Concerns about the role of communication failures in adverse events coupled with the success of checklists in addressing safety hazards have engendered a movement to apply structured tools to a wide variety of clinical communication practices. While standardised, structured approaches are appropriate for certain activities, their usefulness diminishes considerably for practices that entail constructing rich understandings of complex situations and the handling of ambiguities and unpredictable variation. Drawing on a prominent social science theory of cognition, this article distinguishes between two radically different modes of human thought, each with its own strengths and weaknesses. The paradigmatic mode organises context-free knowledge into categorical hierarchies that emphasise member-to-category relations in order to apply universal truth conditions. The narrative mode, on the other hand, organises context-sensitive knowledge into temporal plots that emphasise part-to-whole relations in order to develop meaningful, holistic understandings of particular events or identities. Both modes are crucial to human cognition but are appropriate responses for different kinds of tasks and situations. Many communication-intensive practices in which patient cases are communicated, such as handoffs, rely heavily on the narrative mode, yet most interventions assume the paradigmatic mode. Improving the safety and effectiveness of these practices, therefore, necessitates greater attention to narrative thinking.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
The identification of communication failure among the most common root causes of adverse events1 ignited a movement to improve communication-intensive practices in healthcare. Among these practices, the handoff has received considerable attention, due to concerns that poor communication threatens patient quality of care and safety during transitions of care. Researchers have cited variability in practice, fallible human memory, omission of important information and distractions in the physical environment among the many factors that compromise effective communication during handoff.2
Efforts to improve handoff communication and other patient safety challenges have been heavily influenced by the remarkable success of checklists in reducing central line infections3 and surgery-related complications and mortality.4 Researchers have implemented checklists, mnemonics (eg, SBAR) and similar tools in order to reduce practice variations and standardise transmission of information.2 ,5 While many of these efforts certainly led to improvements, checklists are by no means cure-alls nor replacements for critical thinking,6––8 and it is vital we understand their limits.
For example, the remarkable successes of the Michigan Keystone program3 and the Matching Michigan project9 both entailed far more than implementation of ‘simple checklists’. A host of social and organisational changes accompanied those implementation efforts, and without the attendant changes, it is highly unlikely that the checklists alone would have produced the significant reductions in infections and mortality.6 ,10 ,11 Similarly, checklists, once implemented, may engender a ‘false sense of safety’ if levels of compliance drop and other safety practices have been replaced by the checklists.12
But beyond these implementation and sustainability concerns, the usefulness of a checklist is further constrained by the type of cognitive challenge it is best suited to address. The success of checklists stems largely from their abilities to outline step-by-step instructions for technical procedures for which the particulars of context are immaterial, reliance on human memory is a known problem and variations in those procedures are undesirable.7 Similarly, mnemonics work by prescribing and sequencing broad categories to be covered during information exchange. But the usefulness of checklists and mnemonics alone—indeed, even the appropriateness of relying on such highly structured tools—diminishes when the activities in question involve ambiguous, unfolding, complex situations. In other words, checklists are most appropriate for simple or complicated problems, not complex ones.
Glouberman and Zimmerman13 distinguish between simple, complicated and complex problems. Simple problems (eg, baking a cake) require little expertise and can be addressed using highly standardised, formulaic solutions. A complicated problem (eg, sending a rocket to the moon) arises from multiple causes and involves multiple parts, but it can be solved by breaking it down and managing it piece by piece.13 In contrast, a complex problem (eg, raising a child) also involves multiple causes and parts, but complexity emerges out of the interaction of the parts, such that one must deal with the whole rather than the individual parts. Inserting a central line is a complicated task that can be broken down and outlined in a checklist; whereas understanding the clinical course of a specific patient is a complex and evolving task, in which one must grasp a holistic sense of the interdependencies of biological, pharmacological, social and other interacting processes. Distinguishing between complicated and complex tasks is important because social science theory and research suggest that humans rely on different modes of thought when dealing with these different types of problems. Specifically, humans frequently rely on paradigmatic thought when dealing with simple and complicated problems, but narrative thought becomes most important when dealing with complex problems.
