Cut Bedside False Alarms Up to 99% with Real-Time Analytics

March 4, 2019 Robert Moss

Who doesn’t like a good hospital drama? From Grey's Anatomy to General Hospital, they’re longtime favorites with viewers. And just as in real life, patients are hooked up to a bevy of devices—pulse monitors, ventilators, IV bags, and possibly a dozen more.

But viewers would be surprised if bedside scenes were realistic—with the nonstop distraction of real-life false alarms—endured by patients and healthcare workers alike.

How many alarms are we talking about? As an example, at Johns Hopkins Hospital the number is astounding. The 1,000-plus-bed facility in Baltimore, Maryland reported an average of 350 alarms per patient each day. And research has shown that overall in the U.S., up to 99% of alarms are false.

These numbers explain the high rate of alarm fatigue, in which doctors and nurses become inured to all of the warnings and alerts—no longer taking them seriously—even when a patient may be experiencing a life-threatening condition. The problem is so dire that the ECRI Institute labeled alarm fatigue as a number-one medical hazard.

While alarm fatigue is nothing new, it’s only getting worse. What’s changed? Technology is propelling use of monitoring devices, with an average of 10 to 15 per bed. A shortage of healthcare workers, combined with complex regulations, is increasing the stress on nurses, doctors, and their patients.

Patient monitoring devices are intended to improve medical outcomes. This can happen only when their data is integrated, analyzed, and acted on in real time. But constant monitoring can create havoc if systems are not properly managed.

Continuous Surveillance Goes Beyond Patient Monitoring

In essence, these separate medical devices capture and process a wide range of medical data from very sick patients and those recovering from major surgeries.  But considered separately, the data can be overwhelming and difficult to take action on. To improve outcomes, clinicians and caregivers must receive timely alerts related to the patient’s condition—while avoiding false alarms—so that they can swiftly perform their duties.

A continuous-surveillance solution brings the data from these different devices together to help clinicians recognize a patient’s physiological trends. By pairing algorithms and predictive models at the edge with vital-sign data, changes in patterns can be detected. This includes:

  • Automatically making sense of multiple data types such as heart rate, respiration rate, blood pressure, skin temperature, and more.
  • Capacity to analyze changing patterns in clinical data over time.
  • Applying rules-based data analytics to signal unexpected changes before a life-threatening situation occurs.
  • Flexibility to network with any device and adapt to a wide range of clinical protocols.

Bernoulli Health has devised a system that provides this type of continuous surveillance. The Bernoulli One predictive clinical surveillance platform collects medical device data and runs analytics on that information as shown in Figure 1.

Figure 1. The Bernoulli clinical surveillance platform integrates with a wide range of devices to process data at the edge.

Real-Time Analytics at the Edge

The solution’s middleware detects pattern changes in oxygenation levels, heart rate, pulse rate, carbon dioxide level, and other metrics. It then applies that data in milliseconds to rules-based phenomenon algorithms.

Bernoulli’s low-latency technology helps provide clinicians with advance warning, so they can intervene and stabilize a patient as soon as deterioration is detected.

“What makes us different from basic alarm management systems is that we’re using real-time patient physiologic parameters coming from one or more devices,” said Janet Dillione, CEO of Bernoulli. “If that device is speaking every second, we're getting it every second. If it talks every minute, we see it every minute.”

Helping Patients Recover

Major-surgery patients are often prescribed opioids for pain management, putting them at risk for opioid-induced respiratory depression (OIRD), in which a patient simply stops breathing.

“More than 20,000 patients per year experience OIRD,” said Dillione. “And opioids—while a powerful way to reduce chronic pain—are the cause of half of all medication-related deaths in hospitals,” Dillione added.

Furthermore, the enormous economic impact of OIRD accounts for nearly $2 billion a year in U.S. healthcare costs.

A recent clinical study showed that the Bernoulli One platform reduced the number of OIRD false alarms by 99%. And the system enabled alerts 100% of the time when patients were approaching an episode.

“For this customer, in the first 30 days after Bernoulli went into production, they had reduced 23,000 alarms down to 200,” said Dillione. “We not only reduced the noise by 99%, but we've also identified the patients who are at tremendous risk.”

These remarkable results show how IoT technologies can have huge impacts on improving outcomes. When healthcare workers get the information they need, when they need it, patients are more likely to recover from major surgeries.

Flexible and Secure

Bernoulli One is a flexible and scalable solution, making it simple to configure without customization or rewriting code.The platform, which runs on Intel® technology, including Intel® Core and Intel® Xeon® processors, is FDA Class II approved for patient monitoring and secondary alarm management.

On top of this, the Bernoulli system adheres to the National Institute of Standards and Technology (NIST) Cybersecurity Framework, enabling hospitals to:

  • Meet HIPAA compliance by providing security safeguards that ensure the integrity of electronic protected health information (ePHI).
  • Assess and mitigate cybersecurity risks associated with malware, ransomware, and other threats to patient data.
  • Identify and validate modifications to the platform and medical device security logs and settings.

“We take security very seriously. We have an ability to provide auditing for any change that's made on any device on any parameter,” said Dillione.

False alarms create stress for patients and hospital staff, and are clearly a significant problem. They interfere with the ability of clinicians to provide proper care to patients, often with serious consequences.

The bottom line is that the Bernoulli One platform has been shown to reduce alarms by up to 99 percent, improving patient safety as well as bringing a better environment to healthcare workers, making their jobs a lot easier.

About the Author

Robert Moss

Robert Moss is an independent consultant and strategist who focuses on the value gained through IoT, AI, machine learning and other technologies. He also helps give voice to executives at leading technology companies, enabling their personal stories to show how they encourage innovation, overcome obstacles, and improve their leadership skills. Tweets @RobertMoss_IoT

More Content by Robert Moss
Previous Article
Why embedded world Moved Beyond Tech
Why embedded world Moved Beyond Tech

This year, embedded world was about solutions, not technology. But tech still played a big part, with AI, T...

Next Article
Pump Oilfield Data into Central Offices with Edge Analytics
Pump Oilfield Data into Central Offices with Edge Analytics

Remote and rugged oilfield operations can get a boost from IIoT and edge computing for remote monitoring, a...