**2. Primary research methods**

on clinical indications and appropriateness of use for the existing equipment and monitoring frameworks [3–6]. Beyond the concept of "false alarm," suboptimal implementation of clinical monitoring systems can have much more profound and potentially dangerous consequences [7–9]. One such consequence, and the primary topic of this chapter, is the phenomenon of alarm fatigue (AF). It is defined as the decrease of clinician response caused by excessive alarms, sensory overload, and desensitization, in addition to other occupational and environmental variables [9–11]. Among contributing factors are also high staff workload, long shift hours, and work environments with high noise levels, all of which contribute to the

Hospital patient care units tend to be high-paced and potentially unpredictable environments, with complex workflows. Multiple simultaneous interactions between patients, families, and health-care staff may create an added element of chaos [13, 14]. To help nurses and other staff cope with their many responsibilities, various audible and visual alerts have been implemented to prompt immediate response and clinical assessment of patients [15]. These alerts are relayed from patient monitoring devices, which provide continuous flow of vital sign data with a high degree of sensitivity. The advanced technology used in these surveillance systems has provided a significant amount of physiological data at low cost while being particularly helpful by facilitating the monitoring of critically ill patients to identify deviations of vital signs (e.g., heart rate, respiratory rate, blood pressure, and pulse oximetry) from normal ranges [16]. However, when various clinical alarm systems are superimposed on the need for constant vigilance in the setting of highly challenging and often chaotic environment of the typical clinical unit, the stage is set for the emergence of AF and other forms of cognitive

The prevalence of various monitoring modalities has increased significantly, with most healthcare institutions utilizing some broadly defined combination of different alarm systems. As the use of these systems became more widespread, a major flaw became evident: the excessive amount of triggered alarms was contributing to unintended consequences, both in terms of patient outcomes and staff fatigue/dissatisfaction [8, 20, 21]. The high rate of nonactionable alarms, where immediate action is not required on the behalf of clinicians, was especially problematic [22]. In fact, the increasing frequency of "false alarms" has a significant desensitization effect on hospital staff, whereby some alarms may be erroneously "dismissed by assumption" as being "noncritical" [23]. This desensitization leads to both increased response times and decreased, or even lack of, clinician response. In the setting of a busy hospital, it is commonplace to hear constant chimes and beeps, each coming from different machines and indicating different "alarm conditions" (**Figure 1**). It should be more of an expectation that clinicians become desensitized to extraneous stimuli given the constant sensory bombardment coupled with the need for vigilance and differential interpretation of each alarm [25, 26]. When further compounded by heavy clinical workloads and long shifts, it becomes a matter of "statistical probability" before a critical alarm is missed [27–29]. Given the effect of this potentially dangerous phenomenon on both quality and safety of patient care, closer scrutiny of AF and related concepts is warranted. In this chapter, we will present a vignette-based discussion outlining fairly typical AF scenarios. Opportunities for improvement, including equipment,

personnel, and systems-based considerations, will then be provided.

"desensitization effect" associated with AF [10, 12].

94 Vignettes in Patient Safety - Volume 4

lapses [17–19].

For the purposes of this chapter, the authors performed a thorough literature search using PubMed, Google Scholar™, and Bioline International. Primary search terms included "alarm fatigue," "health-care alarms," "patient monitoring," "provider burnout," as well as secondary terms consisting of various combinations of primary search terms. From over 47,000 unique search results, we distilled 73 most pertinent references immediately relevant to this document. Finally, additional sources that were cited across our primary search results were added, for a total of 101 references included in the final manuscript.

#### **3. Patient monitoring: different types and modalities**

A diverse number of patient monitors are widely used across various health-care settings [30–32]. When employed correctly, they provide potentially valuable, actionable, and realtime information about a patient's clinical status. Different monitoring devices are intended to measure different parameters, potentially allowing for rapid assessment of a patient. This is especially relevant in the context of the current discussion of AF and more specifically the domain of alarm trigger accuracy [32, 33]. As clinical monitoring becomes more sophisticated and better integrated, remote (off-site) implementations also become possible [34–36]. The subsequent discussion will outline major types of monitoring equipment and alarms, including ventilation/oxygenation, hemodynamic, and pressure point alert systems.

#### **3.1. Ventilation/oxygenation alarms**

In general, primary ventilation/oxygenation alarms (VOA) include capnography and pulse oximetry, respectively. More broadly, respiratory parameter monitoring indicates the patient's oxygen saturation, respiratory rate, and end-tidal carbon dioxide [33, 37]. The use of VOA has been particularly important for critically ill patients who require mechanical ventilatory support. In such applications, the monitor is designed to be exquisitely sensitive to detect even the slightest changes in a patient's oxygenation or ventilation status [38]. As demonstrated in *Clinical Vignette #1* later in the chapter, an alarm may be triggered following the detection of a very small respiratory parameter "excursion," regardless of its clinical significance or magnitude of the observed change in the patient's actual clinical status. In this context, apnea and minute volume warnings are among the most common alarms triggered, with majority of such occurrences deemed clinically irrelevant upon further interrogation [39]. Moreover, many VOA triggers can be attributed to artifactual sources (e.g., patient movement, interruption of blood flow by inflating blood pressure cuff, and even atmospheric pressure variations) [37]. Thus, providers should be educated accordingly to ensure that the above considerations are appropriately factored into final clinical determinations and decisions.

