In addition to these specific adverse effects, excessive and poorly managed alarms are a contributing factor to the general sense of clinicians feeling overburdened and stressed in the present healthcare environment.
Alarm fatigue occurs when healthcare workers become desensitized to the sounds of alarms, and may silence or ignore an alarm with the belief that it is not an important or emergent issue. This can lead to failure to rescue events, where a patient’s adverse event is not recognized in time to prevent death or serious injury. Failure to rescue events have been defined as incidents in which a patient in a hospital develops a complication that is not recognized in a timely manner and averted, which in turn leads to higher mortality rates. It is thought that alarms may have been a contributing factor in these events.
Alarm systems have been used in the hospital environment for well over half a century, and are designed to alert healthcare workers to both critical and non-critical changes in patient condition. Alarms can be physiological (e.g. changes in heart rate, blood pressure) or technical (e.g. infusion pump alarms) and there are approximately 10-15 alarms per patient per day in the acute care environment. It is estimated that between 85-99% of the alarm signals do not require clinical intervention and alarm safety was listed in the ECRI institute’s top 10 health technology hazards for the past 4 years.
This section will provide some context to the subject matter, focusing on: alarms in the healthcare setting, the importance of alarm management systems, challenges and adverse effects of poorly managed alarms, and the potential in improved alarm technology to play a role in addressing these issues.
Importance of Clinical Alarm Management Software
There are many reasons that an alarm can be ignored, even when the care provider believes that the alarm is important and doesn’t intend to miss it. Most alarms are false and do not require clinical intervention. In fact, it has been reported that up to 85-99% of alarms in certain units are, in fact, false. Alarm fatigue can also occur when there are too many different alarm tones for a single care provider to remember or keep track of. This is particularly common in children’s hospitals, where each medical device may have its own distinctive alarm tone. Finally, cognitive overload resulting from too many alarms, many of which require a complex intervention, can lead to a numbing of the senses and inability to respond to any alarm.
Alarm fatigue occurs when care providers become desensitized to or overwhelmed by an excessive number of alarms, which can result in delayed (i.e., the alarm signal occurs and is not acted on for a clinically appropriate time) or missed patient care. It is a sensory overload, defined by physiological, psychological, or emotional energy depletion as a result of the demands from the environment. It can result in the care providers ignoring an alarm, turning down the volume of the alarm, or altering the alarm settings to be non-actionable. Staff may even leave their unit without disabling the alarm or call system, resulting in other staff having to respond to the alarm.
Challenges in Alarm Management
In a study by Abreu et al. it was found that the effectiveness of alarm systems is affected by the location of the alarm generation and the requirements for staff to interpret how serious the situation is and what action to take. Alarm types and physiological parameters were often unclear and lead to a lack of staff knowledge on how to respond to an alarm. This confusion could result in a misinterpretation of the alarm, or alarms being turned off because staff are unclear on what the alarm is indicating. CAMS should therefore be designed in consultation with clinical staff to identify alarm types and determine the most appropriate response. Finally, alarm systems can be seen as surveillance devices as they are recorders of clinical events and at times the behaviour of clinical staff. Abreu et al. state that both patients and staff have mixed feelings about being under surveillance, it generally makes them feel uncomfortable. CAMS needs to respect the privacy and autonomy of the patient and avoid being seen as a mechanism used to control staff behaviour.
Implementation of CAMS into the clinical setting poses numerous challenges. The technology currently used in hospitals is diverse and fragmented. In many cases, multiple alarm-generating devices are used for a single patient (i.e. pulse oximeter, ECG monitor, ventilator), many of these devices have their own alarm systems – often with different alarm sounds and required actions. As a result, a general ward in a hospital can experience in excess of 100 different audible alarms per patient per day. The high frequency of clinical alarms and the cacophony of sounds have contributed to ‘alarm fatigue’, where clinical staff become desensitised to the sound of an alarm, sometimes resulting in a failure to act or delayed response. Alarm fatigue has been implicated in several adverse events including an instance whereby a patient on a monitored bed died, with staff supposing the alarm sounding was due to a faulty machine. A CAMS system aimed at preventing such an event would need to intelligently filter and prioritise alarms so that important events are brought to the attention of the relevant clinician without excessive noise interference.
