**6. Alarm fatigue**

intending to check in after tending to her newly assigned patients. When she finally returned to the patient's room, she found him sprawled on the floor and writhing in pain. The patient, emboldened by his rapid recovery, had attempted to ambulate to the bathroom without assistance and lost his balance in the process. The intense pain prevented him from reaching the call button on the hospital bed, so he was forced to lie on the floor in pain for approximately 1 h. A subsequent skeletal survey revealed a left hip fracture, which required additional surgery, prolonged hospital stay, and the need for inpatient rehabilitation stay due to temporary disability involving bilateral lower extremities (e.g., right knee arthroplasty and left hip injury).

A 71-year-old male with a history of multiple myeloma was admitted to the urgent care center after noticing sudden onset of right lower extremity swelling associated with minor pain. The patient began induction therapy for multiple myeloma approximately 1 year prior, achieving adequate disease control. He was subsequently transitioned to maintenance treatment, which he continued for the past 6 months. Evaluation in the urgent care center with venous duplex studies revealed a deep venous thrombosis (DVT). Because of the patient's established history of malignancy, the triage clinician opted for hospital admission and therapeutic anticoagulation. While being transferred to the inpatient unit, unfractionated heparin anticoagulation was started. Per standard protocol, monitoring equipment was hastily fitted to the patient for noninvasive measurement of his blood pressure and heart rate. Overnight, the patient remained stable, with some resolution of lower extremity of pain despite persistent swelling. The on-call physician assessed the patient during morning rounds and ordered to repeat venous duplex for the afternoon to evaluate for resolution/progression of the DVT. Of note, throughout the night and into the morning hours, the patient's hemodynamic monitor had been sending intermittent alarm signals. With the first few alarms, the charge nurse promptly responded and quickly assessed the patient for any signs of instability or distress. However, as the shift progressed, the nurse increasingly dismissed repeated signals as "false alarms" due to a recurring pattern of mildly elevated blood pressure and heart rate secondary to episodic extremity pain. Because the inpatient unit continued to be understaffed during the morning shift, the charge nurse decided to disable the patient's repeated monitor alarms after the patient was assessed during morning rounds and found not to have any acute issues. It was hoped that this decision would eliminate the distraction of the nuisance alarms. However, during the patient's routine afternoon assessment, the rounding physician noted cold and diaphoretic extremities with markedly increased swelling. Interrogation of the monitor system revealed progressive bradycardia and hypotension over the past hour. An emergency CT angiogram showed a massive pulmonary embolism, prompting immediate thrombolytic therapy and patient transfer to intensive care. Despite aggressive management, the patient's

**5.3.** *Clinical Vignette #3***: 71-year-old male with history of multiple myeloma admitted for right lower extremity swelling associated with minor pain**

100 Vignettes in Patient Safety - Volume 4

shock became refractory, culminating in his death several hours later.

The three hypothetical clinical scenarios outlined above share a common theme: dedicated monitoring systems implemented to ensure early detection of clinical deterioration and thus

**5.4. Summation of Clinical Vignettes: finding common threads**

After the general introduction of AF earlier in the chapter, the authors will now discuss this important concept in greater detail. The phenomenon of AF is multifaceted and includes increased clinician response time with simultaneous decreased response rate that is mainly attributed to excessive stimuli from clinical alarms [8]. Depending on patient acuity and clinical monitoring requirements, typical bedside health-care personnel may be exposed to as many as 1000 alarms during a single shift, of which as many as 95% can be nonactionable and thus do not require immediate clinical determination [8, 66, 67]. Given the multitude of clinical alarms, a provider has to sort through during a typical hospital shift, there will be a natural tendency to potentially dismiss certain alarms as insignificant through rationalization. This phenomenon is described in the literature as the natural human behavioral reaction to "deprioritize signals" that have often been proven to be either false or misleading. Thus, staff may begin reflexively disabling or silencing alarm systems, which could effectively mask other alarms that may be clinically significant [68, 69]. To some extent, this behavioral pattern was seen in all three *Clinical Vignettes*, where the actionable alarm was masked by the vast number of nonactionable alarms that preceded it. Ultimately, the resulting delay in response or inadequate response puts patient safety at risk and may result in morbidity and/or mortality [70, 71]. Technologically advanced physiologic monitors bring a lot of promise, both in terms of earlier and more sensitive detection of patient deterioration (or other clinically significant event); however, the sensory overload and desensitization associated with AF will likely continue to present a major opportunity for improvement.

Certain other factors have been implicated in the increased incidence and severity of alarm fatigue, including greater staff workload, higher patient acuity, and the complexity of the modern health-care environment [10]. Nurses serve as key frontline staff in most clinical settings and play a pivotal role in overseeing patient care and monitoring. Moreover, nurses are subject to significant occupational stress that can be attributed to multiple causes, including heavy workloads [72]. This stress, as outlined in previous sections of this chapter, certainly influences AF by forcing nurses to instantaneously adjust their work activities (and priorities) according to perceived importance of near constant clinical alarm activity. Our *Clinical Vignette #2* illustrated the difficult task of ongoing patient triage, with the nurse having to prioritize between the three newly admitted patients and all of her other assigned patients. This constant need for clinical vigilance and prioritization is potentially disruptive to typical workflow, especially when high task complexity is involved. It can also contribute to the development of burnout [73]. Nurses have expressed the internal conflict between having to ignore the constant alarms simply to maintain sufficient focus to finish their routine tasks [74]. It is not surprising that increasing workload or task complexity has been associated with both suboptimal job performance and inconsistent alarm response [10]. Furthermore, the

Institute's Health Technology Hazards list [77, 78]. The subject of AF has been extensively studied, primarily due to its high prevalence across essentially all health-care settings. The underreporting of alarm-related events has been recognized as a challenge, and it should be noted that recorded incidents likely reflect only a small proportion of actual events. Available records from the Joint Commission's Sentinel Event Database show 98 alarmrelated occurrences between January 2009 and June 2012 (**Figure 4**). Of these reported events, several common alarm system issues (**Figure 5**) were directly connected to events leading to

Combating Alarm Fatigue: The Quest for More Accurate and Safer Clinical Monitoring Equipment

http://dx.doi.org/10.5772/intechopen.84783

103

**Figure 4.** Alarm-related events and subsequent results from January 2009 to June 2012 (source: Joint Commission's

**Figure 5.** Major contributing factors of alarm-related events (source: Joint Commission's Sentinel Event Database).

injury or death (**Table 1**) [79].

Sentinel Event Database).

**Figure 3.** The word cloud demonstrating the multifaceted phenomenon of alarm fatigue.

effort of acknowledging, evaluating, and responding to an alarm significantly increases the overall time commitment and workload of the nurses, which further perpetuates the trend of decreased alarm response and task performance [8].

Because multiple factors contribute to AF, many existing models struggle to fully account for (and address) clinician behavioral patterns seen with AF [75]. At the same time, it should be noted that AF is not unique to clinicians. In fact, a similar phenomenon has also been seen among human operators utilizing automated monitoring systems, such as aircraft pilots and nuclear power plant operators. The excessive number of alarm activations leads to the tendency of operators to ignore alerts, particularly when the monitoring system produces a high rate of false alarms or alerts [75]. For these operational environments, it has also been suggested that increased primary and secondary task workloads have a compounding effect on alarm response degradation that may occur in the setting of low alarm system reliability [76]. Similar to the clinical setting, AF can be associated with serious safety risks and represents a similar barrier to the practical application of automated monitoring systems in other fields (**Figure 3**).
