Despite these variations, some common core mechanisms of resilience have been identified ([
3], p. 122) as the ability that individuals, communities, organisational units or larger systems have to return to some ‘normal’ condition or state of functioning after a disruptive event; to cope with pressure and problems by being flexible without compromising system performance; or to adapt to a new normal state, where system functioning is reorganised or enhanced in some way in response to the disruption they face ([
3], p. 122). In this view, there is an emphasis on the individuals’, communities’ or organsations’ ability to regain equilibrium in circumstances of changes, or to adapt to new norms, forms, and practices. Either way, the focus is centrally on processes of learning and changing for individuals, communities, and systems, coupled with remaking or transforming individual, interactive, or organisational practices. These mechanisms may help to explain how, in different ways and in different places, healthcare systems are able to deliver high quality care. Any integrative view of resilience [
3] in healthcare should be able to accommodate all of these different mechanisms at all levels of a system—while also being clear about the differences between them.
A brief overview of resilience definitions and theories
In the following, we briefly synthesise key relevant literature in the field of resilience and describe how resilience has been defined in these literatures. The intention is not to provide a comprehensive review, but to illustrate the different ways that resilience has been defined and the concepts and components previous research has focused on.
Aron Wildavsky [
21] is a key figure in safety research and defines resilience as «
the capacity to cope with unanticipated dangers after they have become manifest, learning to bounce back” (p. 77). A main feature in this conceptualization is that resilience deals with the dangers that have been realized (manifested) and learning emerges as a key element in enabling resilience [
21]. Wildavsky [
21] emphasises the notions of
active and passive resilience (Lovins & Lovins [
22] – in Wildavsky [
21]). Active resilience includes a deliberate effort to improve various abilities to cope with surprises and to use stress as a source of learning, thus actively benefiting from stressful situations. By contrast, passive resilience relates to the ability to bounce back and get back to normal procedures, after an adaptive change has happened, without any further development of skills or systems ([
21], p. 98). It follows that the link between resilience and learning is important to explore and understand, such as how learning processes are associated with adaptations in work performance in the face of manifest risks, dangers, and opportunities.
Parts of the resilience literature focus on societal planning and handling major disasters and extreme events. The main concern here is how to design systems able to withstand and respond to major crises such as hurricanes, earthquakes, floods, and terrorist attacks [
4]. In this this area both intentional (e.g. terrorist attack) and unintentional threats (e.g. earthquake) are included and Comfort et al. ([
4], p. 9) have defined resilience as “
the capacity of a social system (organization, city, or society) to proactively adapt and to recover from disturbances that are perceived within the system to fall outside the range of normal and expected disturbances”. The social system orientation is important in this area because entire societies, communities, and organisations need to mobilise to adapt and recover from massive unpredicted and large scale disruptions, and return to the normal functioning of a society. Interestingly, the definition includes both proactive identification and prevention of risk as well as recovery from disturbances. That is, resilience can happen before, during or after the occurrence of a disturbance [
4]. In addition, in this view the definition of resilience separates individuals and the wider social system, focusing on the latter. This focus on resilience at the level of the social system may be advantageous in the study of organisational and societal responses to large-scale disasters. However, the role of individual actors is also key to understanding how resilience unfolds particularly around the numerous smaller scale disruptions that occur in healthcare [
13,
14,
23,
24].
In the safety science literature, the traditional approach focuses on adverse outcomes and ‘find and fix’ solutions, increasingly referred to as a ‘Safety I’ approach. Recently, a ‘Safety II’ perspective has gained interest in healthcare [
25,
26], focusing on the processes that support resilient healthcare. Safety II research focuses on learning from why things go right in order to improve safety. Understanding what works well is considered key to understanding what goes wrong [
9]. This means researchers should be interested in adaptations made to enable systems to work during stress but also during adaptation to positive changes such as new technology enabling new ways of working. The Safety II approach and Hollnagel et al. ([
10], p. xxv) define resilience in healthcare as “
a health care system’s ability to adjust its functioning prior to, during, or following changes and disturbances, so that it can sustain required performance under both expected and unexpected conditions”. The emphasis is on the system and its adjustments both before and after the disturbances in addition to “changes” that can provide opportunities for the system to transform [
26]. According to this definition, resilience is not only related to risk, hazard and danger – it also relates to positive deviances and changes in terms of success, opportunities and disturbances in positive ways [
3]. Importantly, resilience and adaptation may not always be beneficial for safety or positive for entire systems: there are circumstances where local adaptations become too extensive and could have negative consequences for the broader system (the ‘tragedy of adaptability’) [
27]. An additional trap is that resilience may become constructed in a reductionist and individualistic manner, where the sharp-end operators (e.g. nurses, doctors, patients, next of kin) are expected to handle and compensate for system-caused problems such as ill-designed or under-resourced systems. This can potentially result in burdening the sharp-end practitioners with the responsibility for the resilient performance of a system [
9].
This brief conceptual overview illustrates how some of the key definitions of resilience differ, that resilience must thus be understood as a concept with different meanings in different disciplines [
4], and that there are also potential traps if resilience is misunderstood or misapplied.