INTRODUCTION: FROM LINEARITY TO COMPLEXITY

As medical error statistics have become more widely disseminated among healthcare organizations, patient safety initiatives have proliferated. These, however, tend to be ad hoc rather than systematic, and both poorly researched and theorized. A major issue with framing research-based interventions in the field of patient safety has in general been reliance upon linear, problem-solving methods (e.g., modeled as a jigsaw where missing components can be identified, and a solution can be engineered). Embedding patient safety initiatives, however, is not just about establishing linear protocols, but changing the habits of complex organizations, beginning at the level of the clinical team. For this, we need more knowledge on complexity theory.1,2 With the introduction of Safety II, there is upcoming attention for the complexity in health care. Functional Resonance Analysis Method (FRAM) is a good start when adopting a Safety II approach, but we believe that “work as experienced” and the cultural-historical aspect is underexposed in this method.3

Theoretical framing of patient safety interventions requires productive alignment between learning theory and complexity theory. In this article, we illustrate one such alignment—between Cultural-Historical Activity Theory (hereafter CHAT)3,4 and ideas drawn from complexity theory. While this article acts as a primer for the application of CHAT to patient safety issues, our original contribution to the literature rests with aligning CHAT with complexity thinking as a basis to clinical practice improvement aiming for better patient care. The shift from linear to complex thinking is described in terms of informing values. Where linear teamwork models habitually draw on instrumental values, complexity models embrace a broader, holistic view. These values will be explored and explained throughout the article. First, as background, we will summarize the origins of CHAT.

In the wake of the Russian Revolution of 1917, a new wave of psychology emerged focused on how we learn. This was spearheaded by the psychologist Lev Vygotsky, who developed a model of learning termed ‘activity theory.’ In line with materialist philosophy developed from Marx, any ‘theory’ must be practical. It is only through ‘doing’ (activity) that learning is accomplished. Vygotsky suggested that all learning is object(ive)-oriented or focused on something to be achieved that outstrips current ability and then ‘expands’ activity. Here, cognition is not just located in the learner, but widens to include use of material objects or artifacts (tools) and other persons, including experts, where learning can be ‘scaffolded.’ Here, the learner is helped to move from the known to the unknown. Such a view of learning—as activity-based process, materially enhanced through artifacts, scaffolded, and expanded by collective learning—was not adopted in the West, particularly North America, where a spirit of ‘self-help’ individualism reigned.4 In the 1970s, however, a Finnish educationalist Yrjö Engeström developed Vygotsky’s ideas, graphically representing learning as a set of conversations between the components that Vygotsky had mapped.5

First, a learner (‘subject’) is oriented to a goal (‘object’ or conceived outcome of learning). To meet that goal, a learner draws on material artifacts and a community of practice of other learners. Engeström added to this the fact that within the community of practice there are specified rules of engagement and division of labor or roles. Here, the activity system constitutes a learning ‘climate’ within a ‘culture.’ The conventions of that culture are established historically; thus, the full descriptive term ‘cultural-historical activity theory.’ importantly, each of the components of any activity system talks to and modulates every other component (see Fig. 1 below).

FIGURE 1:

Schematic representation of an activity system.

One dynamic activity system (such as a clinical team) never acts alone. The system is always in conversation with other systems (e.g., a surgical team with a team of anesthesiologists), and then potentially expansive, as shown in Figure 2 below. This visual modeling tends to artificially freeze activity in time and space, where the reality is that learning is fluid, driven by an ongoing process. To realize that process element, in this article, we put CHAT in conversation with the dynamics of complexity thinking.

FIGURE 2:

Schematic representation of 2 related activity systems.

Teams are not problems to be solved but activities to be expanded. CHAT acts both as a theoretical framework explaining and exploring learning activity and as a set of practices. For example, CHAT theorizes that open, 2-way conversation between the elements of different activity systems will lead to expanding the ‘object’ or aim of the activity (such as a team aiming to embed democratic habits). A set of practices can then be planned and executed, such as communication skills training for dialogical briefing and debriefing in a surgical team—the latter aiming to overcome ingrained habits such as surgeons resorting to authority-led monolog within an assumed hierarchy (illustrated in the penultimate section of this article).

