While patient monitoring has saved many lives, the majority of alarms in hospital rooms are clinically insignificant, causing nurses to develop ‘alarm fatigue’. The Cleveland Clinic in Ohio, US, is addressing this head-on by monitoring patients off-site. Dr Daniel Cantillon tells Abi Millar how centralised monitoring can more effectively use hospital resources and improve patient care.
Alarm fatigue is a major problem in hospitals. It refers to what happens when nurses become desensitised to the alarms in the hospital room, and can lead to longer response times or even to alarms being missed entirely.
The upshot is that patients may fail to receive the urgent care they require, which is particularly troubling in the case of cardiac arrest. According to the American Heart Association, less than a quarter of adults survived in-hospital heart attacks between 2012 and 2016, when they occurred outside the intensive care unit.
“Alarm fatigue occurs when a high volume of inactionable clinical alarms drown out [a] true signal from a patient in distress,” says Dr Daniel Cantillon of the Heart and Vascular Institute at the Cleveland Clinic, Ohio, US. “It is important to hospitals all over the world, because studies show that more than 90% of telemetry alarms are inactionable and yet nearly 40% of cardiac arrests are not detected appropriately.”
More monitors, more problems
Concerned about the implications for patient safety, the US Joint Commission introduced a new National Patient Safety Goal in 2014. Since then, hospitals have been required to develop alarm management policies and establish alarm safety as a priority.
While this new requirement boosted awareness of the problem, it didn’t point up any particular solutions. As the Joint Commission stated, “It is important for a hospital to understand its own situation and develop a systematic, coordinated approach to clinical alarm system management.” This means it falls to each hospital to tackle the issue on its own terms, finding ways to minimise non-urgent beeps and buzzes.
As many researchers have identified, one of the chief contributors to alarm fatigue is the overuse of telemetry. On the face of things, it might seem like more is better when it comes to monitoring – if every patient is hooked up to a monitor, then theoretically speaking fewer incidents should fall through the cracks. In practice, however, indiscriminate monitoring can create more problems than it solves. Since the vast majority of alarms are clinically insignificant, they can lead to distraction at best, and systematic desensitisation at worst.
This situation is widely recognised. In 2013, the Society of Hospital Medicine advised hospitals, “Do not order continuous telemetry monitoring outside of the intensive care unit without using a protocol that governs continuation”. In other words, if they want to use their resources appropriately, hospitals need to be selective about which low-risk patients they monitor.
Taking this advice on board, the Cleveland Clinic decided to develop standardised criteria for putting patients on telemetry. The criteria, rolled out in 2014, served to reduce the numbers being monitored without any increase in cardiacrelated adverse events.
“The standardised criteria were derived from our own data to encourage important use of cardiac telemetry, which is an important resource,” says Cantillon. “Telemetry over-use contributes to alarm fatigue. Our study showed that applying standardised criteria resulted in an immediate and sustained 15% reduction in patients on telemetry.”
While there's nothing new about using telemetry criteria, the Cleveland Clinic took the novel approach of monitoring these patients off-site. Rather than leaving the work to nurses, the hospital used a central monitoring unit (CMU).
“The CMU is an off-site ‘bunker’ command centre, providing eyes in the sky to guide the cogs on the ground,” says Cantillon. “Our central station can detect emergencies within an hour before they actually happen in about 80% of cases. When we invoke the discretion to directly engage emergency response teams containing a dedicated nurse, physician, and respiratory therapist, we see an astounding 93% survival rate for in-hospital cardiac arrests.”
A look at the data
These figures, which come from a Journal of the American Medical Association (JAMA) paper published in August 2016, are based on 13 months of data. Between March 2014 and April 2015, the hospital followed a pool of low-risk patients, who met the standardised telemetry criteria. These patients received continuous cardiac monitoring, along with blood pressure, pulse oximetry and respiratory rate notifications on request.
While the patients were based at the Cleveland Clinic and three regional hospitals, all the monitoring was conducted at the CMU. For every 48 patients, there was one dedicated monitoring technician, along with on-site technicians providing oversight.
Over the 13-month period, the CMU received electronic telemetry orders for 99,048 patients and provided 410,534 notifications to the nursing team. The emergency response team was engaged 3,243 times, in responses to changes in patients’ heart rhythm or heart rate.
Every so often (105 times to be exact), the CMU provided the emergency team with discretionary notifications that a patient’s condition was worsening – 27 of those patients went on to have a heart attack, and owing to the emergency team’s readiness, 25 of them (93%) survived.
The night’s watch
These figures are undeniably positive, a resounding endorsement of the Cleveland Clinic’s approach. Cantillon says the results were in line with expectation. “When improving the ratio of true signal to background noise, you can better identify and focus on the patients requiring immediate attention, to prioritise their needs. This confronts the problem of alarm fatigue,” he says.
He adds that centralised monitoring offers advantages over monitoring on-site. “These include removing monitoring personnel from the noise and distractions of normal hospital activities, applying standardised practices, and improving staffing efficiency and oversight,” he says. “We have watchers over the watchers, much like the way pit supervisors observe and manage card dealers in a casino.”
It’s important to recognise that this system does not replace bedside care. Even the best monitor is no substitute for frequent direct observation. Rather, patient monitoring is regarded as a shared responsibility between the CMU and nurses. Communicating frequently, both sides are held accountable for safety.
Since the 2016 JAMA study was published, the Cleveland Clinic has worked to improve efficiency and cut costs.“Over the past two years, we’ve come even further by applying innovative technology called the eCMU,” says Cantillon. “This is a novel risk stratification system that I invented and we developed in-house. It more than doubles the number of monitored patients per human technician, while improving efficiency throughput on communications by 30%, and yielding even better early detection of crisis events one hour before they occur – which is now approaching 90%.”
With this new platform, patients are riskstratified and alerts are generated when their risk crosses a certain threshold. Their telemetry information is automatically synced with their electronic medical record and the monitoring technician reviews that information before notifying the nursing team. Rather than watching 48 patient waveforms, the technician can focus on a single patient.
At present, the Cleveland Clinic is about halfway through the process of phasing in this new system, and phasing out the old CMU. “It will be fully replaced by the end of 2018 for the Cleveland Clinic main campus and seven of our regional hospitals, which includes our facility in Florida,” says Cantillon. “From there, we aim to publish our data. Our leadership may one day consider deploying the eCMU to our hospitals in the Middle East and Europe.”
He feels that the Cleveland Clinic’s work ties into trends in digital medicine. “You can think of this as one way that providing care at a distance is transforming healthcare all over the globe,” says Cantillon. “We can mix and match our Florida monitored patients with those in Cleveland, Ohio and connect directly with their respective nurse by a click of a button. In some cases, our technicians may barely notice where that patient is physically located.”
The JAMA paper highlighted one further area for future improvements: namely, the monitoring technology itself. Of the 410,534 calls made by the CMU, a significant minority were due not to patient arrhythmias, but to connectivity failures. This suggests that current telemetry systems are not always as reliable as one might hope.
“Future studies can use our descriptive data as a benchmark, given [the] limited scientific data to date, and should evaluate new technologies with better skin adherence, such as patch-based monitors,” wrote the authors.
To date, centralised monitoring systems – combined with standardised telemetry criteria or patient risk stratification – remain a relatively new and untested topic. However, as the Cleveland Clinic amasses further data, their benefits will likely become more apparent.