Email Updates
Application of Safety-II Principles
Venkatesan C, Helak K, Sousane Z, et al. Application of Safety-II Principles. PSNet [internet]. Rockville (MD): Agency for Healthcare Research and Quality, US Department of Health and Human Services. 2024.
Venkatesan C, Helak K, Sousane Z, et al. Application of Safety-II Principles. PSNet [internet]. Rockville (MD): Agency for Healthcare Research and Quality, US Department of Health and Human Services. 2024.
Traditional approaches to patient safety have often been reactive rather than proactive, seeking to identify and mitigate contributing factors after adverse events have occurred, and seeking to prevent future events by imposing constraints. This approach to patient safety is referred to as Safety-I, which defines safety as a condition where as few things go wrong as possible. Safety-I focuses on minimizing adverse events by using tools such as root cause analysis to evaluate contributing factors and risk assessments to understand the likelihood of these events occurring. A goal of the Safety-I framework is to implement processes and procedures to prevent the recurrence of these root causes.
Safety-I is a valuable framework for patient safety, particularly when evaluating adverse events. But it falls short because it emphasizes the few things that go wrong rather than seeking to understand the factors that contribute to the many things that go right every day in health care. Healthcare systems have become increasingly complex, and healthcare workers operate in dynamic environments that require flexibility to operate effectively as a team despite variable and often ambiguous circumstances. This increase in complexity in the healthcare system and recognition of the importance of evaluating successes in addition to failures yielded a shift to the Safety-II framework for patient safety, which seeks to understand why things go well in health care. Safety-II views safety from an everyday, system-wide perspective and recognizes the importance of adaptation in contrast to the Safety-I approach of primarily evaluating incidents of failure.
As Safety-II has gained traction, there has been an increase in its practical application in patient safety programs and care delivery across the healthcare landscape. This essay provides an overview of Safety-II principles and discusses examples of how these principles can be applied in healthcare.
Overview of Safety-II Principles
Key principles that guide the Safety-II framework include acknowledging and embracing variability and building resilience by learning from everyday successes. Rather than defining safety as the absence of adverse events, where as few things as possible go wrong, Safety-II defines safety as a condition where as many things as possible go right. Safety-II seeks to understand and replicate processes that allow things to go right most of the time despite variable conditions.
Under a Safety-I framework, the goal is to control, limit, or reduce variability. In contrast, Safety-II views variability as inevitable and critical for obtaining success and avoiding failures. In health care, things often go right due to healthcare professionals’ ability to adapt to the dynamic conditions of their working environment. While Safety-I primarily views humans as a liability or a hazard and aims to implement standardized policies and systems to avoid the recurrence of adverse events, Safety-II understands that specific healthcare circumstances are not always predictable and the decision-making of healthcare workers is an asset. A Safety-II approach assumes that with everyday performance variability, healthcare professionals provide the adaptations needed to respond to varying conditions.
Another key principle of Safety-II is its focus on resilience, or the ability of a system to adjust its functioning prior to, during, or following events to sustain operations under both expected and unexpected conditions. Organizational resilience in health care is characterized by the capacity to monitor, react, anticipate, and learn in order to maximize opportunities for things to go right. From a Safety-II perspective, this means that in order to improve safety, it is important to take a proactive approach to understanding what enables and/or impedes healthcare professionals to succeed in their everyday work.
Application of Safety-II in Health Care
There are several practical applications of Safety-II principles within healthcare systems. One example is to map a process using the Functional Resilience Analysis Method (FRAM) to understand work-as-done, which refers to everyday work as it is actually carried out, compared to work-as-imagined, which is the perception of how the work should be carried out. A FRAM analysis involves identifying functions, describing the performance variability of these functions, analyzing relationships between functions, and monitoring and controlling variability. This model to assess processes and adverse events has been used in settings such as intensive care, transitions of care, and sepsis management.1
Another example is incorporating a Safety-II mindset into existing Safety-I tools, such as root cause analysis. Traditional root cause analysis limits the learning that can be gained from assessing adverse events because it focuses primarily on what went wrong. Safety-II can be incorporated into root cause analysis and adverse event investigations by identifying misalignments between work-as-imagined and work-as-done, discussing what went well in addition to what went wrong, and discussing how to adapt and adjust the work processes that led to the adverse event. Broadening the scope of adverse event investigation by asking what normally happens and what went well can allow for a deeper understanding and strengthening of the work-as-done that creates safety on a day-to-day basis in addition to preventing failures that caused the adverse event. Simulations can also be incorporated into root cause analyses in situ in order to understand patient care situations and highlight differences between work-as-done and work-as-imagined.