RESISTANCE TO COMPLEXITY: A PARADOXICAL AND USELESS FORM OF “SAFETY”?

This article explores how patient “safety” can be reframed as a series of calculated innovations rather than static protocols through appropriate application of a complex model: CHAT.6,7 Where people work together intensively and across disciplines, as in healthcare, complexity is inevitable.1,2 “Complex” should be distinguished from “complicated.” Sending a person to the moon is complicated, but can be engineered through predictable, linear, closed systems planning and execution. Jigsaws can be complicated but ultimately all pieces fit to form the whole. In contrast, complexity refers to unpredictable, dynamic, and open systems, demanding high levels of adaptability. Much clinical work is like this, often referred to as a “VUCA world,” an acronym coined in 1987 based on the leadership theories of Nanus and Bennis, referring to “volatile, uncertain, complex, and ambiguous.”8 In such a world, everyday practice cannot be fully regulated by protocols or guidelines. Nor can it be assumed that there is a unified and shared mental model of such practice, where “work-as-imagined” often deviates from “work-as-done,” or how that work is ultimately performed, a discrepancy that has been known for some time but is only now receiving increasing attention within healthcare research.3,9

If we want to improve the quality and safety of care, it is important to shift perspectives from the linear to the complex, so that planned interventions match complex realities. A key intervention is education into tolerance of ambiguity for healthcare practitioners.8 While complexity in healthcare is the lived experience (work as experienced or WAE), understanding such complexity as is often lacking in both practice and policy. This may be explained by the obvious fact that complex systems are necessarily complex to grasp and more so to respond to innovatively. Thus, there is a tendency to resort to linear, instrumental explanations and practices, seeking closure. The adoption of instrumental or functional values to inform and shape such activities refuses other, enriching, values that may increase quality of diagnosis, treatment, and care. But such values orientations—including the ethical (morality), aesthetic (quality), political (power), and transcendental (meaning) invite uncertainties, ambiguities, and paradoxes.7 In contrast, instrumentalism invites closure—solutions, decision making, and problem solving rather than the more complex “problematizing” and “sensemaking.”10 Again, policy will often be described using linear, engineering visual and verbal metaphors—such as completing a jigsaw, or viewing good teamwork as cogs fitting together—eschewing the application of an imagination of complexity fit for purpose.11,12

We cannot duck out of the challenges of learning to tolerate ambiguity and uncertainty as we adopt complex approaches to complex issues. Indeed, we can come to learn how to utilize contradictions as opportunities rather than threats. Here, a conservative “will-to-stability,” or unproductive habit, can frustrate opportunities for productive change, as “spearheads for innovation.”11

To benefit grassroots practitioners, such models linking complexity and expansive learning must be graphically and linguistically clear, while sophisticated. Developed in particular by Engeström and colleagues, CHAT is one such model.6,13–15 In this article, we show how CHAT can be applied specifically to the arena of patient safety, satisfying our criteria of: providing a stepwise approach to understanding complex learning; adding to the broader project of democratizing healthcare through patient involvement and “horizontal” interprofessional teamwork; and providing a framework to artfully inform and shape research activity.