Paradigmatic and narrative modes of thought
Cognitive psychologist Jerome Bruner identified two distinct modes of thought—the paradigmatic (also, ‘logico-scientific’) and the narrative—each with its own unique way of ordering experience and constructing knowledge.14 Neither is superior to the other, nor are they alternatives for one another. Each contributes to the richness of human cognition by providing strengths suited to different kinds of cognitive challenges. We all depend on both modes in our everyday existence. Table 1 provides a comparison.
The paradigmatic mode of thinking organises knowledge into hierarchical categories, classifying phenomena according to categorical memberships.14 ,15 Its quest is to identify and apply universal truth conditions. Thus, it is context-free, dealing in generalisations. The particulars of circumstance are irrelevant in the paradigmatic mode. Instead, the phenomenon (eg, event or entity) is abstracted from context and matched to its appropriate position within the categorical hierarchy.15 The paradigmatic mode is procedural, technical, rigid and easily reproducible, deftly handling all phenomena that fit neatly into its categorical scheme. Its usefulness breaks down, however, in the face of ambiguity and unpredictable variability, which make categorising phenomena difficult if not impossible. The paradigmatic mode is materialised in taxonomies, lists and standard operating procedures. We evaluate the products of paradigmatic thinking on their verity—their ability to withstand falsification.14
By contrast, the narrative mode of thinking organises knowledge temporally into a plot, linking specific events into a unified whole by emphasising consequential connections among them.14 Its quest is to connect disparate events and information into a meaningful whole—in short, to make sense of a situation. As such, it is highly sensitive to context, dealing in the particular. The narrative mode is adept at handling anomaly, uncertainty and unpredictable variability, which it readily incorporates into its unfolding plot.14 The narrative mode materialises in the form of stories. It is the means by which we establish identities (our own and those of others) and make sense of the continual stream of experience we call life. Finally, the narrative mode aids comprehension16 and appears to be a crucial means by which we organise, retain and recall memories.17 We judge the products of narrative thinking by their verisimilitude, or conformance to what we take to be lifelike.15 ,16 ,18
Each mode of thought is exercised daily in the clinical arena,19 and many tasks may draw on both modes. The checklist for central line insertions provides a stellar example of the paradigmatic mode in action.3 Checklists work because a generalised knowledgebase exists to categorise the practice, irrespective of context. The singularity of the particular patient matters little: the steps required to ensure a sterile insertion do not vary from one instance to the next. In the controlled environment of an intensive care unit, no stories are needed to accomplish the insertion—only a careful adherence to a predefined, well understood procedure. In fact, variation or divergence from standard procedures in such cases can lead to complications and poor patient outcomes.
On the other hand, the act of communicating a patient plan of care—such as when handing off responsibility at shift change or transferring the patient to a different unit—draws heavily on the narrative mode of thinking. Unlike the central line example, telling a patient's story entails focusing on the particulars of the case at hand—rather than abstract generalisations—and attending to uncertainties and ambiguities.
Narrative is central to the practice of medicine.20––22 It is the means by which patients inform doctors of symptoms and doctors inform patients of disease, diagnosis, treatment and prognosis processes. Narrative is the cognitive structure of the case report.21 It is through stories that errors, near misses and adverse events are processed—by both clinicians and patients—and held up as examples to galvanise support for learning and system change.23 ,24 In the following section, using the example of handoff, we explore in greater depth why communicating a patient's care necessarily involves the narrative mode.
Handoff and the narrative mode
Many handoffs include elements of paradigmatic thinking, such as when details about the case are presented in standardised, categorical formats. For example, in box 1 a nurse presents a patient by listing values corresponding to general vital sign categories. This sample handoff excerpt and others provided below are modified from actual handoffs we have observed.