Safety-II principles can also be incorporated into routine debriefing. In addition to weaving a discussion of what went well into adverse event investigations, healthcare organizations can incorporate debriefing as a tool for positive events, where patients have a good outcome, to understand what led to that outcome and replicate those processes. For example, New York Health and Hospitals pioneered what they call a success cause analysis, which uses root cause analysis methodology to understand the factors that contributed to favorable outcomes as well as adverse events.2 The University of Pennsylvania implemented an activity called One Safe Act, which asked perioperative staff to highlight one safe act or proactive safety behavior that they conducted in their daily practice. Participating staff reported that this activity changed their patient safety practices, improved their ability to deliver safe care, and demonstrated their colleagues’ commitment to patient safety.
Safety-II principles can also be applied in communication between healthcare professionals. A newly funded pediatric Patient Safety Learning Laboratory is exploring protocols for communication channels between clinicians in order to design, develop, and test a more resilient patient safety communication system.3
Finally, to understand work-done versus work-as-imagined, the authors (CV and KH) have started implementing a learning team methodology at Inova Health System.4 These sessions are three-hour, in-person meetings with between 15 and 20 subject matter experts from a variety of disciplines. Learning teams may or may not be triggered by a safety incident; the goal is to examine processes from the perspective of those involved to grasp the intricacies and risks of potential mistakes, errors, and harm, and to identify what ensures successful outcomes. The authors use the 4D Question Tool to encourage authentic team discussion of what went well and what could be improved.5
Future Directions
Safety-I and Safety-II are not mutually exclusive; they are complementary perspectives that can be used together to understand safety and to address and mitigate failures as well as to replicate and amplify successes. While we see Safety-II being integrated at the healthcare organization level, many policies and reporting requirements from authorities and regulators still focus on what goes wrong rather than what goes well. Moving forward, successful implementation of Safety-II concepts in health care will require collaboration among stakeholders, including healthcare providers, administrative leaders, and policymakers to consider how we can best build resilient healthcare systems.
References
- Sujan M, Pickup L, Vos D, et al. Operationalizing FRAM in healthcare: A critical reflection on practice. Safety Science. 2023;158:105994. doi:10.1016/j.ssci.2022.105994
- Parkash V, Musser L, Krouss M, et al. Success cause analysis: learning from what works to advance safety. Institute for Healthcare Improvement. January 9, 2024. Accessed July 22, 2024. https://www.ihi.org/insights/success-cause-analysis-learning-what-works…
- Drosey B. Patient safety learning laboratory looks to enhance clinical communication. Children’s Hospital of Philadelphia Research Institute. January 26, 2024. Accessed July 26, 2024. https://www.research.chop.edu/cornerstone-blog/patient-safety-learning-laboratory-looks-to-enhance-clinical-communication.
- Robinson B, Mccarthy G, Sutton B. The Practice of Learning Teams: Learning and Improving Safety, Quality and Operational Excellence. Santa Fe, Nm Pre-Accident Investigation Media; 2020.
- Bryant J, Lyth J, Robinson B, Sutton B. 4Ds for HOP and Learning Teams: A practical how-to guide to facilitate learning from everyday work, critical and dynamic risks with the 4Ds. Santa Fe, Nm Pre-Accident Investigation Media; 2023.
In Conversation with Chalapathy Venkatesan and Kathy Helak about Application of Safety-II Principles
Editor’s note: Dr. Chalapathy Venkatesan is the Chief Quality and Safety Officer, and Kathy Helak is the Assistant Vice President for Patient Safety at Inova Health System. We spoke to them about Safety-II principles and their application at Inova.
