Heed the indicators of in-hospital cardiac arrest

The key to reducing deaths from in-hospital cardiac arrest may be preventing them from occurring.


The rapid response nurse on call was skeptical when the alert appeared on her pager: A 21-year-old woman who had recently undergone a caesarean section was in danger of cardiac arrest, according to the community hospital's electronic early warning system. It was unusual for the team to be called to the postpartum unit, and the patient had no history of heart problems on admission.

Image by Thinkstock
Image by Thinkstock

At the bedside, the patient reported shortness of breath and her oxygen saturation was dangerously low. The rapid response team (RRT) shifted into high gear, initiating a transfer to the ICU, starting positive-pressure ventilation, and ordering a chest X-ray and electrocardiogram (EKG). The eventual diagnosis: Severe mitral valve regurgitation causing excessive fluid buildup in the lungs had triggered heart failure.

Fortunately, the problem was caught early—the patient received a valve replacement and was discharged a week later. However, clinical signs of in-hospital cardiac arrest (IHCA) often go unrecognized until it's too late, said Dana Edelson, MD, MS, a hospitalist at the University of Chicago Medical Center and lead developer of the electronic cardiac arrest triage (eCART) early warning score, a software tool marketed by AgileMD.

“In this case, the patient almost definitely would have died if her deterioration had not been caught quickly,” said Dr. Edelson. “Once a patient has a cardiac arrest, the likelihood of them leaving the hospital alive is only about 20%. If we want to have more survivors, we have to catch patients earlier and prevent arrest from happening in the first place.”

“Prevention is the most powerful intervention,” agreed Clifton Callaway, MD, PhD, professor and executive vice chair of emergency medicine at the University of Pittsburgh Medical Center. “The real value of the rapid response team is in being able to intervene before a patient deteriorates to cardiopulmonary collapse—once that occurs, even the most robust response often can't save the patient.”

Barriers and solutions

It can be challenging to identify inpatients at risk of arrest and get them help quickly. IHCA can and often does happen outside of the ED or cardiology wards, yet many facilities do not have hospital-wide training programs on appropriate staff responses. In addition, unlike outpatients, hospitalized patients often do not report overt symptoms, such as chest pain, dyspnea, or palpitations.

A study published in the April 4, 2013, Journal of the American Heart Association looked at inpatient versus outpatient ST-segment elevation myocardial infarctions (STEMIs) at the University of North Carolina Hospitals in Chapel Hill and found that insufficient training outside the ED and cardiology services helped explain why inpatients with STEMI waited longer to receive an EKG or angiography.

Even in hospitals with robust training programs, clinicians are sometimes reluctant to call the RRT when they see problems developing, said Raghavan Murugan, MD, MS, associate professor of critical care medicine and clinical and translational science and RRT director at the University of Pittsburgh Medical Center. Clinicians may feel that the situation doesn't warrant a full response or that they can manage on their own.

However, clinicians should be trained to err on the side of caution rather than risk letting a potential crisis slip through the cracks, said Dr. Murugan. “We encourage all hospital staff to activate the RRT the moment the thought of an emergency crosses their minds—even if someone falls and does not appear to lose consciousness, they should call us.” One important reason for early activation of the RRT is to quickly bring help, including physicians, nurses, respiratory therapists, phlebotomists, and EKG technicians, simultaneously to the patient's bedside, he added.

At the University of Pittsburgh, RRT calls are categorized as either crisis (condition “C”) or arrest (condition “A”) based on defined criteria, said Dr. Murugan. The criteria for condition C include a wide range of symptoms, including sudden-onset tachycardia, shortness of breath, bleeding, and fever. Condition A calls indicate an arrest or imminent arrest.

On a typical day, the team gets between seven and 10 condition C calls compared to a condition A call every few days, said Dr. Murugan. The goal of the two-tiered system is to prevent cardiac arrests from occurring by encouraging staff to call for condition Cs at the first sign of distress, even if they fear it may be a false alarm.

“We encourage all staff to call in condition Cs as early as possible, and we don't criticize anyone for making an incorrect call,” he said. “Condition As still occur, but we think we prevent many of them by intervening early. As a result our hospital mortality rate is far below the national average.”

Many hospitals use early warning scores, such as the Modified Early Warning Score (MEWS), to identify and stratify patients at risk of a cardiac event. MEWS assigns point values to various vital sign readings and combines them into a single score that is used to activate an RRT.

Such scoring systems are being successfully integrated into hospitals' electronic health records (EHRs), but simplicity limits their effectiveness, said Dr. Edelson in an article published online by Annals of Internal Medicine on May 16, 2017. More sophisticated tools are now being developed that incorporate additional data elements and analyze them in real time.

