Concise Critical Appraisal: Cardiac Arrest in the ICU: An Assessment of Preventability

An intensive care unit cardiac arrest (ICU-CA) often leads to death or severe neurologic dysfunction. In the initial treatment of a cardiac arrest, cardiopulmonary resuscitation is used but this is only a reactionary response to an event that has led to a catastrophic failure of the cardiopulmonary system. In an ICU, the enhanced degree of resources and monitoring gives us an ability to be proactive in treating patients before they have a cardiac arrest. It is the goal of the ICU clinician to obviate any unintentional or preventable cardiac arrest in the ICU. Multiple studies (Galhotra et al. Qual Saf Health Care. 2007;16:260-265; Hodgetts et al. Resuscitation. 2002;54:115-123) have attempted to define a process to identify and avoid potential cardiac arrests but failed to identify reliable criteria.
 
Therefore, Moskowitz et al (Resuscitation. 2019. Epub ahead of print) created a prospective observational study at Beth Israel Deaconess Medical Center that evaluated the preventability of ICU-CAs in a combination of cardiac, trauma, neurologic, medical, and surgical ICUs that totaled 77 beds. The ICUs were all staffed by critical care physicians except for the cardiac unit, which was staffed by cardiologists. Over a 10-month period from 2017 to 2018, 43 patients had ICU-CAs. All ICU-CAs were counted except 5, who were excluded because they arrested within 1 hour of arriving in the ICU.
 
The study used an expert panel and a survey group to analyze the cases. The expert panel was composed of 7 critical care physicians and 3 nurses. The survey group was composed of attending physicians, trainee physicians, advanced practice providers (APPs), and registered nurses. Each group was asked to rate each case on an ICU-CA scale. The ICU-CA scale started at 0 (not preventable) and went up to 5 (completely preventable), with a total score greater than 3 being “potentially preventable” and a score between 1 and 2 being “unlikely, but potentially preventable.” Members of both the expert panel and the survey group then selected from a predefined list of potentially contributing factors to the cardiac arrest or proposed other contributing factors that were not on the list.
 
The expert panel identified 14 (32.6%) patients and the survey group identified 13 (30.2%) patients whose ICU-CAs were “unlikely, but potentially preventable.” The expert panel identified 11 (25.6%) patients and the survey group identified 10 (23.3%) patients whose ICU-CAs were “potentially preventable.” Potentially preventable factors among both groups for ICU-CAs were delayed response to clinical deterioration, delayed intubation for respiratory failure, and the administration of anxiolytics and narcotics. The primary preventable factor was a delayed response to clinical deterioration. It was thought either timelier intervention or alternate treatment may have prevented an ICU-CA. The expert panel and survey group also found a delay in care when families of patients were making care decisions. The clinicians who gave families more time to think about patients’ goals before intubation or invasive treatment may have increased patients’ risks for an ICU-CA. The expert panel and survey groups also noted that symptoms perceived as treatable with narcotics and anxiolytics could have instead been signs of clinical deterioration, which increased patient risk of an ICU-CA. Bedside nurses, training physicians, and APPs were more commonly aware of a patient’s deterioration whereas attending physicians were less aware of the patient’s deterioration. This suggests that improved communication could also be a factor in preventing ICU-CAs.
 
This study had a number of limitations, a large portion of which were in its qualitative design. The ability to obtain tangible findings from a subjective analysis of a cardiac arrest is difficult. The groups’ opinions were not always highly confident, which led to weaker findings. Also, members of the groups were not directly involved in care of the patients, so there may have been a disconnect between participation in care and observation of care, which was probably exponentially exacerbated for the survey group. The study evaluated only events that led to an ICU-CA. Many delays in diagnosis occur every day in the ICU but do not result in an ICU-CA. Further study must be done to investigate what measures can be taken to prevent this disconnect in clinical care.
 
Despite these limitations, some of the study’s conclusions are valuable and likely to aid in prevention of ICU-CAs. Improvement in communication among team members, developing goals of care with family sooner, and looking for other signs of hemodynamic compromise in the setting of agitation may help reduce ICU-CAs.
 
Sean Amedeo, ACNP-BC, is a critical care nurse practitioner at INOVA Fairfax Medical System.
 
James H. Lantry III, MD, is an adjunct assistant professor of emergency and critical care medicine at the University of Maryland Medical Center in Baltimore, Maryland, USA. Dr. Lantry is an editor of Concise Critical Appraisal.