Benefits of Implementing Clinical Alarm Management Software
This module provides important background information about the issues surrounding and the complexity of the clinical alarm process and offers a tactical solution. Healthcare is a dynamic and complex environment. An area of concern that is becoming more prevalent and demanding an effective solution is the management of clinical alarms. Alarm messages are transmitted from medical devices to clinical information systems and alarm endpoints. An alarm message is a signal that an alarm occurred and is treated in some fashion (i.e. a visual message, text message, phone call, etc.). Although alarm messages are critical in identifying potential patient adverse events, the reliability and effectiveness in integrating alarm data into clinical workflow is not sufficient. The Joint Commission Board of Health Systems and the ECRI Institute have identified issues with clinical alarms in 2013 in their top 10 technologies and patient safety concerns. Despite not being marketed as a patient safety concern, without considering alarm endpoint integration, data integration, and proper alarm delivery, potential for adverse events associated with undetected alarms is a patient safety concern. This article is aimed primarily at data integration and proper alarm delivery. Alarm endpoint integration is covered in depth elsewhere.
Features and Functionality
Consider customizability when configuring each type of alarm. Can the parameters be defined by the clinical setting and severity of the individual patient’s condition? Also crucial is ensuring that the alarm system is tailored to roles and responsibilities of the staff. There will be an overwhelming number of alarms for staff to respond to, so it is important to only direct each alarm to the relevant staff, ensuring the correct response and of course reducing alarm fatigue. Essentially the ability to tailor the alarm system to specific clinical needs will result in higher alarm efficacy, and consequently patient safety.
First, you’ll need to consider the goal of the alarm management software – is it replacing existing systems, or being used as a supplement to these? What data do you want to collect from alarms, and what goals do you want to set to determine ‘alarm success’? Determine what alarms you want to be included in the system; different systems will cover different devices and alarm types.
Real-time Alarm Monitoring
Real-time alarm surveillance comes into play if the aim is accurate, sensitive and specific alarm generation (i.e., few false alarms) using the least invasive sensors. Unfortunately, the least invasive sensor is, the less likely it is to provide a highly accurate signal, therefore clinical systems must weigh sensor choice against the likelihood of detecting a clinically significant condition. For example, a highly specific but not very sensitive arrhythmia detection algorithm could be used in conjunction with ECG monitoring in the post-operative setting where the goal is to detect a particular arrhythmia before it causes symptoms. On the other hand, a more sensitive, but less specific ambulation monitor might be appropriate for detection of increased fall risk in an elderly patient. The issue of alarm fatigue can also be addressed in several ways. An alarm suppression policy for specific units of patients with expected clinical events can be implemented so that alarms do not sound during a normal, expected event. For example, a patient recovering from coronary artery bypass surgery is not likely to do much heavy lifting for several weeks, therefore an “off-activity” monitor could be set with upper limits higher than the expected postoperative activity level. Automatic adjustments to alarm settings based on patient condition or clinician confirmation of an event are other potential methods for reducing alarm fatigue.
Customizable Alarm Settings
Integration with Existing Clinical Systems: As illustrated in Figure 1, MIMIC II uses a frontend application to monitor an ICU database and implement alarm algorithms. We wish to extend the alarm algorithms to the point that they are implemented directly with the physiologic data in the database. As such, an externalized alarm would be a trigger from a complex event monitor which monitors a patient’s ECG looking for myocardial ischemia. When the event monitor detects a pattern of significant ST segment depression, the alarm algorithm will be called upon to compare the patient’s current ECG with a reference ECG to confirm the presence of myocardial ischemia and generate an appropriate alarm. This level of clinical decision support requires a deeper understanding of specific clinical events and is only possible if the alarm algorithms are an integrated part of a system’s clinical software.
Alarm delay settings: Alarm delays are useful for tailoring alarm responses to specific patient situations. This ranges from simply delaying the visual and aural notification of an alarm to the complete holding and subsequent release of an alarm.
Integration with Existing Clinical Systems
Beyond inpatient clinical monitoring devices, many adverse events occur in hospital environments as a result of communication failures between levels of care or different healthcare settings. For example, a patient with critical lab results (e.g. high INR) may be seen in a clinic setting by a primary care provider and receiving care from an anticoagulation nurse who is not aware of the lab results. Creation of a unified alerting and alarm platform that extends across different levels of care and settings can allow these patients to be identified and prevent adverse events. The alert can direct the anticoagulation nurse back to the lab results and advise temporary dose adjustment. Similar strategies can be employed for patients with diet changes, medication changes, or upcoming procedures. This type of transitional care alerting has the potential to greatly impact patient safety.