THE LANDSCAPE OF CHAT

CHAT takes the spatial collective, or social, as the unit of analysis, rather than the individual. It then adds motion or time to this, as “activity” based in “history” (such as historically conditioned cultural habits of teams). All healthcare involves at a minimum a dyad of practitioner and patient, while usually patient care is employed through clinical teams (however fluid), and then communication between such teams multiplying up the actors.16 At the heart of such team enterprise is dialogical reflexivity—the ability to individually and collectively make sense of activities and initiate necessary change through dialog, while pursuing ways of improving those activities for better patient care and safety.17,18 Here, expertise is shared and learning democratic habits is paramount, yet historically determined forms of resistance create barriers for the ready uptake of models such as CHAT. Such resistance is grounded not only in a habitual retreat to comforting stability through simplification (as noted above) but also in masculinist and individualist traditions within a dominant discourse of medicine that resists feminizing and collaboration as a natural habit, leading to a persistent return to vertical hierarchies over horizontal forms of power.12,13,19

CHAT, as we shall see, is uniquely placed to address such issues as varieties of patient safety concerns where its primary focus is on “boundary crossings,” extending from interdisciplinary healthcare contexts to cross-border migrations and displacements.20 Grounded in post-Marxist dialectical thinking (where resolving oppositions generates new insight), teamwork is best achieved not in didactic hierarchies but in genuine democratic conversational structures. Here, CHAT is simultaneously a theoretical framework, pedagogic intervention, guide to research, and a social justice activity.

An activity system is a fluid, complex process where component parts (“agents” in complexity theory terms) are constantly in interaction over time (hence cultural-historical). The interacting parts have differing degrees of investment or are given differing values. The activity (such as care of a particular patient by a particular clinical team), as it moves through time, is focused always on improvement (‘expansion’ of care and safety), so it is constantly reflecting on what it has done (history) and what it can achieve (outcome or objective). The primary, and fluid, product of the process of care is the health and safety of the patient as “object” of the activity system.

As introduced earlier, expansion of activity can be modeled as the to turn bottom-line instrumental and technical care and safety (the bare minimum) into a much richer experience by generating and embracing a range of values and qualities at a higher register of knowledge, practice, and meaning. This involves a transformation of the mere technical-instrumental (the signature value of contemporary healthcare in a neoliberal age) into ethical, aesthetic, political, and transcendental forms and interests. Ethical practice is expected, but such practice should be “formed” or elegant (aesthetics), politically sensitive (embracing democratic habits), and oriented to giving meaning to patients’ outcomes and practitioners’ working practices (the transcendental).

As the array of persons and practices interact around producing the best possible outcome (the “object” as patient experience), several factors interact. The subject (any one of the healthcare practitioners) is invested directly in the object (patient care and safety) only as these are mediated by artful use of artifacts (mentioned above), and the formation of a working community (say, a clinical team). Again, participants in the activity are embedded in sets of rules (protocols, conventions subject to modifications) and division of labor or roles (who does what and why, who is qualified to do what, who can improvise in an emergency, who might whistle blow in an ethically transgressive incident, and so forth). Rules and roles of course generate subjectivities within the working culture as professional identities, relating back to the “subject” node or attractor of the complex activity system. This is the blueprint model of an activity system, illustrated earlier in Figure 1. Such a system inevitably interacts with other systems through boundary crossings, expanding both activity systems (Fig. 2 above).

Where translations across systems fail (e.g., in poor communication or technical errors) the activity systems involved collapse or crystallize.10,17 This may be because of institutional managerial economies such as budget cuts, perhaps initiated by wider political decisions, which can compromise patient safety.7

Frustration of expansion of the activity system (including interaction with other systems) results in regression to stable, more instrumental functioning, restricting innovation. To counter this, we need carefully designed input to strengthen values: economic (e.g., increased resources), ethical and political (e.g., redistribution of resources to help disadvantaged communities), aesthetic (e.g., a qualitatively more sensitive way of working), and transcendental (e.g., restoring meaning to a workforce that is burned out and disillusioned).

ARTIFACTS AS “ACTORS”

In the age of artificial intelligence, artifacts, instruments, or tools are increasingly key elements, or “actors,” in expanding activity systems.21 We “think” and “act” through extensions to our minds and bodies, some now historically old (spectacles, walking canes, stethoscopes) and some recent (cardiac stents, computers, mobile phones). Contemporary healthcare is now of course reliant on technologies such as laboratory blood and urine testing, imaging, surgical equipment, prosthetics, and pharmaceuticals. Effective clinical teamwork involves entangled conversations across intersecting complex, dynamic systems that include the “voices” of such significant “circulating” artifacts.17 Indeed, expansion of an artifact may become the object of the activity system (e.g., an organ transplant or insertion of a pacemaker).

