Public Release: 

Procedural changes may reduce cardiac arrests outside pediatric ICUs

American Heart Association meeting report

American Heart Association

WASHINGTON, May 10 - A simple procedural change may significantly reduce cardiopulmonary arrests outside a pediatric hospital's intensive care unit, according to a study reported at the American Heart Association's 8th Scientific Forum on Quality of Care and Outcomes Research in Cardiovascular Disease and Stroke.

Cardiac arrests outside the Pediatric Intensive Care Unit (PICU) at Children's Medical Center Dallas dropped by 63 percent after doctors and nurses changed procedures to call the cardiac emergency team or Code Blue Team, whenever warning signs indicated a patient's lungs or heart might stop working. Previously, staff summoned the team only after starting cardiopulmonary resuscitation (CPR) after a cardiac arrest.

"Our goal was to get patients moved to the PICU and a higher level of care sooner, hoping to prevent an arrest," said Tia A. Tortoriello, M.D., senior author of the study, assistant professor of pediatrics in the Division of Cardiology at University of Texas Southwestern Medical Center at Dallas and medical staff member at Children's Medical Center Dallas. "And if we couldn't prevent it, the arrest could be better controlled in the PICU."

Cardiac arrest is different in children than in adults. In adults, heart stoppages mostly result from a heart attack or electric malfunction of the heart. Children rarely have a primary cardiac arrest. Most often, their arrests evolve from respiratory problems, researchers said.

Tortoriello joined the 411-bed hospital's CPR Committee as co-chair in late 2003. She became convinced that the data it had collected could lead to better ways to reduce the number of patients who suffered a full cardiac arrest outside the PICU.

Tortoriello credits the American Heart Association's National Registry of Cardiopulmonary Resuscitation (NRCPR) with enabling Children's Medical Center to shift its response focus from reaction and resuscitation to earlier intervention and prevention.

The registry collects data from many member hospitals, which allows individual institutions to compare their rates of cardiopulmonary arrests and resuscitations with hospitals of similar size and types of patients. They can also evaluate their Code Blue Team's methods to determine how to improve them.

Tortoriello and her team found that the hospital's cardiac "hotspot" -- the non-critical care area where a cardiopulmonary arrest was most likely to occur -- was the chronic vent floor, where patients are maintained on ventilators.

"Patients may acutely plug their tracheostomy or have significantly increased respiratory issues," she said. Either may cause "a respiratory arrest that can progress to a cardiac arrest." After analyzing data and obtaining hospital approval, the CPR Committee made a symbolic change. It renamed the Code Blue Team to the Pediatric Advanced Life Support (PALS) Team. Team members and their duties remained the same.

"The substantive revision of when to call for the PALS Team was harder to implement," Tortoriello said. "It was difficult to get people to change their mindset from waiting to call for help until the patient had arrested to calling at their earliest concern for the patient."

During the first three months of 2005, she and her colleagues instructed hospital staff about the early warning signs of a potential cardiopulmonary arrest in children. These include increased respiratory rate, labored breathing, increased oxygen use, excessive sleepiness and irritability.

They later compared data gathered from October 2003 through March 2005, and from April 2005 through November 2006. They found some positive results -- and one that puzzled them:

  • The number of cardiopulmonary arrests outside the PICU dropped from 68 over a total of 237 events to 30 over a total of 274 events.

  • Arrests in noncritical-care areas declined from 2.7 arrests to 1 arrest per 1,000 patients discharged, a decrease of 63 percent.

  • Patients needing escalation of care -- those who had not suffered an arrest but whose distressed condition benefited from the expertise of the PALS Team -- increased from 3.0 patients per month in the first period to 3.4 per month in the second timeframe. That difference was not statistically significant.

  • The PICU mortality rate dropped from 4.5 percent in the first period to 4.4 percent in the second, a difference not statistically significant.

The researchers did not track what happened specifically to patients arriving earlier at the PICU under its new system. However, the stability of the PICU death rate surprised them.

"If you recognize these patients sooner, get them to the PICU, and intervene before something bad happens, you would expect to see a decrease in our PICU mortality rate," Tortoriello said. "However, a change in the hospital's status might explain the finding, at least in part."

During the study, Children's Medical Center Dallas was designated a Level 1 pediatric trauma center -- the only one in Texas. Level 1 means a hospital meets strict national standards for the quality of its care in a specific area of medicine. "It may be that we are now seeing more children and sicker children in our PICU," Tortoriello said. She and her colleagues continue to monitor the cardiac arrest data, seeking better ways to improve preventive care.

"It's still a work in progress," Tortoriello said. "The goal is to not have a single arrest outside of the PICU."


Co-authors are Jean Storey, R.N. and Susan Wilhite, R.N.

Statements and conclusions of abstract authors that are presented at American Heart Association/American Stroke Association scientific meetings are solely those of the abstract authors and do not necessarily reflect association policy or position. The associations make no representation or warranty as to their accuracy or reliability.

Note: Presentation time is 5- 6:30 p.m. EDT, May 10, 2007.

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.