In anaesthesiology – as in other medical specialities – we learn from what goes wrong. We register complaints, incidents and adverse events and try to learn from them and improve our management. This is most frequently called a Safety-I approach: clear guidelines and protocols have been developed to agree on unambiguous working methods and thus increase patient safety. And we have to admit that patient safety has greatly improved thanks to Safety-I!
Yet, we also start to notice that this approach does no longer help us to further improve our practice, and we thus have to admit that we keep repeating the same themes concerning perioperative patient safety: suboptimal communication, suboptimal handover, medication safety, medication errors, etc.
It should be realised that improving safety is relatively easy if the starting situation is poor. However, the last five to 10 percent takes considerable effort with more and more difficulties. The impossible requirement for absolute safety may cause so much stress that, on the one hand, it contributes to low morale (too high goals, or too many goals, finally leading to unintended behaviour), and on the other hand -paradoxically enough- can result in a reduction in safety because people finally insufficiently learn and report. People are imperfect and will make unintended errors. In addition, some risks in our daily work are yet unknown, especially those related to any innovation in the medical field. Therefore, these risks cannot always be foreseen, nor can they always be prevented. Do, consequently, we have to change our safety culture?
Culture is about the collectively shared norms, values, assumptions and beliefs in a group. Proactive attention to safety, alertness, room for reflection, giving each other feedback, being critical of one’s actions, and finding a balance between safety and fairness are characteristics of a safety culture. However, nothing is more difficult than changing an inadequate culture.
Within the practice and research of patient safety, in recent years, more and more attention has been paid to the shift from Safety-I to what is called Safety-II culture. While Safety-I is mainly about learning from things that ‘do not go well’ and this approach is based on concepts such as ‘risk’, ‘damage’, ‘incidents’ and ‘causes’, Safety-II sees healthcare more as a complex adaptive system, in which many interactions and adjustments continuously take place. While learning from incidents was already part of Safety-I, this seems to be even more central in Safety-II. However, it is not about learning from negative but positive events. Most things in our daily work usually go well, and why is that?
In a complex system like healthcare, considerations must be made continuously, and care providers must constantly deal with unexpected situations and conflicting goals that cannot be laid down in protocols. With Safety-II you look at the entire process and investigate what really adds value to improve quality and safety. This does not have to be big and complicated. Safety-II is an addition (not a replacement) to Safety-I. It focuses on what is going well and how the adaptability and resilience of professionals can be increased to support safe working. The underlying pedagogical principle seems to be that learning from something good feels better than punishing something bad.
Many people describe Safety-II as ‘learning from what goes well’. They immediately think of learning from successful situations. This is undoubtedly part of it, but probably not the most important. We could see Safety-II mainly as learning from what happens in daily practice on the work floor. Sometimes things go wrong, usually, most things go right, and sometimes things go much better than expected.
In day-to-day practice, there are good practices that you can learn from and may be able to use more broadly, but which are yet not frequently described in detail. It is precisely the insight into the full breadth of everyday practice that helps to find targeted starting points for improvement. Safety-II sees professionals as a source of resilience: the professional can adapt to daily changing circumstances, making processes complex and dynamic instead of linear, as seen in Safety-I. What makes Safety-II thinking powerful is that improvement opportunities are created together with involved professionals, and these improvement opportunities are not imposed by the management. As a result, these opportunities are most likely much more in line with what is feasible on the work floor, thereby significantly increasing the chance of successful implementation.
There is more and more discussion about the ‘how’ of Safety-II. As we are used to from Safety-I thinking, there is a need for concrete instruments and outcomes to measure the effects of Safety-II. However, things are different with the Safety-II thinking. There are concrete tools that will certainly help us, but Safety-II must, above all, become the new way of thinking about quality and safety, and it should not be a ‘trick’ that you can implement in your organisation using a toolkit or a package of requirements.
Instead of only discussing and registering things that went (nearly) wrong (classical Incident Reporting Systems), we should all start to register what is going well, what makes the difference, and what are the positive incidents. Some questions to start with could be:
- Tell me a story about an experience at work where you felt you performed well or made a difference. What happened? What were you particularly proud of? What was the outcome?
- Tell me about 3 of your strengths. How do they positively support your work?
- What do you value about your work? What is the most satisfying thing about your job?
- Can you give me an example of when you improved safety – directly for a patient or within your organisation?
