Hospital’s early warning system lowers post-op mortality

By Jerry Zeidenberg

By combining smart technologies with an Early Warning Scoring (EWS) system in its surgical department, a large hospital in New York City has reduced ‘failure to rescue’ deaths by a whopping 35 percent. Failure to rescue patients are those who have experienced a treatable complication and died in hospital.

It’s an astounding achievement, and Dr. Ron Kaleya, director of GI surgical oncology at the Maimonides Medical Center, in Brooklyn, N.Y., recently spoke at a Philips-sponsored webinar about how the hospital accomplished this.

“We believe we have built a safer surgical service,” said Dr. Kaleya, who was part of the team that devised the Brooklyn Early Warning System, or BEWS. The project was started in 2014, and involved Philips as a partner.

He noted that the BEWS crew at Maimonides Medical Center created a proactive early warning system to catch deteriorating patients before it was too late. The team automated the collection of six vital signs, and produced an algorithm that determined when experienced clinicians should be notified.

Further, they automated the notification process by sending wireless alerts using the Philips IntelliSpace Event Management (IEM) System to the smartphones of clinicians, so the right person could respond in a timely manner, without the delays of voice and pager systems that have traditionally been used in the healthcare environment.

According to Philips, the IntelliSpace Event Management (formerly Emergin) helps care teams manage and respond efficiently to multiple event notifications, alerts and alarms by providing useful information directly to communication devices, improving efficacy.

Numerous hospitals have created early warning systems, but they haven’t tied them to an automatic alerting system. As a result, they haven’t experienced an improvement in outcomes.

“It’s what you do with the score that matters,” said Dr. Kaleya. “You may have a parachute when you jump out of a plane, but unless you pull the ripcord, nothing happens.”

It wasn’t easy to create this system, said Dr. Kaleya, who noted that much of the computerized monitoring and communications equipment was technically incompatible and wouldn’t talk to each other.

It included the Philips spot check vital sign monitors and Philips Guardian Solution, Voalte phones and servers, and the Allscripts electronic health record system. The BEWS team urged the vendors to integrate the various pieces of the system, so that seamless communication could take place – a process that took about a year.

Moreover, the team had to determine what signs to monitor, how they should be weighted, and at what point experienced physicians and nurses should be alerted.

The six vitals the team decided to collect are blood pressure, oxygen saturation, pulse, respiration rate, temperature and mental states. They’re collected every four hours, reviewed by a nurse, and then sent to a server; if a certain threshold has been reached, wireless alerts go out to care providers.

“Most patients have scores less than three,” said Dr. Kaleya. “Patients with scores of up to four can be stabilized with fluids and oxygen therapy, and they can continue to be managed on the ward.” Patients with scores of five or more tend to require more urgent care, sometimes including transfer to the intensive care unit.

Initially, the BEWS team decided that it would alert teams of experienced clinicians when post-op patients hit a score of five. This method was used in 2014 and 2015 without much change in mortality outcomes.

“When analyzing our 2015 results, we were unhappy,” lamented Dr. Kaleya. “There was not the improvement we had predicted. We had chosen the activation threshold of five – and this was just wrong.”

He asserted that the BEWS escalation trigger was then reduced to four – and a dramatic change occurred.

In 2016, there were 637 patients with a score of four or more, and rapid notification of care providers occurred. Only 21 of these patients subsequently died, and some of them were palliative care patients.

“Of 15 patients who were transferred from the floor [to the ICU], 14 were rescued and left the hospital,” observed Dr. Kaleya.
“Most of the others only needed adjustments in fluids, oxygen, diuretics and antibiotics.”

The hospital’s internal record-keeping showed a 30 percent drop in failure to rescue patients in 2016. “We didn’t’ really believe our results, as they were our own internal calculations,” said Dr. Kaleya. However, the New York City patient indicators for 2016 confirmed the breakthrough – there was actually a 35 percent reduction in risk-adjusted mortality.

“External validation is a wonderful thing,” said Dr. Kaleya.

He also noted that BEWS is cost-effective. It requires an investment of $850 per patient in hospitalist staffing – hospitalists are used to check on the status of patients with rising BEWS, but with scores less than the trigger threshold.

However, about $900 per patient was saved in the reduction of complications, as the deterioration of patients was caught at an earlier stage.

This alone pays for the cost of the hospitalists, he said.

There is also an investment in technology, of course. Dr. Kaleya noted that $250,000 was spent on the technology required, including 10 vital signs monitors at $10,000 apiece. There was a large expenditure on communication systems, to perfect the integration and alerting solution.

He said to extend the solution to additional wards, each with 70 patients, would cost $150,000 per ward.

At the same time, there are additional benefits. BEWS results in a shorter Length of Stay (LOS) for patients and reduced liability costs. It also produces better medical outcomes and patient satisfaction.

The BEWS project has been so successful at Maimonides Medical Center that the hospital now intends to expand the system to all parts of the hospital.

Moreover, the plan is to become even more proactive, and to start preemptive interactions at the BEWS score of three. This will catch complications at an even earlier stage.

Reflecting on the original project, Dr. Kaleya explained that the success of surgeries depends as much on the level of post-operative care as on what happens in the operating room.

He noted that in the past, in hospitals without early warning systems, nurses have often been afraid to escalate and to notify an experienced physician when post-operative patients deteriorate.

When they did notify a doctor, it was often an intern or resident they contacted. “The most inexperienced person was asked to assess the sickest patients. It just didn’t make sense,” said Dr. Kaleya. These inexperienced doctors were in turn reluctant to call their superiors, fearing they would be deemed inadequate.

As a result, patients on the verge of crashing often didn’t receive the care they need.

A 2005 article in The Hospitalist journal estimated that 60,000 American Medicare patients under the age of 75 die each year in failure to rescue situations – episodes in which they have developed treatable complications, but did not receive appropriate care.

The traditional model of care leading to this situation has been dubbed ‘track and trigger’. Nurses and other clinicians track patients on the floors, and determine when to trigger an escalation alert.

“This model is, and always was, a complete disaster,” said Dr. Kaleya. “It had to go, and it did.” More hospitals are now using early warning systems as a way of escalating the alerting process at an earlier stage.

However, as Dr. Kaleya noted, there are still problems, as the early warning systems used in most other hospitals haven’t been automated. Nor have they been tied to an alerting system.

Studies have shown, he said, that when EWS scores are calculated by hand, only 20 percent are correctly tabulated. Further, the same resistance to escalate occurs when nurses and residents have responsibility to alert more experienced clinicians.

With BEWS, however, the vital signs are automatically collected and scored. When the threshold is reached, moreover, wireless alerts are sent out to experienced care-givers, without human intervention.

“Immediate wireless alerts are generated, according to the algorithm, without clinical input,” said Dr. Kaleya. Experienced caregivers respond quickly, resulting in saved lives. “BEWS delivers the right care, right now,” said Dr. Kaleya.

Written by Editor

2 Comment responses

  1. Avatar
    August 09, 2017

    It sounds like the cost is worth the results. Hope more hospitals will introduce this method and save more unnecessary deaths. Way to go Brooklyn.


  2. Avatar
    March 16, 2018

    sounds great and amazing. hope this saves a lot of lives.
    I just wondered however, what if there is connectivity problem or failure to alert problem due to some malfunction of the system. Will that costs lives? because when humans depend so much on technology to alert us to act, what if this technology one day fails to work or malfunction suddenly?


Leave a comment