The integration of CMS with other clinical systems has the potential to significantly increase the safety and efficiency of patient care. Currently, many different alerting devices are used within the clinical environment. It is not uncommon for an individual patient to have several different types of alerts and alarms being monitored by clinical staff. For example, a patient in a typical hospital room may have cardiac telemetry, pulse oximetry, and blood pressure measurements all with their own corresponding alarm mechanisms. Often, these different alerts become disconnected from the true severity of the patient’s condition and can lead to alarm fatigue from staff. Integration of these alerts into a single platform will allow the software to use a predetermined set of clinical rules to establish the severity of the patient’s condition based on vital sign data and provide a single unified alarm when intervention is required. This method has been shown to significantly reduce the number of alarms generated and increase the correct identification of patients in crisis.
Reporting and Analytics
While still directly involved in patient care, nurses and other clinicians working in hospitals are painfully aware that redundant, non-actionable and false clinical alarms are a worsening side effect of the proliferation of medical devices. It is increasingly recognized that these alarms pose threats to patient safety, as they can lead to alarm fatigue, the desensitization of caregivers to alarms, possibly resulting in missed alarm signals and failure to rescue. Alarm fatigue was one of the ECRI Institute’s Top 10 Health Technology Hazards for 2013. While monitoring the condition of their patients across multiple devices connected to patients, clinicians are seeking more sophisticated and automated ways to help them separate the true clinical alarms from the spurious ones. This need is driving a boom in the development of surveillance and clinical alarm notification tools and middleware. Our recently published global market report on patient monitoring devices covers some of these tools under the category of secondary alarm notification and alarm management software. Global Markets for Patient Monitoring Equipment. The alarm notification and alarm management tools are designed to take data from patient monitoring or other medical devices, analyze that data based on configurable algorithms, and provide notification to caregivers when there is a high probability that a true clinical event has occurred. Examples of companies active in secondary alarm notification include extension360 (from Spacelabs), iSirona, Philips (in 2012 Philips acquired Emergin and partners with both Epic and Cerner) and Welch Allyn. These and other vendors typically have proprietary user interfaces and methods for directing alarms to caregivers, which can make integration of these alarm notification tools challenging for clinical end users. Yet, there is a growing recognition that the changing dynamics of alarm notification and the need to make it truly actionable will necessitate improvements in standardized methods for alarm notification and alarm transmission between medical devices and clinical information systems (CIS) ancillary or Electronic Medical Record (EMR) systems. Standardization and integration of modes for physiologic data and alarm and alerts.
Implementation and Integration
Assessing organizational needs is often said to be the first step of system implementation. In the case of clinical alarm management software, the gap between how much the software can offer and the current status of alarm management in a healthcare facility is apparent. It is important to analyze where the organization currently stands with alarm management and where it ideally would like to be in the future. This can be achieved through conducting a SWOT analysis of how alarm management is conducted at present. This is an analysis of the organization’s strengths, weaknesses, opportunities, and threats in relation to alarm management. Strengths and weaknesses are usually easy to identify from an internal analysis of the current state of alarm management. Opportunities and threats may require some benchmarking against other similar healthcare organizations to identify different practices of alarm management that may be more effective, or to learn from the mistakes of others when alarm management has gone wrong. Identifying these opportunities and threats can provide a better outlook on where the organization would like to improve or where priority areas for change in alarm management should occur. This information can be used to develop specific goals and objectives that the organization would like to achieve with the implementation of new alarm management software. These may range from meeting set standards as defined by the Joint Commission, reducing the number of adverse events related to alarm management, or freeing up nursing time so that they can better attend to patient needs. Having identified a clear direction of where it would like to head with alarm management, the organization is then in a position to investigate the best software to meet their needs.
Assessing Organizational Needs
Part of the reason for said frustration lies in the breadth of area that must be assessed. A clinical alarm is any notification of a change in the medical condition or life support of a patient generated from monitoring equipment or medical devices. Such alarms are an integral part of patient care, and when designed effectively, alert caregivers to potentially life-threatening changes in a patient’s condition. However, because of a lack standards as to how the alarms are programmed and managed, there is also a plethora of non-actionable nuisance alarms, and it is these nuisance alarms which are of critical concern in patient safety today. It is estimated that 85-99% of all clinical alarms do not require clinical intervention and may be considered nuisance alarms. Environmental alarms and device technical alarms may also sound when there is no significant patient condition change. This too can be cause for concern in patient safety. All of the aforementioned alarm types contribute to alarm fatigue. Alarm fatigue is sensory overload when caregivers are exposed to an excessive number of alarms, and is also a growing concern in healthcare today. A recent ECRI Institute report cited medical device alarm hazards as the number one health technology hazard that puts patients at unnecessary risk. All of the above poses the question: how does an organization effectively identify any and all problems related to clinical alarms, and how does one identify specific areas where an alarm management solution will bring added safety and efficiency to patient care?
