Broadening our Approach to Safety: A Safety II Analysis of an Endovascular Thrombectomy Case


Emily Miner, MD, MPH

And Patient Safety Subcommittee of the Education Committee: Surya Kumar Dube, MD, DM; Amie Hoefnagel, MD; Kiran Jangra, MD, DM; Ian McCollough, MD; Sonal Sharma, MBBS, MD; Brie White-Dzuro, MD


In terms of patient safety, our field has come quite far. In the 1940s, 1 in 1,000 patients undergoing anesthesia died, versus 1 in 100,000 in the 2000s.1 This improved safety record is not only thanks to better medications but also because of a rigorous investigation of the “messy details” of safety incidents. These investigations led to the implementation of standard monitors and improved interfacing of humans and machines.2 While this Safety I approach has been vitally important to get us to where we are, there are diminishing returns with rigorous investigations of what are now rare events. As industries begin to adopt a Safety-II approach, the field of anesthesia can again lead the way, not just learning from events when things go wrong but also delving into lessons from everyday work when things go right.

In this edition of the SNACC newsletter, by focusing on a specific case, we will reflect on the Safety I versus Safety II lessons that could be learned from this case. As you read the case below, consider how to work, as done in this case, differed from how the work was “imagined” in the policies and procedures that typically dictate behavior in such a case at your own institution. Deviations from “work as imagined” occur frequently; how does this affect patient safety? Most respondents to the SNACC survey mentioned something about communication or information sharing being a barrier in these cases so this case will focus on communication breakdown.

 The Case: An Excellent Outcome from a Different Path

You have just reunited a grandchild with their beloved grandmother who, 2 hours prior, could not utter comprehensible speech or coordinate her hand into a tender grasp. You take this moment to reflect on the many factors that coalesced for this seemingly miraculous outcome – your patient is already back to a good functional status after what could have been a debilitating stroke.

The stroke occurred during working hours, just as your cases in interventional radiology (IR) were finishing. You began to hear people talking about a potential incoming stroke a full 15 minutes before the official communication over the paging system was sent, which your team did not even receive. You alerted your team and were assigned to the case. You found the patient in the scanner and reviewed the images in real-time. Since the proceduralist’s phone number was saved on your cell phone you discussed the patient, and the proceduralist helped coordinate with the Neuro IR (NIR) technologists. You received a handoff from the ED staff at the scanner and transported the patient directly to the NIR suite. Already aware of the proceduralist’s preference for an endotracheal tube from your phone call, you efficiently induced after a brief time out and allowed the proceduralist to begin while you worked on additional access.

Safety I Approach

A more traditional approach to patient safety may choose to focus on the failure of communication from the paging system within this case. A root cause analysis would look at why the anesthesia team failed to receive the stroke notification. The investigation may involve retrieving old paging logs, investigation of the pager in question, and discussions with involved team members. The solution may involve carrying an additional paging device in the event of a similar situation. Such an approach would miss out on a huge learning opportunity about how communication is handled more broadly for such cases because the “root cause” was found within the paging network.

Safety II Approach

The Safety II approach would see this case as an excellent one to learn from, not because of the thing that went wrong (page not received), but because of all the procedural violations carried out brilliantly by the teams to deliver the best care for this patient. It would ask – how did it happen that a patient was not harmed even though the page was not received? How can our organizations build resilience to allow more of this behavior to happen? These lessons will not be learned through investigations carried out in a conference room. Instead, we must understand the messy details of how everyday work keeps patients safe. This approach recognizes that safety is not some measured quality entity but instead a formed-in-the-moment interaction of humans with risks and hazards within a space.2,3

Safety II Learning Teams

To answer the question of how the patient was not harmed despite no page being received requires learning from the safety net of people that deviated from typical behavior. Convening a learning team can be helpful to ensure that those who do the everyday work are the ones empowered to define the problem and come up with solutions.4 The best teams have different mixes of experience and duration at your institution, welcoming different perspectives. For this case, you may choose to involve: a nurse, resident and attending physician from the ED, members of the stroke team, a CT scanner technician, a radiologist, resident, attendings, nurses, and technicians from Neuro-interventional Radiology (NIR), and anesthesia providers who meet with a facilitator trained in a Safety-II approach. The first meeting may involve a case presentation and then having the group define the problem but not yet venture any solutions. The following meeting can focus on solutions. From there, a micro-experiment may be appropriate to test the solution and see how things change.

The problem, in this case, may be defined as inter-team communication. There were factors that helped promote inter-team communication that day that were not always present. Each team may have a unique perspective- the anesthesia team had the flexibility that day to have someone nearby and able to transition from IR cases to focus on this thrombectomy case and go to the scanner. The ED staff may emphasize the importance of a pre-arrival notification from an incoming ambulance allowing direct transportation to the scanner, though this is not typical practice. The NIR team may remark on how much easier it was to discuss the case on the phone than through the paging system.

The solutions will focus on how to enable resilient deviations. Resilience in Safety II relates to the engineering concept that a resilient system is one where adjustments can be made within a system of changing conditions to sustain function when both expected and unexpected conditions arise.5

Perhaps a secure messaging platform or a launchpad for strokes or any number of different approaches may be appropriate. These solutions should be the scaffolding to support humans who do the work every day to make the system more resilient and adaptable. Cases like this that have successes may have more important information than cases that do not go well. Focusing on what went right may be the way to ensure more things go right. Safety could look like building our systems to enable more such solutions to blossom from the people who show up to do the work every day.


The field of anesthesia has led to patient safety. We have become orders of magnitude safer than we were. With this safety comes the need for a broader view. In a Safety II framework, we can learn from everyday work! Policies and procedures are violated frequently, oftentimes with benefit to patients- we must stop seeing our workforce as rule breakers who need more rules. Instead, we need to understand the problems in their everyday work and how they are being held back, and most importantly, how they muddle through despite it all.6  When we see those violations and the purposeful muddling through as the building blocks of resilient systems, we learn to create freedom within a framework for people to create safety.


  1. Li G, Warner M, Lang BH, Huang L, Sun LS. Epidemiology of anesthesia-related mortality in the United States, 1999-2005. Anesthesiology. Apr 2009;110(4):759-65. doi:10.1097/aln.0b013e31819b5bdc
  2. Wears R, Sutcliffe K. Still Not Safe. Oxford University Press; 2019.
  3. Cook R. Lessons from the War on Cancer: The Need for Basic Research on Safety. Journal of Patient Safety. 2005;1(1):7-8.
  4. Dekker S, Conklin T. Do Safety Differently. Pre-Accident Investigation Media; 2022.
  5. Merandi J, Vannatta K, Davis JT, McClead RE, Jr., Brilli R, Bartman T. Safety II Behavior in a Pediatric Intensive Care Unit. Pediatrics. Jun 2018;141(6)doi:10.1542/peds.2018-0018
  6. Braithwaite J, Hollnagel E, Hunte G. Resilient Healthcare Volue 6 Muddling Through with Purpose. vol 6. Resilient Healthcare. CRC Press; 2021:178.

Back to top