Sarah Mossburg: Welcome. Could you tell us a bit about yourselves and your current roles?
Kathy Helak: I am a critical care nurse by training and the Assistant Vice President for Patient Safety at Inova Health System. Dr. Venkatesan and I oversee the execution of safety science practices across our health system, working with our staff across all levels, from the most senior executives to frontline workers.
Chalapathy Venkatesan: I am the Chief Quality and Safety Officer at Inova Health System. By clinical background, I am an internal medicine hospitalist. At Inova, my role is to help lead the organization to do the greatest good for the greatest number of people. We believe we have a moral imperative to be open-minded to all the different possible approaches to thinking about safety and how to improve it.
Sarah Mossburg: For our readers who may not be familiar with the Safety-I and Safety-II frameworks, could you describe them at a high level?
Kathy Helak: Safety-I uses a traditional, reactive, and failure-focused approach to managing safety events when they happen. That includes looking at the events retrospectively with root cause analysis (RCA) or apparent cause analysis (ACA) to determine what went wrong, identify the cause and effect, and then implement a change that prevents similar occurrences in the future. In contrast, Safety-II focuses on what goes well and seeks to understand how we can learn from and replicate that. Safety-II considers the complexity of healthcare systems by acknowledging that the day-to-day work of our team members is not linear. There are many dependencies for things to go well, and there are varying conditions that occur throughout the course of the day-to-day work. Proactively understanding these variabilities and what is necessary for the right performance and outcomes brings a more comprehensive lens to understanding how we build capacity and agility for success in varying conditions. Safety-II emphasizes understanding the true work that is done, not the work that we imagined.
Chalapathy Venkatesan: Safety-I, the traditional mindset, defines safety as the absence of harm. The assumption here is that the system is not complex, is inherently safe, and that bad things only happen when things go wrong. Safety-II assumes that the system is intrinsically dangerous and that it is only through constant adaptation that we're able to keep it safe. From a methodological perspective, Safety-I really focuses on adverse events that have already occurred, and Safety-II focuses on learning from everyday work.
Sarah Mossburg: Are there ways that Safety-I and Safety-II overlap?
Kathy Helak: With both, the aim is to improve patient safety and reduce harm. They are complementary. Safety-I emphasizes compliance with policies, procedures, and ways to minimize risk. Safety-II stresses critical thinking and what it takes to be agile and to adapt to different circumstances to get the best outcomes. Policies and procedures are operationalized more effectively when there is a full understanding of workflows, so there is room for both a Safety-I and Safety-II approach.
Chalapathy Venkatesan: I agree. They are complementary, and organizations should not choose one versus the other. The approaches themselves, including the methodology, mindset, and definitions of safety, are very different. The overlap is really in what the organization is trying to achieve, which is ongoing safety and less harm. Both also require some universal capabilities. To carry out either framework, you have to have leadership commitment from your board and executive leadership all the way to the frontline for safety, and you have to have an environment of psychological safety and inclusion.
There are still instances where the Safety-I approach makes the most sense. The reason for the Safety-II hypothesis is that most of the time, things go right despite everyday variation. However, there are times when you feel like something is “off.” For example, if we had gone almost three years without a patient suffering a surgical site infection after coronary artery bypass grafting, and then in two months, we had three patients experience SSI, a Safety-I approach may be appropriate.
Sarah Mossburg: How are you using Safety-I and Safety-II at Inova Health System?
Kathy Helak: At Inova, we try to take the best of both approaches. For example, when a safety event occurs, it’s important to do a post-event review. We integrated the principles of Safety-II into that post-event review by adding a discussion of what went well and who should be recognized early in the review conversation. We want to optimize and understand those pieces of the story because we may need to capitalize on them. Having curiosity for what went well, in addition to understanding where we can learn and improve, changes the entire tone of the conversation. When we come in wanting to hear what went well, such as the strong teamwork that occurred or the recognition of an individual’s helpful actions, we promote psychological safety that helps people speak up. We create an environment where team members appreciate that their efforts are being noted and an understanding that what went well is equally important to what we can improve. Shifting the tone of the conversation in this way has created greater engagement of our team members and enhances the learning that can occur.