However, the success of the oncoming project will ultimately hinge on the depth, accuracy, and honesty of this self-same assessment. And despite the wealth of tools and professionals available for such an assessment, numerous healthcare organizations find the process an overwhelming and ultimately flawed endeavor.
Selecting the Right Clinical Alarm Management Software
In order to improve the clinical alarm safety policy of any organization, the first and most important step is ensuring that the clinical alarm management software chosen is the right fit for the organization. Thus, it is crucial to evaluate the capabilities of different software and how well they fit the individual organizational needs. The time taken to effectively compare different solutions to find the best one to meet the organization’s need is a worthwhile investment. The software selected should be able to cater to the unique needs of the specific organization, helping to reduce the alarm burden, reduce alarm fatigue, and improve patient safety. A software may perform well in achieving this in some areas, such as long-term care, but not so much in others. Another factor to consider is the reliability and validity of the alarm sounds used in the software. As mentioned before, the misuse of alarms can create a desensitization to them, which is due to the amount of false and non-actionable alarms. This often results from default alarm settings that are often irrelevant to the patient. Thus, it is important to be able to customize alarm settings based on the physiological and environmental factors specific to individual patients to minimize clinically insignificant alarms. An alarm is only useful when there is an appropriate action response to follow it. In the case of an event occurring to the patient that requires attention, there is a need for an actionable alarm to alert the appropriate clinician. Thus, the software should have the capability to filter and direct alarms to the specific qualified healthcare professionals pertaining to the severity of the alarm. This will help to establish a culture of safety and more effectively monitor patients.
Integration Process with Existing Systems
The first step with any software integration process is to identify the systems and devices that will be replaced or upgraded. This will involve a detailed inventory of medical devices and their PCMs, along with an evaluation of the significance of each device in relation to the hospital’s clinical workflows. Bothe (2006) notes that it is often necessary to modify the existing infrastructure to support the new technology, particularly in cases where the new technology should enhance the system. This was evident in a study of home telehealth technology that required both organizational and technical adaptations to the existing healthcare system (Darkins, 2008). Device modifications and system changeovers have their associated risks and costs, thus a detailed cost benefit analysis will be required for each change. This is particularly important for devices that are not replaced upon failure, as costs to replace PCMs on devices no longer in production can exceed the cost of the device itself.
The integration of new software with existing systems can be a complex and challenging process. Blike and Kelchner (2008) warn that the vast majority of medical devices are not designed with interoperability in mind. Even when software and devices are labelled as “interoperable”, the minimum standards of operability, and the functions that should be achievable are not well defined (Sittig, 2008). Treating interoperability as a plug and play situation can lead to disappointment, cost overruns and a sense of failure. Bothe (2006) used ethnographic research methods to study the integration of a clinical decision support system, noting the temptation for local staff to oversimplify the technology and the complexities of the method involved, potentially leading to disparities in the new and old methods of working.
Training and Education for Staff
Healthcare staff must be well informed about the reasons their organization is adopting an alarm management solution and understand the implications of the change. One cannot expect staff to simply accept a change without a clear understanding or educational induction into the change. Staff need to be trained on the specifics of the alarm system, who will respond to certain alarms, customize alarm settings, and manage the alarm information output and reports. They also need to be educated on alarm system resources, alarm response time requirements, and alarm management objectives. This will provide staff with the information required to perform their alarm management-related duties. Simulation-based learning has proven to be an effective method of education for healthcare professionals. The use of games or simulated experiences has shown to increase learner enjoyment, satisfaction, and self-confidence, and improve knowledge and retention. This may be an approach healthcare organizations would consider to effectively train and educate their staff on alarm management. Now, more than ever, simulation is a critical component to delivering safe, quality patient care. A number of years ago, The Joint Commission called for a reduction of alarms given the high number of non-actionable alarms and the resulting alarm fatigue in healthcare staff. In response, hospitals are now limiting the use of certain clinical alarms and creating alarm response teams, which has resulted in some confusion and danger for new, inexperienced staff members. These trends are affecting patient care delivery in a variety of practice settings — clinicians need to be proficient in determining life-threatening situations and managing and responding to clinical alarms. Simulation can help give staff the hands-on practice needed to properly manage and respond to alarms, as it mimics real work experiences and allows for immediate feedback on practice performance. Simulation can help prepare staff members to become competent in alarm management without compromising patient safety. During this critical time of change in healthcare alarm management, education is essential in preparing staff to keep pace with the changes and develop the competencies needed to manage alarms in both current and future practice settings.