Handoff paradigmatic mode example
Nurse: He's alert and oriented, Blood pressure was 145 over 97 initially; however, his heart rate is 129 sinus tach. Afebrile on 90% room air. Really a nice guy. He really wants to eat.
While the paradigmatic mode can provide an overall structure to the conversation, or at least to portions of it, the complexities of many cases require narrative thinking to be employed as well.20 ,21 Explicit attention to the role of narrative, however, is largely missing from current research focused on improving the safety of communication-intensive practices such as handoffs. Admittedly, some commonly used improvement approaches, such as collaborative cross-checking25 and crew resource management,26 employ practices that touch on the narrative mode of cognition, but even these approaches leave many aspects of narrative thinking underexplored. In the following sections, we discuss four aspects of handoffs that demonstrate why the narrative mode is crucial.
The singularity of the case is most important
First, every patient case is unique in ways that bear directly on understanding it.20 ,21 In fact, one of the most crucial functions of handoff may be to highlight what is unique about the case or care plan. For this reason, the narrative mode is essential to activities involving the communication of patient cases. Even where standardised patient stories may be established, handoffs will necessarily entail careful attention to the ways in which the present case deviates from those standardised ones.27 It is the singularity of the case—the particular details of this patient—that holds the most informative clues, not only for understanding the patient, but also for anticipating the future trajectory of illness and for the selection of approaches to care that are most likely to provide benefit.20 To be certain, generalised knowledge of disease and treatment are essential, but knowing how best to apply that knowledge to a specific case often demands a rich understanding of the case itself.21 In short, generalised knowledge must be adapted to the specific details of the present case through narration. Box 2 provides an example of a physician in a critical care unit using the narrative mode to relate specific details in the unfolding story of one patient.
Handoff narrative example 1
Physician: He was brought to the Emergency Department when his family found him coughing up some bright red blood. In the ED, GI did try to scope him initially and ended up having to intubate him to complete the scope, and they saw an oesophageal ulcer with a visible vessel, and they squirted some epinephrine in there. They were able to evacuate about 80% of the clot, but they wanted him to remain intubated, come up to the ICU, and then they want to re-scope him tomorrow.
Understandings of patients must be co-constructed
Second, comprehending a singular patient case involves constructing a coherent, holistic understanding of that case: a whole that is greater than the sum of its parts.28 The tendency to think of handoffs as information transmissions has led to a focus on identifying the necessary pieces of information to be transmitted and on measuring the accuracy and completeness of transmission and reception.2 ,27 Such an approach presumes the paradigmatic mode of thinking: generalised categories to be applied to all or subsets of cases. What the information transmission perspective overlooks, however, is the work entailed in assembling those various pieces of information into a coherent, meaningful whole. Here again, it is the singularity of the case that matters most. Even if it is possible to identify a required set of categories of information to be covered during the communication, a holistic understanding of the case does not automatically arise from reception of various pieces of information about the case, nor are the interconnections among that information objectively apparent. Understanding is the product of co-construction.28 It is only by the narrative mode that a holistic understanding of a specific series of events and other details can be constructed. The process of that construction is known as emplotment,29 and it emphasises the active role of the narrator in building the plot.20
In narrating stories, we do not simply relate events in chronological order.29 Instead, we posit connections between those events, including causality, by ordering our stories in particular ways, supplying details at specific points in the narration where their significance to other details can be asserted.16 The novelist E M Forster famously explained: ‘“The king died and then the queen died” is a story. “The king died, and then the queen died of grief” is a plot.’30 By adding the simple preposition ‘of grief’ at a key point, the narrator asserts a relationship between the two events, shaping the audience's understanding of each of those events. Thus, emplotment is crucial to the construction of meaning and identity, the arguing of causality and the demonstration of human intentions. Clinicians engage in emplotment frequently. As Charon20 notes, ‘Diagnosis itself is the effort to impose a plot onto seemingly disconnected events or states of affairs.’ When we recognise the narrative mode is crucial to handoff, our attention is turned not only to the specific pieces of information to be included in the telling, but also to the way in which the various details of the patient's case are fitted together into a plot.
