The patient subsequently died, despite efforts to resuscitate. Alarm fatigue was targeted as a contributing factor in the death.

An investigation of the incident showed that the patient showed signs of a problem on the cardiac monitor for a full 20 minutes before collapse.14 It was also found that the patient's bedside monitors had been disabled. In addition, the system had easily disabled alarm controls that could be inadvertently changed at the bedside.

Providing Safe Passage With Improved Alarm Use

The ECRI Institute has developed a comprehensive list of criteria to make alarms safer.7 Manufacturers can use the criteria when developing life support equipment, and hospitals can use them when making decisions related to the purchase of new monitoring equipment. The criteria include specific recommendations on disabling alarms. Alarms should not be allowed to be defeated unless there is an automatic reenabling function. A prominent visual indicator of a silenced alarm should be displayed along with a periodic audible indicator. If a secondary condition occurs while the initial alarm is silenced, a secondary alarm should be displayed automatically.

Human factors also can be manipulated to reduce the incidence of alarm fatigue and other alarm hazards. False cardiac monitoring alarms can be reduced by providing careful skin preparation and making sure that electrodes are fresh and properly applied to the skin.8 Patients who no longer need cardiac monitoring should be promptly removed from the device. Alarm audibility should be ensured by testing remote corners of the care area for sound transmission. Unnecessary alarms and hospital noises should be reduced or eliminated. Alarm systems themselves should be fine-tuned by biomedical engineering to reduce the number of false alarms and nuisance alarms. Finally, a mechanism for reporting problems should be easily accessible and user friendly for all staff members.9

Graham and Cvach12 reported that they were able to reduce the total number of alarms in a medical progressive care unit by 43% through human factors, engineering, and fine- tuning of alarm parameters. They used a well-designed quality improvement initiative combined with small tests of change to improve overall alarm management. Their method can be used in both progressive care and intensive care units to reduce nuisance alarms and help prevent alarm fatigue.

"Smart alarms" have been proposed as a solution to alarm fatigue because of their ability to reduce the total number of alarms.12 Smart alarms use multiple physiological parameters and expert systems to limit alarms to clinically relevant scenarios. An experimental smart alarm program from Penn Engineering samples 4 separate physiological alarm parameters and uses an algorithm-based Fuzzy Expert System to make decisions similar to those that would be made by an acute or critical care nurse.15 Such systems are a promising solution to alarm fatigue, although they have so far been restricted to specific groups, such as cardiac surgery patients.

Cardiac monitor alarms are a ubiquitous fact of life in modern acute and critical care. They are a necessary tool to promote safe passage for our critically and acutely ill patients. Judicious use of technology with appropriate clinical oversight and attention to human factors can ensure that our monitored patients get the best care possible.

REFERENCES