- Tell me 3 small wishes you have for your team and/or your work.
The Learning from Excellence initiative (1) gives additional tools that may help to start the new Safety-II thinking. Analysis tools that would work in the field of Safety-II are, e.g., the Resilience Assessment Grid (RAG), which is helpful for understanding how frontline healthcare professionals manage the complexity of everyday work (2). To start, it would be necessary to measure the resilient potentials of Respond, Monitor, Learn and Anticipate of your own department.
One practical tool is “Appreciative Inquiry,” which has been successfully implemented in a couple of paediatric intensive care units (3). Functional Resonance Analysis Method (FRAM) may be used to prospectively analyse a complex process and for retrospective analysis of incidents (4). Based on interviews with all persons/staff groups involved, a visual map can be made of how a process works in practice. The FRAM fits well with the Safety II concept because the focus is on learning from the variation in daily practice. It is not so much about establishing that people are not working according to the guideline but about identifying differences between the guideline (work-as-imagined) and daily practice (work-as-done). Based on this, it is possible to determine very specifically which variation is desirable and which is undesirable. Improvements can be made as a result of this.
There are some other (simpler) opportunities to develop into a Safety-II culture, e.g., by simply using positive language (5). For example, according to local protocols, pain is regularly measured in post-surgical patients, with quality measures focusing frequently on whether or not the pain measurement is registered. The purpose of this pain measurement is to provide sufficient attention for the patient, as well as to use the right interventions so that the patient experiences that enough is being done to keep their pain acceptable. The patient is always involved in the implementation of the postoperative pain policy. Based on the Safety-II concept, we could handle the question differently by asking more about comfort instead of talking about pain experience or pain figures. This gives a different pain experience and will ensure a more appropriate pain treatment. Specifically, suppose a patient gives an eight as a pain rating, while the nurse herself would not have estimated the patient’s pain as high. If the nurse then asks this patient other questions, e.g., evaluating if he is comfortable lying down or moving comfortably, then this patient could probably be in less pain than the number given would imply. With a score of 8, the nurse would be fully committed to pain relief, while less medication may be needed if the patient indicates that he is still comfortable.
Involving the patient and their family is also worth taking into account. In James Reason’s Swiss cheese model, the patient can form an important safety barrier as a ‘safety buffer’. After all, maximum effort is guaranteed if you can contribute to preventing damage to yourself or those you love. The patient is also unique in his field of observation: in contrast to the fragmented view of care providers, the patient is present in almost every form of care provided to them. Patients are the continuity factor in each care process. Previous surveys indicated that around 85 percent of patients are willing and able to take on this role. A validated measurement list is available that patients can use to provide structured feedback about their safety (6).
To conclude, a couple of instruments are available to start now with the work in your own department to bring the concept of Safety-II into your clinical work, thereby leading safety culture to an even higher level than we reached in the former years.
- The application of resilience assessment grid in healthcare: A scoping review. Safi M, Thude BR, Brandt F, Clay-Williams R. PLoS One. 2022 Nov 4;17(11): e0277289. doi: 10.1371/journal.pone.0277289.
- Positive approaches to safety: Learning from what we do well. Plunkett A, Plunkett E. Paediatr Anaesth. 2022 Nov;32(11):1223-1229. doi: 10.1111/pan.14509
- Preoperative Anticoagulation Management in Everyday Clinical Practice: An International Comparative Analysis of Work-as-Done Using the Functional Resonance Analysis Method. Damen NL, de Vos MS, Moesker MJ, Braithwaite J, de Lind van Wijngaarden RAF, Kaplan J, Hamming JF, Clay-Williams R. J Patient Saf. 2021 Apr 1;17(3):157-165. doi: 10.1097/PTS.0000000000000515.
- Hypnosis and communication reduce pain in peripheral intravenous cannulation: Effect of language and confusion on pain during peripheral intravenous catheterisation
Fusco N, Roelants F, Watremez C, et al. Br J Anaesth. 2019;124(3):292-8.
- Developing a reliable and valid patient measure of safety in hospitals (PMOS): a validation study. McEachan RR, Lawton RJ, O’Hara JK, Armitage G, Giles S, Parveen S, Watt IS, Wright J; Yorkshire Quality and Safety Research Group. BMJ Qual Saf. 2014 Jul;23(7):565-73. doi: 10.1136/bmjqs-2013-002312