Unfortunately, the one true positive alarm became lost in "a sea of false negatives." The practicality of BCPS alarms is also diminished by the inability of staff members to immediately assess/respond to the triggered alarm. Instances have been noted in which the alarm signal is transmitted after the event already transpired, as patients tend to fall immediately upon

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

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

97

In summary, the above-referenced monitor/alarm types have become an important part of the modern health-care fabric. Despite their ubiquitous use and great potential for constructive and practical clinical application, each type of device carries inherent flaws that providers must be aware of. Detailed knowledge of the risk-benefit equation associated with each device and clinical alarm type is important not only for patient safety but also required to help

Patient monitors are designed to have high sensitivity to predefined changes in various measured parameters, including vital signs, respiratory/ventilator status, and patient movements. However, the major drawback associated with high alarm sensitivity is the poor specificity and inherently disproportionate number of nonactionable (or nonclinical) alarms triggered [22, 53, 54]. Depending on the specific alarm and clinical setting, the estimated in range of "false positives" may be as high as 80–99% of all triggered alarms [8]. Broadly speaking, nonactionable alarms can be categorized as false alarms, nuisance alarms, and technical alarms (**Figure 1**). To elaborate further, false alarms occur in the absence of an actual patient or system trigger and typically result from a measurement artifact [55]. Technical alarms mandate the provider to attend to some operational aspect of the monitoring system, such as when readjustment of monitor leads/sensors is required [21]. Nuisance alarms are defined as clinically insignificant alarms that may interfere with patient care [10]. In aggregate, these nonactionable alarms are a major cause of the overall desensitization of hospital staff that may

Furthermore, to be effective, the alarms transmitted by monitoring systems must trigger some degree of cognitive response in health-care providers. This equates to introducing stress and the need for constant vigilance, both of which further heighten the risk of AF [56, 57]. When multiple clinical competing priorities collide, it becomes increasingly difficult for a provider to proactively address all ongoing problems, thus forcing them to resort to only partially addressing acute issues while at the same time disrupting other (parallel) activities due to multitasking [58–61]. Consequently, an ideal alarm should be perfectly audible and easily recognized by health-care providers working within the patient care unit [8], while at the

same time minimizing the amount of stress imposed on the responding clinical staff.

The increasingly complex environment of modern health-care systems has led to several important considerations regarding the practical application of monitoring systems. For example, space-related issues deserve special mention, with overly crowded clinical units creating an abundance of alarm-related stimuli and geographically larger clinical units

improve the quality and accuracy of the next generation of monitoring devices.

leaving the bed or chair [52].

**4. Patient monitor alarm design**

ultimately result in AF (**Figure 2**).

#### **3.2. Hemodynamic alarms**

Hemodynamic alarms (HA) monitor a variety of parameters, of which the most common ones include heart rate, systolic/diastolic/mean blood pressure, and various other intravascular pressure measurements via both invasive and noninvasive approaches [37, 40]. Hemodynamic monitoring has become a useful tool for the bedside assessment of patients in a number of clinical scenarios, from routine telemetry applications to advanced intravascular catheter utilization. There is some degree of predictability based on measured parameters, especially when trend determination and volume responsiveness are being considered [41, 42]. Hemodynamic monitors are particularly important in the setting of an unstable (or potentially unstable) patient, similar to the one described in *Clinical Vignette #3* later in the chapter. In such capacity, HAs can help facilitate rapid intervention and prompt correction of emergent issues. Still, HAs are far from perfect, with significant shortcomings in their discriminatory capabilities. More specifically, HAs are unable to identify a patient as "stable" or "unstable," especially when physiologic compensatory processes mask any underlying instability or in the setting of rapid change in hemodynamic status [43]. Thus, when using any particular monitoring modality, there is no substitute for an astute clinician who is able to effectively correlate HA findings with the clinical reality [44–46].

#### **3.3. Bed and chair pressure sensors**

Bed and chair pressure sensor (BCPS) alarms are utilized across many hospitals and other health-care facilities to help reduce mechanical falls among patients who experience ambulatory or balance difficulties [47, 48]. Falls typically occur as patients attempt to mobilize and/or ambulate without the required assistance of trained health-care staff [49]. Consequently, the use of BCPS alarms serves to alert staff—typically by a pressure-sensitive mechanism—when a patient attempts to move from a bed or chair without assistance. However, the weightsensitive pads are easily triggered by very slight patient movement, resulting in a significant number of false alarms [50, 51]. This challenge was readily apparent in *Clinical Vignette #3* later in the chapter, as the majority of BCPS alerts were likely due to the patient merely shifting slightly in the bed, and not by an actual attempt to independently mobilize and/or ambulate. Unfortunately, the one true positive alarm became lost in "a sea of false negatives." The practicality of BCPS alarms is also diminished by the inability of staff members to immediately assess/respond to the triggered alarm. Instances have been noted in which the alarm signal is transmitted after the event already transpired, as patients tend to fall immediately upon leaving the bed or chair [52].

In summary, the above-referenced monitor/alarm types have become an important part of the modern health-care fabric. Despite their ubiquitous use and great potential for constructive and practical clinical application, each type of device carries inherent flaws that providers must be aware of. Detailed knowledge of the risk-benefit equation associated with each device and clinical alarm type is important not only for patient safety but also required to help improve the quality and accuracy of the next generation of monitoring devices.