As an artifact, such as a tissue sample, passes from one activity system (e.g., surgery) to another (e.g., pathology laboratory) so the artifact (actually, the patient) is treated differently in each context. Activity systems must therefore learn to “speak” with one another in understandable translations (as well as to patients and families). While healthcare practitioners may use reductive technical language that is readily shared across activity systems’ boundaries, patients need, nontechnical language (often metaphors) for ease of translation.18

“Safety” of course is a key concept in this article, but our contention is that such safety is not achieved through resorting to habit, but rather in looking to future-focused innovation. Ironically, expansion of an activity system is more likely to occur through calculated risk and use of (rather than resistance to) contradictions. For example, Gabbay and Le May’s groundbreaking work shows that physicians tend to follow inherently ambiguous local, culturally established “clinical mindlines” rather than standardized evidence-based “clinical guidelines” (including protocols) as a means to craft patient care interventions.22 Here, the individual case in vivo is taken as the marker, rather than the abstract and generalized case in vitro, where “work as done” is preferred to “work as imagined.”22

CONTRADICTIONS AND DEMOCRATIC DIALOG

Systems seeking to expand through coordinated activity may fail to do so as that coordination fails to materialize. Here, activity systems may crystallize or collapse where attempted boundary crossings fail or are frustrated (e.g., by retreat to unproductive habits or differences that cannot be reconciled). Teams may fall apart because of personality clashes or disagreements about methods, and teams meeting other teams may have large areas of disagreement or differences in practice approaches, leading to conflict.

However, as noted, conflict or contradictions can be beneficial in promoting change, sparking teams to shift practice patterns and abandon unproductive habits. For example, there are sometimes very different ideas about concepts such as “critical patient,” “severe bleeding,” or “monitoring and surveillance.” Such differences can subsequently lead to friction. Because the underlying social, cultural, and/or historical explanation is not immediately visible, and thus does not become negotiable, such frictions can persist for a long time. However, even within the same activity system, frictions can exist between its various components, mostly historically grown. For example, the instrument used may no longer be suitable in relation to the desired goal, or the rules within the community may conflict over time with changing patterns of division of labor.

With CHAT, these frictions do not have to be eliminated—as traditionally tends to be the case within patient safety thinking—but are instead a starting point for learning through contradiction, employed as resource.19 Here, learning is not seen as an individual, personal matter, but as a collective event seeking meaning.23 The collaborative, dialogical-democratic basis to CHAT creates reflexive spaces: environments in which an awareness of the various visions and experiences between the different caregivers arises and ensures that alignment becomes possible.24

FROM THEORY TO PRACTICE

Thus far in this article, we have concentrated on CHAT as a theoretical framework. In this section, we give an illustrative example of a patient safety intervention informed and shaped by CHAT. As a longitudinal training intervention, a program called “Theater Team Resource Management” (TTRM) was introduced to one hospital in the United Kingdom, focused on improving patient safety through informed, reflexive teamwork in surgical teams.24 The theoretical framework for the intervention was CHAT. Because of a serendipitous situation, one half of the surgical theater personnel across the hospital site worked in a different location to the other half, and there was little movement or communication between the 2 locations. There was thus a readymade situation where one group could be used as an experimental cohort and the other group as the control cohort. (After 1 y, when results from the intervention could be established, the control cohort also received the educational intervention).24

In the experimental group, surgical teams were videotaped in live practice. Videotaped examples would be edited to provide illustrate examples of excellent and poor communication and used to provide feedback and to generate discussion and action plans, in debriefing sessions.24 Data were gathered as follows: (1) quarterly validated Safety Attitude Questionnaire returns to gauge the quality of team climate and culture; (2) close call reports analyzed for content; and (3) videotaped surgical activity analyzed for patterns of communication such as incidence of facilitative communication (dialog) as opposed to authoritative communication (such as monolog, statements rather than questions, and closed questions).