For example, through an interaction of questions and answers, the two residents in the handoff in box 3 engage in narrative thinking as they co-construct an understanding of one aspect of a patient's treatment. In addition, we see emplotment in this example as well, as Resident 2 explains the intentions that guided the choices and sequencing of medications administered. Emplotment is also illustrated in the handoff in box 2, when the physician describes the related, unfolding events in a patient's admission to the critical care unit.
Handoff narrative example 2
During handoff, two residents discuss medications for an insulin-dependent diabetic patient with a history of gastroparesis and chronic kidney disease who ‘isn't very compliant with her medications’.
Resident 1: Does she have an antiemetic that she likes?
Resident 2: She likes Phenergan.
Resident 1: Oh, so Phenergan does work for her? Tried Compazine or anything like that?
Resident 2: Not yet. So, she has Phenergan and Regalan at home, and those are the two she likes. Compazine is what I would go to next. Zofran I don't think does a single thing for her. I just figured: try the Phenergan because she likes it, and then I tried the Regalan, because I figured with gastroparesis, the body would like it.
Resident 1: Sure, sure.
Ambiguity and uncertainty must be accommodated
Third, narrations of patient cases often occur while other relevant activities are still unfolding, and therefore, while understandings of those cases are still evolving. Consequently, in telling and interpreting patient stories, clinicians must deal with ambiguities, uncertainties and emerging contingencies. Whereas the rigidity of checklists, modelled on the generalised knowledge that is central to the paradigmatic mode of thinking, cannot readily incorporate unexpected variation and ambiguity, the flexible structure of narrative easily stretches to accommodate the emergent and unpredictable. In fact, emplotment is an act of imposing order on what might otherwise be disorderly and of conjecturing where insufficient information produces uncertainties or inconclusive data gives rise to ambiguities.14 If the narrative mode has a weakness of which we should be aware, it is likely the fact that it can so easily accommodate equivocality, including to the point that one might over-rationalise or underemphasise some ambiguity, thereby presenting an inaccurate or overly simplistic understanding of the case.16 This danger arises not from an inherent flaw in narrative thinking, but rather from the way such thinking may be practiced as individuals deal with norms and expectations. For example, where doctors feel pressured to appear knowledgeable and confident, they may be more likely to tell stories in ways that downplay uncertainties. The solution, then, lies not in faulting the narrative mode of thinking, but rather in adjusting the socio-cultural norms that shape behaviours. In box 4, a physician uses narrative thinking to acknowledge uncertainties and open up thinking to possibilities beyond the accepted diagnosis.
Handoff narrative example 3
Physician: She's had frequent bouts of pneumonia in the past, and it's usually the same findings on chest X-ray, but this is worse than previous X-rays have been. So, part of me says this may actually not be pneumonia at all, and some other process is really driving all of this—that we're just watching things get progressively worse and worse and calling it pneumonia.
Patient stories must be adapted for the audience
Finally, handoff situations vary in terms of the relative status and experience of the involved parties, their prior knowledge of the patient, and the extent and quality of their existing relationship.2 ,31 Effective handoff communication—and by extension, effective co-construction of understanding—demands adaptation to these factors.
