Focusing on What Goes Right: The Safety-II Mindset

Amy C. Edmondson, known for her work in psychological safety and organizational behavior, squarely places healthcare organizations in the category of “complex systems.” In her article “Strategies for Learning from Failure,” she suggests that “failures are due to the inherent uncertainty of work.”

The very nature of the work requires healthcare professionals to be skilled in responding to change and adapting to uncertainty. I imagine that anyone who has spent time at the bedside will appreciate the obvious nature of this statement. Yet, some of the systems on which healthcare rests are built on methods that do not account for the uncertainty of the work and the adaptive nature of those performing the work.

Recently, one of my colleagues from Quality and Patient Safety introduced me to the concept of Safety-II. What does this mean?

Here’s the short version of what I have learned from hanging around people in the quality and patient-safety world: Early safety concepts emerged from studying behaviors, some of which led to the development of the manufacturing line. Most of you will be familiar with more modern concepts such as Lean Six Sigma, which aims to standardize work, streamline processes, and eliminate waste. In this kind of system, when things go wrong, the safety response seeks to understand where the system failed via tools such as root cause analysis. This is defined in this white paper as Safety-I. According to the authors, Safety-I culture is bimodal (success/failure) and reductionistic.

Safety-II, on the other hand, offers an alternate, additive paradigm that more aptly accounts for the complexity of healthcare environments and working with human suffering. Safety-II aims to additionally account for “work as done” instead of “work as imagined.” Simply put, Safety-I provides tools to help systems understand what went wrong, whereas Safety-II thinking focuses on what goes right. In doing so, Safety-II calls attention to the “adaptive variabilities in performance” by those working within these environments (see Learning from Excellence for more examples).

When viewed through this lens, human behavior is not viewed as something to “constrain…by reinforcing compliance and by eliminating variability.” To illustrate the need for both models of thinking, I return to the mistake made by the PGY3 resident I referred to in a previous blog post.

The story went something like this: The resident was called to consult on a very sick patient overnight, and based on his consultation, the fellow managing the patient ordered imaging. Because of the way the medical record works, in this situation the findings are sent directly to the fellow managing the patient, not to the resident providing the consult. However, as the resident used to tell me, “Time is brain!” Therefore, in a case like this, consultants on this service frequently go into patients’ charts unprompted to check for results.

But on this call shift, the resident was the only one on and was getting bombarded. I don’t recall whether there was no back-up call system to bring someone in when necessary or whether it was just unheard of and would be considered a sign of weakness. Either way, the resident felt that he needed to fend for himself.

The first mistake he made was forgetting to go back into the patient’s chart to check on the imaging, presumably because he was preoccupied by the steady stream of incoming pages. Time passed. When he finally remembered, he quickly checked, did not see anything remarkable, and returned to his other patients. More time passed.

At some point the Fellow, who I imagine was also preoccupied overnight, received the results, saw that intervention was needed, and paged the resident consultant.

When viewed through the Safety-I lens, the delay in treatment cannot be captured because the system functioned as intended. And when adaptive responses are made — for example, checking the patient’s chart unprompted to improve time to treatment — these will also not be captured in a Safety-I system.

Without the addition of the Safety-II lens, it leaves one to wonder: How much exceptional work remains unseen? As long as significant gaps exist in systems thinking between “work as imagined” and “work as done,” it would seem that opportunities for learning, recognition, and restoring hope will continue to be missed.

Follow Medscape on Facebook, TwitterInstagram, and YouTube

About M. Chase Levesque

For more than a decade, M. Chase Levesque, PsyD, has provided care to healthcare providers within a large rural academic medical center. She is a clinical psychologist and clinical assistant professor of psychiatry at the Geisel School of Medicine where she also serves on the Committee for a Respectful Learning Environment. Her clinical training was centered on providing integrated treatment healthcare within family medicine. As a result of her training and work experience caring for caregivers, she is a strong believer in the biopsychosocial (with an emphasis on occupational) model of care. She has a private practice in White River Junction, Vermont. You can reach her on LinkedIn.


Leave a Comment