Safety-II and the study of healthcare safety routines: Two paths forward for research

Perspectives on safety: understanding the system

Looking at compliance with the surgical safety checklist from a human factors perspective, the assessment of compliance (i.e. looking for non-compliance) is associated with what Hollnagel et al. ¹¹ call a Safety-I approach. Safety-I focuses on when things go wrong based on the assumption that things that go right and things that go wrong (i.e. adverse outcomes) “are due to different modes of functioning” ^{11 (p. 10)}. The obvious way to avoid errors from a Safety-I perspective would be to “‘find and fix’: look for failures and malfunctions, try to find their causes, and then eliminate those causes or introduce barriers, or both”. ^{11 (p. 10)} In relation to the checklist, failure is something that happens during surgery, and the surgical safety checklist is a symbolic barrier, that is, a reminder to take certain measures before surgery starts in order to avoid failure. ²² Similarly, the assessment of mortality and complications and their correspondence with checklist usage are also aspects of a Safety-I approach. A complicating factor is that a failed checklist is, by all means, not the same thing as an adverse event happening to the patient. In fact, in most cases, i.e. when no error or mishap related to the checklists contents has occurred before the checklist, non-compliance with the surgical safety checklist or even the total absence of it, does not do the patient any harm at all. But in some cases, when an error or mishap has occurred, it does.

In contrast to Safety-I is the Safety-II approach. It assumes that things that go right and adverse outcomes “have a common basis, namely everyday performance adjustments”. ^{11 (p. 20)} According to Safety-II, it is much more fruitful to look at all occasions that go right, instead of the few that go wrong. In this sense, Safety-I is focused on failure, and Safety-II on resilience, i.e. a system's ability to cope with the unexpected. ²³ While safety-I essentially is founded on a linear conceptualization of safety, safety-II rather sees safety, and the absence of safety, as emergent and dynamic phenomena in a complex adaptive system.^11,24 An underlying assumption in Safety-II is that most work situations are underspecified, that is, partly unpredictable. ¹² While an important part of Safety-I is to decrease and control variation in performance, Safety-II recognizes that variation is inevitable. ²⁵ To improve safety from a Safety-II perspective, one must strengthen the system's ability to successfully adapt to variation in normal behavior, rather than redesign the whole system in order to handle specific adverse events. ²³ Since the system is not stable, the ability to, anticipate what could happen and to monitor what is actually happening, including to know what to look for, becomes crucial. ²³ An importance aspect of this is to gain an updated knowledge about the boundaries of acceptable performance and of gaps in the system.^26–28 The successful handling of incidents and near misses are, despite that no harm is done, valuable to study since they reveal information about the boundary conditions of the system as well as how the system behaves close to the boundaries. ²⁷ Information that could both facilitate anticipation of what could happen in the future and reveal patterns that are important to look for when monitoring performance. As Safety-II is concerned with the dynamics of the organization, it puts more emphasis on the actual work that is done and its underlying assumptions, compared to the formal organization of work.^12,25

Perspectives on safety: understanding the users

To gain insight into why the surgical safety checklist is used the way it is, one needs to know more about the behavior shaping mechanisms, including boundary conditions and gaps that affect performance; and in order to improve behavior, one needs to change these mechanisms.^22,26 Research shows that the content and number of team members involved in the checklist is adapted to meet the demands of the current situation. ²⁹ However, in the end it is always the individuals that decide how they act and respond to the checklist. For that reason, it is not only the behavior shaping mechanisms on the system level that is of interest but also those that influence the behavior of the actual users. While the implementation processes and compliance with the checklist has gained much attention,^{14,15,18–20} the team members’ mental models of how the checklist improves safety, including their perception of its usefulness, is often overlooked, as I see it. ³⁰ Mental models of how the checklist works constitutes input to the rationality that the surgical team members apply when using the checklist. This is important because people, at least most of the time, act in ways that are locally rational. ³¹ Given their understanding of the current situation, their presumptions, prior knowledge and the cognitive resources at hand, they act as rational as can be expected. As pointed out by Wagenaar and Groeneweg, accidents “… do not occur because people gamble and lose, they occur because people do not believe that the accident about to occur is at all possible”. ^{32 (p. 596)} Thus, in order to alter the users’ behavior in relation to the checklist and similar safety routines and tools, the foundations upon which the users’ base their rationality in the actual use situation, must be understood.

Understanding the users: focus on what the users value

The first approach is that instead of being preoccupied with what people do not do with the checklist, investigate what they value with the checklist. The perceived value or importance of different questions on the checklist can shed light on the team members’ mental models of the routine, and thus also on the local rationality that guides their choice of actions when using the checklist.^30,31 This approach is in line with safety-II in the sense that it focus on the foundations of normal behavior rather than on behavior associated with failure.^11,12 Knowing how the actual users perceive the surgical safety checklist, it will be much easier to design interventions that alter their behavior. The interventions can then be attuned to their own understanding of how their practice actually works.

What the users perceive as valuable or important with a safety routine in part reflects the users understanding of how risks emerge in relation to that specific safety routine. Thus, in order to increase the users’ adaptive capability, making these underlying assumptions or mental models explicit makes it possible for them to reflect on them and challenge them, which in turn can expand their understanding of how risky situations emerge. Hence, investigations into what users’ value with a safety routine, could both inform redesign of the routine, and serve as point of departure for proactive analysis of how adverse events can occur.

Understanding the system: investigate when the routine works

The second approach is that instead of looking for failure, investigate and highlight when the checklist works. It works when it captures something that potentially could lead to adverse events for the patient. Surgery is a complex activity. ²⁴ Before an action trajectory results in a potential adverse event, and before it hits the checklist, there can be several occasions in which the circumstances leading up to the adverse event could have been discovered. This shows that there could be possible barriers, and organizational gaps, other than the checklist that would also have to fail before the adverse event occurred.^28,33 Thus, a potential adverse event discovered by a safety routine, is also an opportunity to learn about the boundaries and dynamics associated with the complex adaptive system that the routine is part of.^24,26,27 Furthermore, when the checklist captures something that could have resulted in an adverse event, we have a concrete examples of the checklist's usefulness. If these types of examples, i.e. incidents, are highlighted for the personnel, they become more aware of the routines usefulness. Their behavior can then be altered in a positive direction through what psychologists call the availability heuristic. ³⁴ These types of events can also be used to illustrate and investigate the variation in, and dynamics of, situations that can lead to adverse events, i.e. to learning about boundary conditions and about gaps in the system as well as about what could be expected to happen and what to look for in the future. Knowledge, which can increase the capability to assess future events and to adapt to them.^11,23 For the individual user of the checklist this implies an opportunity to develop more complete mental models of how adverse events can occur. ³⁰ Remember, Wagenaar and Groeneweg states that accidents “… occur because people do not believe that the accident about to occur is at all possible”. ^{32 (p 596)} Potential adverse events, captured by safety routines, can also serve as entry points for analysis with more systematic methods to model performance and risks in complex systems, such as the Functional Resonance Analysis Model or Cognitive Work Analysis.^35,36 Another vialbe approach could be to use information about potential adverse events and the users mental models in relation to safety measures, to conduct work system analysis influenced by the SEIPs framework. Particulary the focus on journeys in SEIPs 3.0 could be of interest.^37,38 However, since the SEIPs framework focuses on the patient journey, I suggest a widening of the journey concept to include the interaction of several journeys, including not only patients and personnel, but also tools, and information, in relation to the potential event or object of study.