Overcoming Implementation Challenges
Recognition of the potential issues at hand and planning contingencies and strategies to mitigate these issues will serve to enable a more fluid and efficient implementation of the alarm management system and the changes it brings to the clinical alarm environment. With this in mind, it is evident there are multiple steps and often complex challenges in fully transitioning to the alarm management system; however, thorough implementation will lead to a safer clinical environment for patients.
Usually, the most difficult and time-consuming implementation of the alarm management system targets the individual care unit. This alarm system is considered to be an extension of the central alarm system, in which the device triggers can still be seen by nurses and forwarded to other medical staff on a case-by-case basis. In this situation, the desired target system will still eventuate; however, the overnight switch from the old alarming system to the new system will be difficult to enforce. The way to overcome this obstacle is a phased implementation with constant reassessment and education.
In order to switch off all irrelevant beeping and pagers around the clinical environment, there needs to be a system in place to ensure that every department is catered for. To do this, an inventory of all devices that have alarm capabilities needs to be identified before devising plans to either eliminate the device or modify its output in the case of a false alarm. This process will then need to be prioritized into which alarm needs attention first, and lastly, this change to the alarm or device needs to be documented in case of future re-installation.
In order for the alarm management system to be considered a success, the complete implementation of the system to override all other alarms in the clinical environment must be completed. Although this idea is feasible, in reality it will prove to be a complex and time-consuming challenge. However, ensuring a safe alerting environment for patients can be achieved by breaking down the process of implementation and addressing the challenges that arise.
Best Practices for Clinical Alarm Management
A major issue the authors point out is that most alarm software is not very efficient as the alarms are generated without any mode of prioritization. It is important that the correct alarms are set to go off when needed and that staff can distinguish between an urgent alarm and a more general one. Clarian Health Partners in Indianapolis achieved a 43% reduction in alarm signals and a 52% reduction in calls to the unit for help regarding alarms through modifying the default settings. This was done by forming a multidisciplinary team who defined default alarm settings and identified the parameters for customizing alarms. A major pitfall of Clarian’s project was that it implemented one of the first stages in introducing an alarm management system but failed to follow it through due to a lack of monitoring and maintenance. Default settings were altered by staff over time and monitoring was lost. It is important that alarm settings are continually monitored and maintained, this tends to be forgotten and abandoned once an initial system is in place. Regular maintenance and testing is also enforced in the ECRI Institute’s guide. Measures taken include conducting annual performance evaluations of all alarm systems as well as conducting periodic testing of individual alarm components. Each year at a set date, designated by the alarm system technician, each alarm system is silenced, and each individual alarm signal (such as that produced by a vent) is tested to assess its audibility and function.
Establishing Alarm Management Policies and Procedures
Once an organization has created policies on the use of alarms, procedures should be established for the customization of alarm parameters. In the acute care setting, medical devices are often purchased with the manufacturers’ default settings. In many cases, the default settings are programmed to monitor the sickest patients and can result in a high number of false alarms for less acute patients. An extensive list of the types of alarms and desired settings for each alarm should be made and approved by clinicians familiar with the population of patients being monitored. This will involve changing some alarm parameters and turning off some alarms for different patients. The procedures should outline who will make these changes and how and when they will be evaluated. Once all desired changes have been made, it is important to evaluate the effects of customization on adverse patient outcomes so that changes in alarm parameters can be linked to patient results and so that changes that do not result in better patient outcomes can be reverted.
Alarm management policies and procedures are a critical component of a successful alarm management program. Written policies should address the development, implementation, and maintenance of alarm systems. The objective of the policies is to ensure that alarms are used appropriately and safely in a manner that supports the best patient outcomes. Specific areas to address in the policies include the assignment of accountability for alarms, the process for deciding when an alarm is necessary, the type of alarm to be used, and default alarm settings.
