Originally Published in the August 2003 issue of Critical Connections, Volume 2, Number 4

Christine Holzmueller, BLA
The Johns Hopkins University
Baltimore, Maryland, USA

Lisa Lubomski, PhD
The Johns Hopkins School of Medicine
Baltimore Maryland, USA


The ICUSRS team walked into the lives of health care providers in intensive cares units one year ago.  Their goal was to present an alternative method for reporting adverse events and near misses than the traditional mandatory incident report. The hypothesis was that people would feel safe and comfortable reporting to a system that was anonymous, confidential and focused on addressing system failures not blaming individuals.

Last July this web-based Intensive Care Unit Safety Reporting System (ICUSRS) was piloted in three adult ICUs at Johns Hopkins and collected four reports.   Since that time, numbers have exploded to 18 adult and pediatric ICUs and over 900 report submissions. This is great news for a team of outcomes researchers interested in transforming useful data about system failures into safety initiatives for the critical care community.

So what has this innovative project shown thus far?   Preliminary findings reveal that multiple factors effect an incident, poor communication and use of policies and procedures are significant problems, nurses believe workloads are unrealistic and medication errors often start outside the intensive care unit.

Looking more closely, we found that most of the incidents involved 2 or more system factors, supporting previous evidence by James Reason and others regarding the negative impact of complex work structures on safety.  Our finding that many medication events started outside the ICU, often in the Pharmacy, further illustrates this complexity, making it necessary to peel away the layers involved in caring for a patient to find the causes of an incident.

Investigating why policies, procedures and protocols were underused unearthened a variety of reasons.    Some providers did not know a policy existed or noted that no standardized guidelines were written for a procedure.   One pertinent reason for not following an established protocol or guideline was the logistics of finding the dust topped manual in the face of a patient crashing after throwing a pulmonary embolus.    All of these motives can be easily modified with front end training, continuing education, and clearly written and accessible protocols and guidelines.

Finally, a hot topic in the news these days and a big problem in 65% of incidents reported to the ICUSRS was poor communication.   While failure to converse and miscommunication were undoubtedly several causes of incidents, we also found that staff's perceptions that their opinions were not valued and ignored suggestions were also significant causal factors. 

As our data collection for this demonstration project is ongoing and funding provided through the end of August 2004, we plan to continue our analyses and disseminate findings through publications and partnership with the SCCM.