FDA Approves Crisis-Predicting Algorithm to Save Hospital Patients From Early Death

Photo: AP

The FDA has approved a new clinical platform for hospital staff that uses an algorithm to predict and prevent sudden patient death, makers Excel Medical announced Monday. Called the WAVE Clinical Platform, the system monitors patient vitals and sends alerts to connected smart devices up to six hours before patient suffers a potentially fatal heart attack or respiratory failure. This is the first such algorithm to receive FDA approval.

As many as 400,000 people a year die in hospitals prematurely, according to a 2013 study in the Journal of Patient Safety. Excel’s algorithm, called the Visensia Safety Index, tracks the vitals of very sick patients and calculates their risk of falling into “early deterioration,” the six to eight hours preceding a potentially fatal cardiac event.


Excel Medical’s chief strategy officer, Mary Baum, told Gizmodo that WAVE saves lives by saving precious time.

“The typical rapid-response team in a hospital today gets about 15 minutes,” she begins. “It takes you 10 minutes to get to the floor. Now I have five minutes to make a difference to your life. [But] if I can give you six hours... could we intervene in a productive way and be able to change the outcome?”

The WAVE platform looks at patient’s vital signs, including heart and respiratory rate rate, blood pressure, and body temperature. As Baum explains, the relationship between different vital signs can indicate something’s wrong hours before any single vital system goes out of whack enough to trigger emergency alerts to medical staff. So a slight decrease in respiratory rate alone may not flag traditional monitoring software, but a decrease at the same time as a blood pressure spike would be an early warning sign of deterioration.

“It’s not just raw data,” she says. “You see the trend and the correlation between the data. There’s not singular data points. So the algorithm is about the correlation between the data points.”

Excel Medical

Patients are assigned a number between one and five, depending on how healthy they seem based on their vital data, and hospital staff are alerted if patients are ranked a three or above (this is considered the danger zone). Instead of sending in staff after unstable vital signs trigger an alert, the WAVE platform calculates an at-risk patient’s likelihood of early deterioration. The system is always on and sends alerts to connected phones, tablets, and desktops.


In clinical trials at the University of Pittsburgh Medical Center, two cohorts of elderly patients were compared, one with the WAVE system and one without. The control group had six unexpected deaths, while the trial group had zero.

Death prediction is an emerging field in medical technology and artificial intelligence. Another such company, Aspire Health, claims its algorithms can save families thousands by predicting the point at which elderly patients should shift from hospitals, where they receive ongoing treatment, to in-home palliative care, where they’re made comfortable but aren’t expected to recover. The families save money by foregoing expensive procedures that won’t buy much time.


Excel Medical told HCA News it will eventually pursue wearable devices, integrating WAVE with the software and adding the trove of biometric data they provide to their research.

[Healthcare Analytics News]


Share This Story

Get our newsletter

About the author

Sidney Fussell

Of course I have pages. I had pages five years ago. How anyone can believe I don’t defies belief.

PGP Fingerprint: 6595 DBB6 1A35 1D33 2459 1643 3315 A092 4BA2 05E1