Cvach, Maria M. DNP, RN, CCRN; Currie, Andrew MS; Sapirstein, Adam MD; Doyle, Peter A. PhD, CHFP; Pronovost, Peter MD, PhD, FCCM
Too often clinicians are apathetic to alarms, thus they silence, disable, or ignore them. In fact, patients have been discovered unconscious and not breathing while leads-off alarms went unnoticed.1 Failure to respond to alarm signals in a timely manner has led The Joint Commission (TJC) to introduce a new National Patient Safety Goal on alarm management beginning in 2014.2
Alarms are common on clinical units, but most are false, and these false alarms distract nurses from performing important work, making them immune to true alarms. Studies indicate that 85% to 99% of alarm signals generated by medical devices are false and/or clinically insignificant, resulting in a phenomenon known as alarm fatigue.3–6 Alarm fatigue is the lack of response to an alarm due to excessive numbers, resulting in sensory overload and desensitization. From 2005 to the middle of 2010, 216 reported deaths have been linked to monitor alarm systems.1 ECRI (formally known as the Emergency Care and Research Institute), a nonprofit organization that uses applied scientific research in healthcare to establish best practices for improving patient care, lists alarm issues as the number one health technology device hazard for the past 2 years.7 The Association for the Advancement of Medical Instrumentation issued a challenge that by 2017 no patient will be harmed by adverse alarm events.8
While higher nurse staffing is associated with reduced hospital mortality, false alarms distract nurses and occupy their time with interventions that don't benefit patients. Thus, these false alarms effectively reduce the value of increased nurse staffing. Hospitals are packed with an ever-growing number of medical devices, each one with its own alarm, and hospitals lack policies defining alarm accountability, as well as technologies to integrate and prioritize alarms.
Because of the challenges stated above, TJC has created an alarm management goal (NPSG.06.01.01) aimed at improving clinical alarm system safety. This goal will be phased-in with two Elements of Performance (EP) being implemented by July 2014 and two EPs being implemented by 2016. The first 2 EPs address hospital leadership establishing alarm system safety as a hospital priority, as well as identifying the most important alarm signals to manage. EP 3 requires hospitals to establish alarm management policies and procedures for those alarms identified in EP 2, including appropriate settings for alarm signals; defining when alarm signals can be disabled; who can change alarm parameter settings; when alarms can be turned “off;” and who and how staff respond to alarm signals. The final EP indicates that hospital staff must be educated about the purpose and proper operation of alarm systems for which they're responsible.9
The Johns Hopkins experience
The Johns Hopkins Hospital (JHH), a 1,000-bed academic medical center, began an alarm management initiative in 2006. To improve patient safety, JHH implemented teams called Comprehensive Unit-based Safety Programs (CUSP) to implement JHH safety goals. CUSP teams identify and mitigate unit-specific safety risks and improve safety culture.10 The teams are led by unit staff and hospital executives who help staff navigate complex organizational systems. The medical progressive care unit (MPCU) CUSP team identified missed alarms as a serious problem requiring attention.
To facilitate the process of alarm management on the MPCU, hospital leadership requested the Director of Clinical Engineering and the Assistant Director of Nursing, Clinical Standards, to assist the CUSP team. Initially, a small interprofessional group analyzed the issue and pilot tested alarm reduction strategies, including making modest monitor default parameter changes to reduce audible alarm signals and educating staff on how to customize alarms based on patient need. The unit experienced a 43% reduction in critical alarms as a result of these rapid-cycle tests of change.11
Within a year, the small group expanded to include representation from all monitored units at JHH and purposefully addressed the issue in a systematic fashion, getting input from all stakeholders. Today, the JHH Alarm Management Committee consists of nurse representatives from all monitored areas, a human factors engineer, a respiratory therapist, clinical engineers, an information technologist, physicians with expertise in safety, and a patient representative. Hospital risk management and vendors are requested to attend, as needed.
Analyze the problem of missed alarms
To understand the problem of missed alarms, the Committee used a fault tree analysis (FTA) identifying seven potential top-level failures that can lead to a missed alarm. FTA is a robust tool to proactively identify system failures. We identified root causes for each failure type and used the root causes to identify solutions for corrective action. This enabled the Committee to implement changes to reduce risk and to develop redundant systems to improve patient safety. (Figure 1illustrates the top level of the FTA.)
