How the algorithm that predicts death works

How the algorithm that predicts death works


Imagine if you could know that you are going to suffer a heart attack hours before it happens.

This – which sounds like science fiction – today is already possible. At least for some patients who are interned in hospitals in the United States.

The Food and Drug Administration (FDA) has just approved the first algorithm that can predict sudden deaths from cardiorespiratory events.

It is a software that analyzes the data of patients that are monitored in hospitals and calculates the risk of suffering from either a heart attack or a respiratory failure.

The algorithm, which was developed using the medical information of thousands of patients , can detect any of these episodes up to six hours before they occur, and alerts doctors and nurses to prevent them.

“Nowadays, health services are going through a perfect storm: people live longer but have more chronic illnesses, while many of the more experienced doctors are retiring,” says Lance Burton, general manager of ExcelMedical. technological medical company that developed the algorithm.

“Patients are monitored but many times doctors and nurses react only when a catastrophic event occurs, not before, ” he tells BBC Mundo.

The reality is that in many hospitals the staff can not cope with so many patients, but even if there were people dedicated exclusively to collect and analyze the vital signs, it is unlikely that they could anticipate all the problems.

Patient monitored in hospital
Copyright of the ISTOCK image
Image caption The algorithm analyzes all the information that comes out of the monitoring equipment connected to the patients. Does not require additional equipment.

“Humans can not process all the information that comes out of a patient and it becomes very difficult to know when it will deteriorate,” explains Burton.

What the system created by ExcelMedical – called WAVE Clinical Platform – does is to cross- link all the information and ” anticipate patterns in the information that are not recognizable to the human eye”.

Third cause of death

According to a study by Johns Hopkins University, 10% of all deaths occurred in the USA. they are due to medical errors.

And these unexpected or preventable deaths that occur in hospitals, often due to the inability of health personnel to detect, communicate and act in the face of signs of clinical deterioration, represent the third cause of death in the country.

The creators of the WAVE platform believe that many of the more than 250,000 people who die this year, according to Johns Hopkins, could avoid death.

The system is very simple to use: it does not require special machinery or new hardware . It simply uses existing monitoring systems in most hospitals.

The software analyzes five key variables : heart rate, respiratory rate, blood pressure, body temperature and oxygen saturation.

Through an app that can be viewed through a smartphone , a tablet or a computer, health personnel can access all those vital signs, in real time, of each of their patients.

WAVE Clinical Platform
Copyright of the EXCELMEDICAL image
Image caption Healthcare staff can see all the vital signs of the patient through an app that they can look at their phone or computer, anywhere.

That is, they can monitor them without being by their side. You do not even have to be in the hospital.

But the great novelty of WAVE is its unique algorithm, whose official name is the Visensia Safety Index .

This algorithm – the first of its kind to receive the approval of the FDA – crisscrosses all the information about each patient and quantifies their level of risk in a simple scale ranging from 0 to 5.

Thus, doctors and nurses have only to check a simple number for each patient to know in what state they are.


If a patient exceeds number 3 on the scale, the system automatically sends an alert for the medical team to intervene.

WAVE Clinical PlatformCopyright of the EXCELMEDICAL image
Image caption This is how the WAVE application looks. Each box is a patient and the number calculated by the algorithm (ranging from 0 to 5) indicates the level of risk. Above 3 is red and sends an alert.

In the clinical tests carried out by ExcelMedical at the Medical Center of the University of Pittsburgh (UPMC), the success of the system was resounding.

Among patients who did not use the WAVE there were six unexpected deathsduring a period of eight weeks.

With WAVE, there was no unexpected death during the same period.

Although the system sounds enormously promising, at the moment it can not be used outside the setting of an intensive care room.

However, the creators of the platform also hope some day to develop a system that anticipates and prevents cardiorespiratory attacks in the general population.

Another limitation of WAVE is that it can not anticipate strokes , another of the most frequent causes of disability and death today.


On the other hand, Burton recognizes BBC World that not all patients who are detected an imminent heart attack or respiratory failure can be saved.

“There are patients who enter a hospital and we know they will not leave,” he says.

Patient in hospital
Copyright of the ISTOCK image
Image caption The system is designed to save the lives of those who suffer unexpected problems. Medicine still can not cure everyone.

“We want to specifically help those who die in an unexpected way, who suffer a deterioration either due to complications after a routine operation or due to the effect of certain medications,” he says, as an example.

But the WAVE does not intend to guarantee immortality .

“The only incurable disease is life,” says Burton.

However, the employer says they are working on developing several algorithms that could anticipate – and therefore prevent – causes of known deaths.

“Our next project is about septicemia, ” he reveals, referring to the infection that affects some 20 million people a year in the world and kills about 8 million.

About author

Rava Desk

Rava is an online news portal providing recent news, editorials, opinions and advice on day to day happenings in Pakistan.


Leave a Reply

Your email address will not be published. Required fields are marked *

Your email address will not be published. Required fields are marked *