Setting Priorities and Thresholds
Setting priorities for various kinds of alarms is fundamental to an effective alarm management strategy because it ensures that the most clinically relevant alarms are presented to the end user. A data-driven, methodical approach should be taken to allocating categories and priority levels to different types of alarms. This can be achieved in a multidisciplinary fashion using a consensus approach. Aboukhalil et al. provide some excellent examples of how this can be done in the intensive care setting. They have defined a 3-tiered priority system for ECG alarms where life-threatening arrhythmias are designated as priority 1, urgent but non-life-threatening arrhythmias as priority 2, and technical ECG faults as priority 3. They then determined the frequency of each type of alarm and found that 18% were priority 1, 11% priority 2, and 71% priority 3. This allowed them to set the ECG arrhythmia alarm to only monitor for priority 1 and 2 events and filter out the remaining 71% as they determined it was not clinically relevant to the end user. This is a good example of a clinically driven consensus approach to determining the relative importance of different types of alarms in our case ECG arrhythmia events. Certainly, the highest priority should be given to those alarms which are of greatest clinical importance and those for which there is a significant impact on patient outcome if they are missed. This is often not the case with certain technical arrhythmia alarms or some physiological parameter threshold crossing alarms, but may be more relevant to monitoring for ST segment changes that signify myocardial ischemia or alarms related to respirator dysfunction in postoperative patients. In these cases, it is important to have an explicit list of what is being monitored and why it is important to the end user. This information can often be summarized on an ‘alarm criteria sheet’ which can aid subsequent decision making on which alarms can be filtered and which must remain with default settings. An example of a respirator alarm criteria sheet is provided by Cvach who suggested that a team may need to create a scoring system to determine the acuity or severity of the alarm and its potential impact on the patient before determining whether it is a candidate for filtering.
Regular Maintenance and Testing
Regular maintenance and threshold testing is essential in upholding a functioning and efficient alarm system. It is important to constantly assess the numerous changes in unit environment, staffing, and patient conditions to determine the impact on clinical workflows and alarm system management. Regular testing will also identify whether alarm system changes are required and ensure that any changes do not impair the overall alarm system effectiveness. According to the ECRI institute, one of the top 10 patient safety concerns for healthcare organizations is the testing of alarm systems to ensure that it is supporting patient safety. They recommend the following as a general principle guideline to adhere by: The alarm system’s ability to detect a clinically significant event, the effectiveness, and the impact on clinical workflow. There are many alarm signals that are not routinely checked that serve a vital function for the safety of the patient. Regular maintenance and testing will identify whether these alarms are being heard or seen, help determine that there is a qualified responder for the event, and whether the response is timely and appropriate. By continuous testing of the alarm system, it is also possible to devise correlations between certain alarm types and adverse patient events which will provide evidence to set a specific priority and threshold for that alarm ensuring that it is heard and acted upon by the right person at the right time.
Continuous Improvement and Evaluation
While alarm settings are being adjusted, it is important to monitor the effect it has on the ability of the alarm to capture the event and the incidence of false alarms. This can be done by comparing alarm data before and after a setting change, and by directly observing certain types of alarms at the bedside. Any changes that are found to be detrimental can simply be reversed. Settings that are found to be beneficial can be made permanent and incorporated into alarm management policies.
As this is done, it is important to continuously monitor for any unintended consequences, often with the Plan-Do-Study-Act cycle. Changes to alarm settings should be done conservatively, as each change represents a new untested alarm. There may be a trade-off between alarm sensitivity and specificity for capturing a given event. Increasing the threshold for an arrhythmia alarm, for example, may reduce the number of false alarms, but it also may delay detection of the arrhythmia. This should be done with a thorough understanding of the clinical context and input from those directly involved in patient care.
The continuous evaluation of the safety and efficacy of the alarms being used in the clinical environment is essential for making future improvements in alarm system design and for ensuring patient safety. This task can be complicated by the sheer number of alarms that occur on a given day, as well as the slow rate at which adverse events occur as a result of false alarms or the inability of a particular alarm to effectively capture a true clinical event. The first step toward improving alarm systems is the identification of the false or non-actionable alarms and the events surrounding those alarms. Once the frequency and type of non-actionable alarms are known, changes can be made to the alarm settings to reduce the number of false alarms.