Use data to drive change
To establish a baseline for measuring alarm burden, Clinical Engineering extracted monitor alarm data. We analyzed alarms for each monitored unit using the metric alarms per monitored bed per day. Initial alarm analyses of our ICUs ranged from 100 to 771 alarms per monitored bed per day. We also analyzed the data to learn the types of alarms and which were actionable. Monitor alarms occur in a hierarchy of high, medium, and low priority. Each hierarchical alarm produces a different audible tone. Clinicians use the tones to determine the degree of urgency and the proper alarm response. The more frequently clinically insignificant alarms signal, the more likely alarms (in general) will be ignored. Thus, reducing the overall alarm burden will improve the likelihood of alarm response, thereby improving patient safety.
By surveying practice and examining literature, we learned that there are no standards for monitor default parameter settings. Each unit's settings varied, and often the default settings on bedside monitors also varied within units. There were duplicate alarms, and parameter thresholds were set to non-actionable levels, resulting in excessive alarm noise. Making modest changes to monitor default parameters and empowering nursing staff to customize alarms for patient need resulted in large reductions in clinically insignificant alarms and a quieter environment.11 Today, Clinical Engineering provides weekly monitor alarm data reports to Alarm Committee members. Most units maintain an average of fewer than 100 alarms/bed/day. Details regarding weekly alarm events are provided to Committee members for review and use for education purposes.
Prepare an alarm inventory
The JHH Clinical Engineering Department conducted an alarm inventory, which included monitor default settings and arrhythmia levels for all cardiac monitors on each unit, and provided it to Committee members. The Committee's nurse and physician chairpersons met with each unit's nursing and medical leadership to review the purpose of the alarm management initiative and review current monitor default settings and arrhythmia levels. We assisted unit leadership in understanding how to reduce alarm signals by setting monitor defaults to actionable levels. When possible, we standardized default settings, resulting in the following population-specific parameter sets: telemetry, progressive care, surgical intensive care, medical intensive care, cardiology care, oncology, and pediatrics.
After monitor alarm parameters were established for each unit, the next step was to conduct an alarm inventory of other unit-based medical device alarms. The FDA Manufacturer and Use Database provides direction on which medical devices have been associated with alarm-related deaths.12 We used this information to identify “at risk” medical devices, along with JHH error reports and clinicians' perceptions of risks. Not surprisingly, cardiac monitors and ventilators were among the devices most often linked to alarm-related events.
We identified over 20 medical devices on acute, inpatient hospital units and prepared an alarm inventory of those devices. Table 1 represents a sample medical equipment device alarm inventory. We classified the priority of the alarm signals as A) requires immediate attention, B) requires attention as soon as possible, and C) timely response required. We categorized the level of medical device alarm oversight as high, medium, and low. Finally, we discussed the type of secondary alarm notification approaches that could be used when a clinician isn't in immediate proximity of the alarming medical equipment. Secondary alarm notification is a method of communicating an alarm through another mechanism such as a wireless device, waveform display, marquee sign, or human monitor watch. Recognizing that secondary notification devices add to noise and distraction, the Committee was careful to design its use to provide back up with minimal negative effect.
Use a rapid cycle change approach
Using a systems-based, rapid cycle change approach, we tested methods to address failures identified in the FTA with the purpose of alleviating false and clinically insignificant alarms. The rapid cycle approach enabled us to examine independently the effects of various strategies and sustainability challenges in a stepwise manner on select units. By doing so, it allowed us to implement the appropriate and cost-effective measures to reduce alarms house-wide.
Since implementing these improvements, we continue to monitor alarm frequency data. (Table 2demonstrates the pre- and post-intervention results showing reduction in alarm signals on some units as a result of implementing all strategies.) We continue to examine new strategies to minimize alarm fatigue, such as use of a nurse-managed telemetry discontinuation protocol and use of “smart alarms.”