With our traditional Safety-I approach, we heard from team members that there was often not psychological safety to speak up because of the focus on what went wrong and “What are we going to improve?” without acknowledging that every story has complexity. We adopted the “harm event and learning” or “HEAL” acronym from Intermountain Healthcare because words matter. We infused Safety II principles of “What went well?” and “Who should be acknowledged?” into our post-event discussion and stress these points before talking about what can be improved. Because of the experience we are creating, team members now often ask for HEAL sessions to talk about safety events.
Sarah Mossburg: How did your team become first interested in integrating Safety-II principles?
Kathy Helak: I became interested in Safety-II principles when we recognized that we were stuck. Safety outcomes were not changing in the way we wanted. We needed to take a pause, do a critical assessment of our current approach to safety events, and see where we could improve.
Studying Safety-II really highlighted the importance of creating space to explore our team’s day-to-day work and adaptations to an extremely complex and ever-changing environment. I could see that using traditional methods was limiting our line of sight. We were missing the bigger picture of what goes well most of the time and why. We were not tapping into the organizational wisdom and expertise of our team members as effectively as we could.
Chalapathy Venkatesan: At Inova, a lot of things converged at the same time. A study about the safety of inpatient care revealed the frequency of harm had not really changed over many years, as well as the worsening of outcomes before, during, and after the pandemic. As a healthcare system, we have a lot of external objective measures of high performance in safety, such as CMS Star Ratings and Leapfrog safety grades. Still, underneath those objective measures, we saw inconsistent performance and recurrent safety events that were unacceptable. We could have gone back to basics and doubled down on the traditional way of thinking about safety. Still, we felt we needed to think differently about safety altogether, especially in light of the increasing complexity of health care.
At the time, many of the published articles on Safety-II focused on abstract theories and complex topics such as resilience engineering. However, in early 2023, we attended a conference called “Practical Applications of Safety II.” At that conference, we learned from other industries, such as aviation and nuclear power, sharing practical Safety-II applications. Those industries adopted Safety-II at a much higher rate than health care. That's when we started to see that Safety-II was not just a theory but something we can practically apply.
Sarah Mossburg: Where did you begin when you first started to incorporate Safety-II?
Kathy Helak: The patient safety team led the operationalization of Safety-II by looking critically at how we were managing safety events and any challenges that might be contributing to undesirable outcomes and experiences for teams and leaders. For example, feedback from team members indicated discomfort with speaking up in traditional post-event debriefs, and managers shared that they did not always feel comfortable and confident leading those exercises with their team members. We also did a “4D” assessment of our Root Cause Analysis and Action (RCA2) process to identify (1) What doesn’t make sense, (2) What is difficult to do, (3) What is different from what we think, and (4) What is dangerous in this current work.
This assessment and learning led to a refresh of our event management toolkit. We replaced Safety-I debriefs and ACAs with our new Safety-II HEAL sessions. While we continue to use the RCA2 methodology when necessary, we have shifted our attention to using the “learning team” methodology more often for insight into processes that drive risks of safety events and harm. Learning teams are multidisciplinary and include frontline experts and the 4D Question Tool for candid conversations on the real work done every day and dependencies for things to go right.
Sarah Mossburg: Can you tell us more about how you have used learning teams to operationalize Safety-II?
Kathy Helak: We use the learning team methodology to understand work done versus work imagined. This may be precipitated by a safety event, but the intent is to look at processes through the eyes of the stakeholders to understand complexities and risk concerns for mistakes, errors, and harm, as well as what makes it go right.
