Adding a Second Layer of Safety to the Care of Children with Critical Heart Disease with a Pediatric Tele-CICU Model: A Pilot Study of Virtual Surveillance Activities Aided with an Artificial Intelligence Tool

Objectives: The occurrence of dangerous events (i.e., cardiac arrest, emergent endotracheal intubation, and ECMO cannulation) is still very common among children with critical congenital and acquired heart disease despite refinements in surgical techniques, training of bedside team, and traditional quality interventions; hence, innovative ways are needed to advance safety during the care of this very sick children. An additional layer of safety can be added with virtual surveillance enhanced by AI tools, as well as targeted communications delivered from a tele-CICU to the bedside team. We describe here our initial experience with our own CICU.

Methods: Initial tele-CICU operations started in 9/4/2019. Virtual surveillance activities with our own CICU patients included review of remote monitoring, video feed in patient rooms, electronic medical records, and imaging and were targeted according to risk stratification based on patient daily census and an artificial intelligence tool displaying physiological trends and an algorithm trained to follow oxygen delivery (Etiometry). A systematic approach was used aiming to support escalation of diagnostic or therapeutic interventions whenever granted, with a focus on physiological trends and residual disease. Communications with the bedside team were established via secure text messaging, phone call, or virtual robotic tele presence pending on acuity level of information to be discussed. Additional activities at the request of the bedside team were introduced in July 2019. Tele-CICU activities were prospectively recorded in a custom-made electronic database.

Results: Between January 1 and August 31, 2019, there were a total of 1,827 virtual surveillance activities including 1,251 remote monitoring (surveillance with no communication with bedside team), 538 noncritical communications. Thirty-eight critical communications with direct voice conversation were conducted from May 1 to August 31, 2019.

Conclusion: A tele-CICU model has the potential to increase safety of patient care for children with critical heart disease, given the significant number of dangerous trends and additional clinically relevant information found during our virtual surveillance activities. Further refinement of workflows and research is needed to establish the most effective model of care.