Identify best method(s) to notify staff
Getting alarm signals to the right person at the right time is critical. The Alarm Committee explored many possibilities for effective alarm notification. There's no single best solution for all units. Using bedside monitor split-screens to allow multiple patients to be viewed, hallway waveform displays, and automatic view-on-alarm are examples of possible solutions. Our hospital was undergoing a large renovation project, which afforded us the opportunity to implement an alarm integration system with wireless acknowledgement devices in our new clinical buildings. This system enables alarm signals to be directed to wireless devices according to hospital-defined algorithms that ensure alarm coverage.13 Multiple algorithms were developed for monitor alarm signals. For instance, the crisis monitor alarm algorithm sends high-priority signals immediately to the nurse's wireless acknowledgement device. If the nurse is unable to address the alarm signal, he or she can escalate the alarm to another “buddy” nurse. When there's no response to an alarm in a predefined period of time, signals are automatically escalated to a charge nurse. For non-crisis monitor alarm signals, a slight delay in the algorithm allows for alarm correction, which helps to reduce the number of clinically insignificant alarms sent to the nurse's wireless device.14
Develop an alarm policy and provide initial and ongoing education
The Alarm Committee has incorporated alarm management into device-specific policies. These policies address alarm response, who can change alarm default parameter settings, secondary alarm notification, when alarms can be turned off and by whom, and permission for clinicians to customize alarm settings for patients based on need. Alarm review has been incorporated into our electronic medical record as a reminder to nurses to perform this task minimally each shift. (SeeTable 3.) Each unit has “alarm champions” who are members of the Alarm Committee and guide the Committee regarding policy and education. The members of this Committee are engaged and empowered to recommend changes that directly benefit patients. Monitor and alarm policy reeducation was provided house-wide when we began the alarm initiative. We've incorporated monitor alarm management into orientation; it's reinforced by the unit alarm champions. The policy provides a critical behavior checklist for monitor set-up and alarm management, which is used by preceptors to train new employees.
This article identifies a systematic approach to identify and control problems related to alarm management. We provide data from our experience in evaluating corrective measures and recommended strategies useful by other institutions. Specific results are always dependent on patient acuity, staffing variations, and other uncontrollable factors in a dynamic facility. Therefore, the results obtained can't be considered repeatable or taken as absolutes. However, the trends show benefits received from the strategies employed. We acknowledge that the results of incremental changes from these improvements may diminish over time. We also recognize that we didn't evaluate the impact of these changes on patients or nurses. Once put into place, alarm management strategies must be re-evaluated and pertinent steps taken to ensure continued control of alarm-related issues.
3. Lawless S.T. Crying wolf: false alarms in a pediatric intensive care unit. Critical Care Medicine. 1994; 22:, 981–985.
4. Tsien C., Fackler J. Poor prognosis of existing monitors in the intensive care unit. Critical Care Medicine. 1999; 25:(4), 614–619.
5. Chambrin M.C., Ravaux P., Calvelo-Aros D., Jaborska A., Chopin C., Boniface B.. Multicentric study of monitoring alarms in the adult intensive care unit: a descriptive analysis. Intensive Care Med. 1999; 25:, 1360–1366.
6. Atzema C., Schull M.J.. Alarmed: Adverse events in low-risk patients with chest pain receiving electrocardiographic monitoring in the emergency department: A pilot study. American Journal of Emergency Medicine. 2006; 24:(1), 62–67.
10. Timmel J., Kent P.S., Holsmueller C.G., Paine L., Schulick R.D., Pronovost P.J.. Impact of the comprehensive unit-based safety program (CUSP) on safety culture in a surgical inpatient unit. Jt Comm J Qual Patient Saf. , 2010June; 36:(6):252–60.
11. Graham K., Cvach M.. Monitor alarm fatigue: standardizing use of physiologic monitors and decreasing nuisance alarms. Am J Crit Care. 2010; 19:(1) 28–34.
12. Weil K. 2005–2006 Alarm Related Death by Device, Food and Drug Administration.
13. Hoglund D., Elms J.. The use of mobile devices to improve alarm systems. BI&T. Sept/Oct 2011; P. 381–84.
14. Cvach M., Frank R.J., Doyle P., Khouri-Stevens Z.. Use of pagers with an alarm escalation system to reduce cardiac monitor alarm signals. J Nurs Care Qual. 2013; DOI: 10.1097/NCQ.0b013e3182a61887.
15. Cvach M., Biggs M., Rothwell K.J., Charles-Hudson C. Daily electrode change and effect on cardiac monitor alarms. J Nurs Care Qual. 2013; 28:(3); 265–71.