There are many different ways to tell the same story, but the ‘best’ narration is one that is adapted to this inter-subjectivity of narrator and audience. Events may be sequenced differently as a way of emphasising different details, asserting different ideas or arguing different interconnections. If narrators are to create in their audiences the understandings they hope to create, however, they must attend to the perspectives of their audience and let these perspectives shape the narrating. When individuals engage in perspective-taking,28 ,32 they attempt to see matters from another's point of view and to understand what that person might be thinking. Gaining insight into another's perspective enables more effective communication because speakers can better interpret their audiences’ emerging understanding, anticipate needs and adjust communication accordingly. Perspective-taking highlights the fact that effective narration entails sensitivity to context and complex, nuanced social interactions that cannot be neatly predefined by a checklist. The failure to consider another's point of view may contribute to ineffective communication during handoff,33 and a few researchers have begun exploring the influence of various perspective-taking practices on these interactions.34 ,35
All of this suggests that the narrative mode may be highly important in handoffs involving parties who are particularly inexperienced or to a clinician who has had no prior experience with the specific patient. In such instances, more work is often required to make explicit links between events or among details. For example, the detailed narrative relayed in box 2 is necessary insofar as the receiving party is unfamiliar with the details surrounding the patient's admission. By considering the perspective of the receiving party, the speaker is able to predict what may be the most appropriate level of details to provide.
Storytelling has been fundamental to human communication since at least the beginning of recorded history, as our ancestors’ drawings on cave walls would attest. Far from being merely a tool for entertainment, narrative is integral to the way we make sense of unfolding situations, form identities, address the unique aspects of particular situations, and cope with contingencies and ambiguities. The narrative mode of thinking is not inferior to the paradigmatic mode14; it is complementary. Both modes are crucial to many clinical tasks, particularly handoffs. We have focused on the narrative mode only because it has been largely overlooked in reported handoff improvement efforts. If the narrative mode is more chaotic, less predictable and less amenable to standardisation than the paradigmatic mode, that is because its chief task is to cope with the complexity of an evolving, unpredictable world that is never perfectly knowable.16 As our social world and care models grow increasingly complex, the importance of the narrative mode will likewise increase.
Where checklists and mnemonics are used to structure communication-intensive practices, it is highly likely that clinicians are still engaged in the narrative mode as they co-construct understandings of unfolding cases. While the checklists and mnemonics used in such scenarios would guard against forgetting particular details, it is not likely that such tools offer any assistance with the challenging work of crafting plots and co-constructing useful, holistic understandings of patients. For example, Hu et al36 have shown that during complex, unanticipated situations in the operating room, which is heavily governed by the checklist mentality, human resilience is what rescues the provider and patient. Resilience is the ability to adapt to and absorb variations and disturbances, particularly those that fall outside of what the system has been designed to accommodate.37 We argue that it is the narrative mode of thought that is most essential to such resilience. Therefore, to build more resilient systems of care, we must supplement current improvement efforts with approaches that honour the strengths and challenges of narrative thinking. What such approaches might look like, we can only speculate since we are not aware of any research that has attempted to examine how different narrative practices shape comprehension or effectiveness of handoff. A starting place might be simply to pay closer attention to how specific patient stories are being narrated and to encourage clinicians to examine what it is about certain narratives that make them easier or more difficult to comprehend. In the meantime, research into the narrative structure of effective handoffs would deepen our understanding of how transitions of care might be further improved.
To be clear, we are not arguing against the value of checklists but urging caution relative to simplistic applications of such approaches.6 ,7 Indeed, when incorporated into a multi-dimensional, inter-professional patient safety programme, checklists can be an effective piece of the solution.38 They are not, however, the holy grail of communication. Our argument, then, is that we need to differentiate between complicated tasks that draw primarily on the paradigmatic mode of thought and complex tasks that draw on the narrative mode and then fashion interventions accordingly, recognising that many activities involve both types of tasks. The nearly complete absence in the recent communication patient safety literature of any discussion of the role of narrative thinking and the paucity of interventions shaped to address the unique demands of narrative tasks suggests that we are missing opportunities to intervene effectively. We may even run the risk of making matters worse.
The authors acknowledge the helpful feedback from Dr Vineet Arora.
Contributors BH developed and drafted the initial manuscript. SDM-B helped refine the ideas and contributed to the revisions of the manuscript.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.