We have conducted learning teams on early ambulation, bedside medications, oxygen safety, and care of patients with behavioral health needs on medical-surgical units. These are three-hour, in-person sessions with 15 to 20 subject matter experts from various disciplines. Each team is formally chartered by an executive leadership team accountable for operationalizing the learning team recommendations. The session kicks off with a team welcome and setting the tone for psychological safety and candor. We then move into open team discussion using the 4D Question Tool, provide lunch and time to reflect on the discussion, and conclude with the team’s recommendations.
It's been incredible to watch the “aha” moments as frontline team members engage and learn from each other. Participants feel heard and appreciate the opportunity to share their experiences and collaborate with other disciplines to solve real issues that impact the safety of their work.
We finish every learning team with a photo of the team, which is included in the formal report to humanize the experience and highlight the Inova team members who took the time to come together, learn from each other, and collectively make recommendations to improve processes. This work is then formally reported to and acted on by the accountable leadership team. We also showcase the team’s work in a one-page “learning team highlight” for collective awareness across our health system.
Sarah Mossburg: When operationalizing Safety-II, were you using these tools in one hospital, one unit, or across the entire system?
Kathy Helak: We rolled out a full toolkit across our entire health system at the same time through formal presentations and leader trainings as part of our larger Safety-II transformation. We have continued to be intentional during this transition, regularly obtaining feedback to ensure we are doing safety differently and in a positive way. In addition to the Learning Team and HEAL Session methods, we have a Safety Pause. This is a proactive executive leader briefing or huddle with direct frontline team members regarding a specific safety concern, such as patient falls, or patient identification. Leaders open this with a request for help solving a safety concern—the why—and then actively listen to feedback and suggestions.
For us, it’s not just about event management; it’s about how we can incorporate Safety-II principles into daily conversations and the work that we do. Just incorporating Safety-II into event management is a limitation, and we recommend organizations think about how to take these concepts and integrate them into everything that they do in terms of care delivery. This includes the words we use to talk about safety, how we talk to patients differently around safety, and how we do our huddles and rounds.
Sarah Mossburg: You mentioned shifting the words you use to talk about safety when integrating Safety-II. What are some examples of words or vocabulary you have changed when referring to patient safety?
Kathy Helak: One example is that we don't refer to event reviews as investigations. We call these learning sessions to tell a story of the event to make it more human-centered for open conversations. A key premise of our mission, vision, and values is that every person comes to work every day to do the right thing, and so words do matter. Labeling a review as an investigation may imply someone did something wrong. We know that even when an event is not the best story, there are things that went right, and people were trying to do the right thing.
Chalapathy Venkatesan: We keep a list of traditional versus team-centered language. For example, instead of incident, we’ll say event. Most of the time, when we refer to events, we are referring to situations when there was an undesired outcome. However, as we shift our culture, we have also shifted to referring to great catches and patient care experiences that went well as events. Anything that is reported in our safety reporting system is noted with this label.
Instead of a causal factor, we'll say condition/contributing factor. Instead of non-compliance, we’ll say variation, adaptation, or improvisation. We realized that no matter how well and how compassionately you facilitate an event review or an RCA, sometimes the walls are still up from team members. Anything we can do to enhance learning and bring the walls down, such as a shift in language, is what we're going to do.
Kathy Helak: I think all organizations embrace a learning culture, but we have been deliberate in operationalizing this. We use the word “learning” frequently to emphasize our intent. From learning will come improving, but we don't focus on the improvement; we focus on the learning.
Chalapathy Venkatesan: When I first started to learn about Safety-I and Safety-II, I thought we do Safety-II only because it feels good, but it’s really about optimizing learning and squeezing out as much learning as possible. Most of the time, things go well. It makes sense to spend more time on that, not because it feels better to review things that went well, but because that's what happens most of the time. If things didn't go well most of the time, then that's what we would focus on, but that's not the case. For example, most people don't fall or get a wrong site surgery. We should not ignore those instances when they happen, but we should also really be mindful that the learning from those situations is going to be limited.
Sarah Mossburg: What are some challenges you've encountered as you've shifted to incorporating a Safety-II mindset?