Dr. Edelson and her team have been working on and refining the eCART model for over a decade. The algorithm uses vital signs, lab results, and demographic data to calculate individualized cardiac arrest risk scores, she explained. It can predict the probability that a patient will need intensive care, experience IHCA, or die within the next eight hours and stratify patients from lowest to highest risk. The RRT nurse is automatically paged when a patient crosses to the high-risk category, Dr. Edelson said.

In a recent analysis by Dr. Edelson and her colleagues of over 30,000 postoperative inpatients, published Jan. 12 by Annals of Surgery, eCART was found to be significantly more accurate than MEWS in determining risk of IHCA. The authors noted that eCART's effectiveness is likely attributable to incorporating many more variables—33 versus five for MEWS—and using complex statistical modeling as opposed to expert opinion.

Although early warning and scoring systems that help identify at-risk patients hold great potential for improving care, implementing them can be complicated, said Dr. Callaway. In order to meet individual hospitals' needs, many tools require additional software and programming that must be built on top of existing EHR systems.

“There are logistical and financial barriers,” he noted. “We might adapt an early warning tool for our hospital, but we would then have to update it every time our EHR vendor updated our underlying system—and our adaptation may not be compatible with other hospitals that use the same EHR vendor but a different installation.”

In addition, automated warning systems may be more suited to small community hospitals than large academic centers that have residents and intensivists in-house 24/7 as well as extensive experience dealing with complex patients, said Dr. Murugan. “We tried one tool but found that we picked up on problems before the system was triggered.”

Once a team arrives at the bedside, organization and communication are critical to making things run smoothly. At the Hospital of the University of Pennsylvania in Philadelphia, a simple paper-based intervention aimed at defining team member roles led to less confusion at the bedside, which could ultimately improve survival.

The study was triggered by evidence that absence of leadership and clearly defined roles are associated with negative outcomes after IHCA, said the study's lead author Marion Leary, RN, MSN, MPH, director of innovation research at the Center for Resuscitation Science at Penn's Perelman School of Medicine. Before implementing the system, the research team observed videos of actual cardiac events at Penn, which gave them a clearer picture of what happened in the room, what different team members were doing, and any obstacles to accomplishing tasks efficiently.

The intervention consisted of appointing a nurse and physician leadership dyad and distributing preprinted identification stickers with predefined roles, such as “code recorde” or “CPR.” When a general code is activated, the physician leader focuses on the patient while the nurse leader selects a bedside team, assigns roles, and oversees CPR quality, said Ms. Leary. Those who do not receive a sticker leave the room.

“We've learned that leadership is very important. Someone needs to step up and take control,” she said. “Also, we were able to significantly decrease overcrowding in the room by deciding on an optimal team size and clarifying role definitions.”

Hospitalists' role

Hospitalists contribute to such systems in a variety of ways.

“In some hospitals, the hospitalist is the RRT point person,” said Dr. Callaway. “They help determine what the criteria should be for calling a physician to the bedside, and how and when to escalate care. Hospitalists are well suited to the role because they have situational awareness and are much more familiar with hospital resources than those of us in the ED.”

Hospitalists can also help prevent arrests from occurring by identifying patients at high risk and appropriately triaging them to the ICU or more intensive monitoring, said Rohan Khera, MD, a cardiologist at the University of Texas Southwestern Medical Center in Dallas.

It may also be helpful to assess whether individual patients should be targets for the RRT's interventions. “Hospitalists care for many patients who are near the end of life and have DNR orders or are poor candidates for CPR,” said Dr. Khera. “Identifying these patients helps us make better treatment decisions and avoid delivering CPR to patients who are likely to have poor functional outcomes.”

From a systems standpoint, hospitalists can highlight process-of-care issues that could be contributing to poor outcomes. Dr. Murugan holds a weekly code review meeting in which hospitalists and other clinicians discuss such issues.

“When we review events in detail, it becomes clear that condition As don't just happen—usually there were warning signs hours before an arrest occurred or poor communication that led to slower response times,” he said. “Those are the issues we really want to address and that hospitalists can help us identify early and thus prevent true cardiac arrests from happening.”

Ultimately, preventing IHCA must be an institutional priority that involves everyone on staff, said Ms. Leary.

“One reason our hospital has been successful in doing quality improvement projects is that we include the entire team,” she said. “Besides nurses and physicians, we involve respiratory therapists, pharmacists, security and transport services, and others. Many people have vitally important roles in improving response.”