When compared with the control group, the experimental group in this study created a significant safety climate over time: (1) increasing scores on the Safety Attitude Questionnaire and scoring significantly higher than the control group that remained at a stable level; (2) establishing democratic habits, where briefing and debriefing (led by nurses and not surgeons) were established; (3) surgeons in particular showing increased situational awareness and a significant shift to dialog, away from habitual monolog; (4) nurses were empowered to ‘speak up’ where patient safety was at risk; and (5) individuals identified as ‘team’ players even in highly fluid teams, challenging habitual identities and roles to turn multiprofessionalism into authentic interprofessionalism,24 In transforming historically autocratic structures into collaborative structures, there was a shift first in surgical team ‘climate’ (now open to patient safety issues) and then team ‘culture’ (now democratic). Importantly, team members showed an active interest in better understanding, and deepening, of patient-centered practices, with reference to ethical, aesthetic, political, and transcendental values, away from bottom-line functionalism.25 CHAT-informed feedback sessions further led to greater understanding of how innovation in roles (division of labor) and rules (especially protocols) could occur, shaping identity formations.26

CHAT AND VIDEO REFLEXIVITY ETHNOGRAPHY

Crucial to the TTRM project described in the previous section was the research method of debriefs facilitated by video feedback. CHAT is primarily neither a research methodology (a theory of research) nor a discrete method (a way of doing research), but rather a grand narrative concerning collective activity. However, in this guise, it can readily invent, inform, and shape research activities artfully. Put simply, CHAT does not normally tell you primarily how best to collect data, but rather how to make collective sense of data. However, CHAT can act to shape data collection methods as forms of innovation, such as the video reflexivity (video reflexivity ethnography) approach adopted for the TTRM project.27

In both cases, video feedback was used in structured and facilitated debriefing settings within a CHAT framework to generate boundary crossing activity between teams as a means of improving patient safety. Video reflexivity ethnography involves collecting ethnographic data of daily work practices through real-time video and audio recordings. In the TTRM project, longer sections of footage were converted into short video clips by seasoned theater practitioners (nurses, anesthesiologists, and surgeons) working with researchers (in this case, psychologists and medical educators). The edited footage was later played in debriefed reflexivity sessions for the participating professionals. The ensuing dialog can be considered a form of collaborative research: the meetings designed to jointly examine specific, everyday clinical practices with the aim of reflecting critically on what members of the community or different communities tend to take for granted.

Through CHAT, information (data) is not worthwhile until it is transformed into knowledge and then into meaning, as in our example of video reflexivity. As described by Frambach et al, CHAT readily embraces different qualitative data collection methods such as interviews, simulations, observations, and focus groups.28 CHAT is a quintessential metacognitive (thinking-about-thinking) process.

In addition, at the Maastricht UMC+ video reflexivity was used to reflect on shared safety practices (in this case putting on sterile gloves in the neonatal intensive care unit).24,29

Here, however, CHAT was not used as an explanatory theoretical framework. In a follow-up study, in which video reflexivity ethnography will be applied within the emergency department during the care of trauma patients, CHAT will be used to give meaning to the data and to generate insight into, for example, interdisciplinary collaboration.

CONCLUSIONS

We encourage readers unfamiliar with CHAT to investigate its potential for medical and healthcare education and research. This does not need to be centered on patient safety, the focus of this article, but on any issue of practice. CHAT’s wider value is that it is not limited to epistemology as exploratory or explanatory modeling but describes an ontology or way of being, and an axiology or set of values informing that way of being. CHAT at its best is a lifestyle grounded in democratic habits and is restlessly innovative.