**9. Technological advances in patient monitors**

Additionally, the US Food and Drug Administration's Manufacturer and User Facility Device Experience (MAUDE) database has identified 566 alarm-related patient deaths between January 2005 and June 2010 [79]. Reports detailing alarm-related events have prompted thorough investigation into AF and possible strategies to address this important phenomenon in

Considering the potential for very serious clinical consequences of AF, quality improvement measures have been proposed to help reduce both nonactionable alarm occurrences and the incidence of AF. Successful quality improvement projects must address multiple facets of the overall problem, including root causes that lead to AF (**Figure 6**). For example, poor usability and lack of user-centered devices have the potential for elevating clinical personnel stress levels, creating unnecessary workload and interjecting workflow inefficiencies into an already

Potential solutions for reducing the incidence of AF include multipronged approaches consisting of staff education, equipment (hardware and software) enhancements, and implementation of more efficient clinical protocols or guidelines [82–84]. From an educational perspective,

**Figure 6.** The different aspects of alarm fatigue that can be addressed through different quality improvement approaches

the clinical setting.

**Event** Falls

Delays in treatment Delays in ventilator use Medication errors Source: Ref. [24].

104 Vignettes in Patient Safety - Volume 4

**8. Quality improvement**

**Table 1.** Common alarm-related events leading to injuries or deaths.

tense environment [81].

(source: Ref. [80]).

In general, clinical monitoring is based on a careful balance between sensitivity and specificity of alarm signal recognition, as well as the associated threshold setting required to trigger "alert condition" [90, 91]. Increasing monitor sensitivity helps ensure that truly significant events are not missed, primarily using single-parameter alarms and default thresholds [8]. However, as a trade-off this increases the incidence of nuisance alarms that are nonactionable. This issue may be remedied by the development of "smart alarm systems" that use algorithmic approaches to evaluate multiple parameters prior to determining whether the detected change is truly critical, and only then sending an alert to the operator [15]. This improvement in device specificity would result in significantly fewer false alarms and therefore reduce


combined with heavy clinical workload, is known to have negative effects of hospital staff, including alarm desensitization and subsequent delay and/or lack of caregiver response. The resultant AF poses a serious risk to patient safety and has been associated with significant adverse events, including the need for additional or prolonged hospital care, excess attributable morbidity, and even mortality. Prevention of AF requires a multipronged approach consisting of quality improvement measures, staff training, better equipment management (e.g., monitor

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

threshold adjustments) to reduce false alarms, and focus on optimizing staff workload.

, Giuseppe Guglielmello<sup>3</sup>

1 Medical School of Temple University/St. Luke's University Health Network, Bethlehem,

2 Department of Surgery, Division of Acute Care Surgery, Traumatology and Surgical Critical

4 Department of Research and Innovation, St. Luke's University Health Network, Bethlehem,

[1] Kesselheim AS et al. Clinical decision support systems could be modified to reduce 'alert fatigue' while still minimizing the risk of litigation. Health Affairs. 2011;**30**(12):2310-2317

[2] Oppenheim MI et al. Design of a clinical alert system to facilitate development, testing, maintenance, and user-specific notification. In: Proceedings of the AMIA Symposium.

[3] Konkani A, Oakley B, Bauld TJ. Reducing hospital noise: A review of medical device alarm management. Biomedical Instrumentation & Technology. 2012;**46**(6):478-487 [4] Edholm K et al. Decrease in inpatient telemetry utilization through a system-wide electronic health record change and a multifaceted hospitalist intervention. Journal of

[5] Chen EH. Appropriate use of telemetry monitoring in hospitalized patients. Current

[6] Ivonye CC et al. Is telemetry overused? Is it as helpful as thought? Cleveland Clinic

Bethesda, Maryland: American Medical Informatics Association; 2000

3 Department of Medicine, Section of Pulmonary and Critical Care Medicine, St. Luke's

Care, St. Luke's University Health Network, Bethlehem, Pennsylvania, USA

and Stanislaw P. Stawicki<sup>4</sup>

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

\*

107

**Author details**

James Nguyen<sup>1</sup>

Pennsylvania, USA

Pennsylvania, USA

**References**

, Kendra Davis<sup>2</sup>

\*Address all correspondence to: stawicki.ace@gmail.com

University Health Network, Bethlehem, Pennsylvania, USA

hospital medicine. Aug 2018;**13**(8):531-536

Journal of Medicine. 2009;**76**:369

Emergency and Hospital Medicine Reports. 2014;**2**(1):52-56

**Table 2.** Applications of artificial intelligence in the development of intensive care monitoring.

AF. At the same time, the challenges of "unpredictable code" and "interrupted or corrupt data" have been noted and may represent an important safety issue due to the potential for missing data or data misinterpretation, especially when using memory-intensive applications on devices that are continually operating for prolonged periods of time [89, 92–95].

The ideal patient monitor would have high sensitivity, as well as high negative predictive value for life-threatening clinical scenarios. This would result in excellent "event detection rate" while reducing the number of false and nuisance alarms. Still, any improvement of sensitivity/negative predicative value for monitors must be accompanied by corresponding adjustment to specificity/positive predictive value, ensuring that clinically significant events are captured efficiently [33]. The accomplishment of the above goals may be possible using the application of artificial intelligence (AI) in monitoring systems, wherein AI would be incorporated into logic-based, decision-making systems. The ultimate goal would be the development of clinical monitoring capabilities that reflect and mirror human cognitive/ decision-making processes [37]. In the context of this chapter's *Clinical Vignettes*, the application of such AI-based systems might be helpful in minimizing the number of nonactionable alarms, thus reducing the subsequent AF associated with adverse clinical events. So far, the utilization of AI has been explored in several different applications (**Table 2**).

#### **10. Conclusion**

Given the proliferation of advanced monitoring equipment, AF continues to be a major patient safety issue across modern health-care systems. While technological advances show great promise in improving patient care, significant barriers to more optimal implementations exist, including the ongoing struggle to balance the need for high sensitivity versus the excessive number nonactionable clinical alarms. The high frequency of clinical alerts, especially when combined with heavy clinical workload, is known to have negative effects of hospital staff, including alarm desensitization and subsequent delay and/or lack of caregiver response. The resultant AF poses a serious risk to patient safety and has been associated with significant adverse events, including the need for additional or prolonged hospital care, excess attributable morbidity, and even mortality. Prevention of AF requires a multipronged approach consisting of quality improvement measures, staff training, better equipment management (e.g., monitor threshold adjustments) to reduce false alarms, and focus on optimizing staff workload.
