When Elizabeth Seton Children’s Center set out to expand its ventilator program, the team had one overriding goal: keep every resident safe while distributing medically complex children across a larger facility footprint. What followed was a journey that transformed clinical surveillance from an alarm management tool into the backbone of individualized care.
During a recent webinar hosted by Becker’s Healthcare and Philips Enterprise Informatics, Yelena Yadgarova, director of respiratory services, and Christopher Moore, director of quality at the Yonkers, N.Y.-based pediatric long-term care facility, shared how their organization built a data-driven care model — and what other healthcare leaders can take from it.
A care environment unlike any other
Elizabeth Seton Children’s Center serves children and young adults with complex, technology-dependent medical conditions. Of its 169 residents, 158 have active pulse oximetry orders, 102 receive some form of mechanical ventilation and 121 have tracheostomies. Unlike an acute care hospital, the facility is the residents’ home — they live in neighborhoods, go to school and celebrate birthdays despite significant medical needs.
As the ventilator program grew too large to cohort on one or two units, the team made a deliberate decision not to relocate residents based on acuity. That choice required a new approach to clinical visibility.
“If a child would get sick, you wouldn’t take them out of their bedrooms,” Yadgarova said. “We needed to ensure that we had some sort of visibility on that unit.”
From noise to actionable insight
The team implemented a Philips-powered Capsule Surveillance platform initially to address alarm fatigue. With hundreds of ventilators and pulse oximeters connected, the volume of alarm notifications threatened to overwhelm staff — and in a decentralized care environment, a nurse managing four residents across different rooms couldn’t maintain line-of-sight monitoring.
The customizable dashboard, accessible from any computer in the facility, gave clinicians real-time visibility across their entire neighborhood. A respiratory therapist gowned up in one room could assess whether an alarm from a different floor required an immediate response or was self-correcting — directly reducing unnecessary room entries and PPE use.
But the team quickly realized the data could do more than manage alarms. Over a three-month period, Moore’s team analyzed desaturation data and identified two residents each generating roughly 10% of all desaturation alarms across the 169-resident facility.
The first resident had oxygen saturation parameters set tightly at 94% — calibrated conservatively at the request of a concerned family. The data showed the setting was creating significant alarm noise without improving safety. Through months of data-sharing conversations with the clinical team and family, the team reduced the threshold in stages to 92% then 90%, dramatically cutting alarm burden while improving the resident’s quality of life.
The second resident had a different root cause. A social, highly active child, his peak desaturation episodes aligned with times when staff were occupied with other tasks and couldn’t provide one-on-one engagement — leading him to hold his breath or interact with his ventilator tubing. The team worked with the interdisciplinary neighborhood team to build a structured engagement schedule that addressed the underlying behavior.
“He was really trying to communicate with us,” Moore said. “The data gave voice to what he was asking for.”
Objective data, shared decisions
The surveillance platform also shifted care conversations from subjective experience to shared evidence. Moore convenes monthly meetings where neighborhood-level alarm data is reviewed across nursing, respiratory therapy and clinical leadership — creating a common language for care decisions that holds across shift changes and individual staff observations.
“Prior to this, you’re relying on the reporting of staff,” Moore said. “What the data does is tell us this is what happened over this period of time — and this is information that is objectively true.”
The team also integrated trend data into ventilator weaning programs, identifying windows where residents consistently thrived off the ventilator and using that evidence to incrementally extend trial periods. And for one resident with neurological instability prone to bradycardia, the data revealed a consistent daily pattern that allowed staff to stage warming equipment proactively before the expected window.
Key lessons
For healthcare leaders considering a similar program, both speakers emphasized extensive pre-implementation testing — including working within a virtual environment to configure every threshold, alarm delay and dashboard element before go-live — as the most critical success factor. They also stressed the need to build a durable internal team committed to ongoing data review long after implementation ends.
“Find a way to harness the data you get, and think about how you plan to analyze and share it,” said Giridhar Poondi, NA regional marketing lead for Philips’ Acute Care Informatics portfolio and moderator of the session.This article is based on a webinar hosted by Becker’s Healthcare and Philips Enterprise Informatics.
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.