Chalapathy Venkatesan: One challenge early on was that when you're asking frontline teams to describe what actually happened, the tendency of leaders can be to potentially modify the reality of what happened. When we involve frontline staff, we must be committed to taking their word as the facts of the work as done. Leaders sometimes tend to filter and say, “But that's not how we do it” or “I thought we said this is what we were going to do.” When you involve the front line, you have to commit to not filtering or curating what they authentically report.
Kathy Helak: Our new Safety-II approach is creating robust, innovative improvement ideas, and we believe those are the right things to do. However, we then have to execute these improvement ideas with a sense of urgency to maintain credibility of the process. Sometimes, is it hard for leaders who have put a process or system in place to hear “it doesn't work that way”. As we gain more collective experience, leaders are getting more comfortable and eager to use this method.
Chalapathy Venkatesan: I agree. There's also some shock value that has to be managed. People expect that if something happens, we do an ACA, regardless of whether it is effective in learning and improving. Foregoing that step can be an adjustment for some people. We noticed that with hospital-acquired infection, it took some time for us to get to the point where we would not do an ACA for every single central-line acquired bloodstream infection (CLABSI) patient but would instead try and figure out what in the system of care is problematic. For example, is it the insertion? If so, we'll do a learning team about catheter insertion. For the most part, people have gotten used to foregoing traditional pathways following incidents because there are alternatives. I also had a mistaken belief that regulatory bodies would insist that you've done an RCA. We've had a few instances of serious safety events where we did not do an RCA, but instead maybe did some immediate mitigation and a learning team that resulted in a robust action plan.
Sarah Mossburg: Are there any lessons learned or advice that you would provide to other organizations that are interested in integrating Safety-II into their safety programs?
Chalapathy Venkatesan: It is important for organizations to first assess where they are when it comes to patient safety and care delivery. I think there are a lot of ways to do that, with the National Action Plan being one of them. Within that, it’s important to assess honestly where organizations feel they are from an intra-organizational learning perspective, which is on the National Action Plan. I imagine many feel like they have not arrived when it comes to safety, so I think coming to that conclusion first is important. It's very hard to have a mindset that's counter to contemporary safety. The second step is to tap into external expertise, sometimes outside of health care. And third, it’s important to understand that what has gotten us to this point is not possibly going to get us to where we want to be when it comes to safety. There comes a time to try a new approach. With the learning teams, there were a lot of different entry points. Sometimes it is a serious safety event, a recurrent event that we've heard about, or sometimes it's problematic or high-risk processes, like our response to work, workforce safety events, our inter-hospital transport, or early ambulation
What helped us was discovering accessible ways of practicing Safety-II and Safety-II tools. Once you hear about the 4D tool and learning teams, it does require expertise from our patient safety team to facilitate, but it's accessible and doable for us. You don't need engineers or complex processes to do it. There are other sorts of Safety-II tools besides learning teams. In your existing causal analysis, you can ask what normally happens, not what happened in this instance, which can help you learn about everyday work. A hospital in New York presented the concept of performing a success cause analysis after care is provided and a patient experiences a great outcome. They use the same root cause analysis approach, but the entryway is a successful outcome. Some organizations just ask, What is one thing you do to keep your patient safe? If you ask registration, perioperative techs, or bedside nurses, you'll hear about things that folks do outside of policy and outside of work as imagined to keep patients safe.
Kathy Helak: It is also important to acknowledge that this is not a unit tactic or a pilot to be implemented at a micro level. This is safety culture transformation at the organizational level and an important overall strategy for doing safety differently to achieve better, sustainable outcomes.
Getting started can be as simple as opening safety event reviews following undesirable outcomes with a discussion of what went well and if there is someone who should be recognized for going above and beyond. Highlighting the positive early in the discussion changes the tone of the conversation for more open discussion and learning for true improvement opportunities. Additionally, acknowledging the complexity of the environment illustrates awareness of the daily work experience. As we mentioned before, changing the language is another straightforward way to start to weave in different mindsets about safety. And last, I encourage leaders to be open to their learning and tap into other organizations doing Safety-II.
Sarah Mossburg: Those are great examples of thoughtful, interesting things organizations can do. Thank you both for your time today.
