EMOTION AWARENESS AND INVISIBILITY IN AN EMERGENCY ROOM:  A SOCIO-TECHNICAL DILEMMA

Abstract

The expression of emotion is often an overlooked aspect of work and consequently is easily neglected in the design of collaborative information systems. However, in collaborative work, emotion expressions can play an important function in how each individual works as well as how the team works together. As work environments become increasingly computerized, the unintended consequence of the introduction of information systems on emotion awareness emerges. This can be a serious issue for certain highly coordinated, information intensive critical environments such as in the healthcare field.

In this dissertation, I investigate the occurrence of work-related emotion expressions in the real-world work environment of a hospital emergency room (ER). Effective communication and coordination between personnel in an ER can literally mean the difference between life and death. Due to the high rate of information sharing errors, hospitals are beginning to push for the integration of electronic patient records and other information systems in the hopes it will improve communication, and that this in turn will increase safety and efficiency. At the same time, there have been many examples of unintended consequences that have emerged as the result of the implementation of electronic patient records. A central premise of this dissertation is that the expression of emotion is one of these informal aspects of coordination that may be lost or mishandled in the move toward increased technology support.

To examine the occurrence and function of emotion expressions, I conducted a six-month field study of interaction in a large urban hospital in the Northeastern United States. This field study consisted of observations of ER personnel, formal and informal interviews, artifact analysis of formal ER patient records and administrative documents, and a reflexive diary to reflect on my own emotional reactions.

The first contribution of this dissertation is to explain why emotions are expressed in the ER. This includes uncovering the sources of work-related emotion in an ER, the functional roles of the expression of those emotions, and the consequence on the work of the ER. The second contribution of this dissertation is to explain how emotion expressions are modified in the ER. This contribution shows the effect of various stakeholders’ constraining and enabling behaviors on the expression of emotions in the ER. The third contribution is to expand and articulate the role of emotion expression in the development of emotion awareness in work. The final contribution is to discuss how these findings can guide the design of systems for the emotion awareness in the ER as well as discuss the possible unintended consequences of a greater technology driven ER.

 

Chapter 1. Introduction

Effective communication and coordination between personnel in highly collaborative environments such as hospital emergency rooms can literally mean the difference between life and death. Contextual factors in these environments such as time pressure, complex and ambiguous problems, stress, and interruptions (Orasanu & Connolly, 1993; Salas, Driskell, & Hughes, 1996) can raise strong emotions that must be managed in order to perform effectively as a team. The following is a vignette, which captures the intensity of these emotions as well as the repercussions on the team.

“Three young and tired residents in an emergency room were wondering why the pizza they had ordered had not come yet. A nurse shouted, ‘GSW^[1] trauma one, no pulse, no blood pressure’. The residents recognized the unconscious teenager. He was the delivery boy from their favorite all night restaurant. He had been mugged bringing their dinner. That made them work even harder. A surgeon cracked the kid’s rib cage to expose his bullet torn heart, but they could not even stabilize him for the O.R.^[2] After 40 minutes of resuscitation they called it. Time of death, four a.m. The young doctors shuffled into the empty waiting area. They sat in silence. Then one of them said what all three were thinking. Where is the food? They found their pizza where the delivery boy dropped it a few steps away from the emergency room’s sliding glass doors. They put it on the table. The overnight shift was brutal. The doctors were hungry; the delivery boy was already dead. But they just stared at the food. Then one of the residents spoke. ‘How much do you think we ought to tip him?’ They laughed. Then they ate.” (K. Watson, 2006) 

This story highlights the occurrence of emotions in the emergency room; the sadness and fear that the young residents felt after fighting death and losing. It also shows the toll the emotional experience took in that the residents could not even satisfy their base instinct to eat. And, like every profession that confronts suffering or death, the vignette shows that humor around tragedy is vital for effective functioning. The doctors needed to laugh before they could move on, before they could eat and be the best they could be for the next patient, or before they could recoup as a team and know they can rely on one another to recover for the next patient.

It is important to recognize the impacts of the occurrence and expression of emotions in critical work. By emotion, I am referring to a specific, identifiable affective state such as angry, afraid, happy, and excited. Such emotions can play an important function in how each individual works as well as how the team works together. In the design of information systems that support such environments, we cannot simply assume that work activities are rational and prescribed. We must be aware of how the participating workers feel, how collaborators express their feelings to one another, and how those feelings inform their individual and shared work. Developing these understandings is an essential step in the design and construction of information systems that provide effective support for collaboration in critical environments. Broadly, we must understand the role of emotions in collaboration in order to determine whether we need to design collaborative information systems to support emotion.

1.1     Problem Motivation

In a healthcare environment, effective information sharing and coordination between healthcare workers is essential. Fumbled handoffs have been depicted as the “Bermuda triangle of healthcare” (Landro, 2006). Errors or omissions that lead to patient harm can occur in the gap when a patient moves from the supervision or responsibility of one healthcare provider to another either during or at the end of a shift. Gandhi (2005) provides a compelling case of a missed tuberculosis diagnosis due to the lack of effective information sharing between caregivers. Despite multiple visits to the emergency room, a three-week stay in the hospital, multiple radiographs and CT scans, and the continued notation of possible tuberculosis from radiologists, the patient’s diagnosis of tuberculosis was delayed for four months, resulting in death. Gandhi argues that this tragic outcome resulted from simply the lack of directed and motivated sharing of the abnormal findings from the radiologist to the physician.

In fact, failures in information sharing are the primary reason for the majority of fumbled handoffs and medical errors – recent data indicates this could be as high as 70% (Joint Commission on Accreditation of Healthcare Organizations, 2004), with 25% of the errors occurring in the emergency room (Sucov, Shapiro, Jay, Suner, & Simon, 2001). By information sharing I mean both the explicit and implicit information exchanges of already acquired information towards the solving of a problem (Talja & Hansen, 2006). It is clear that accurate and efficient information sharing is required throughout the patient’s illness trajectory – both within a single shift and as when one shift hands off to another.

Due to the high rate of information sharing errors, hospitals are beginning to push for the integration of electronic patient records and other information systems in the hopes it will improve communication, and that this in turn will increase safety (i.e., reduce deaths) and efficiency (i.e., reduce costs). At the same time, many problems and issues have emerged with this move to the paperless hospital (Ash, Berg, & Coiera, 2004; Embi et al., 2004) – primarily due to a rash of unintended consequences from a lack of understanding how medical workers realistically coordinate and share information.

There have been many examples of unintended consequences that have emerged as the result of the implementation of electronic patient records. Ash et al. (2004) produced a summary of errors caused by the implementation of patient care information systems that were pulled from a set of qualitative studies conducted in the United States, the Netherlands, and Australia. As a result of this survey, they concluded that the “intended strengthening of one link in the chain of care actually leads unwittingly to a deletion or weakening of others” (p. 105). For instance, a qualitative study of the implementation of computerized physician documentation argued that the increase in documentation for efficiency and thoroughness might be detrimental to clinical practice and education due to increases in documentation that is not useful (Embi et al., 2004). Because of templates and the pervasive ability to cut-and-paste content, the patient charts overflowed with “clutter” and some of the physicians felt that this might hinder efficient patient care. This was especially a problem with less experienced residents who inappropriately used automated features like template generation of notes that in the end created documents that were useless. Faculty physicians also lamented that they believed the residents lost the opportunity to reflect on the patient’s data while crafting the proper input because they came to depend too much on automation. This caused the faculty physicians to distrust the work of residents because they questioned how much the residents thought through individual problems when they were easily able to cut and paste clinical assessments and treatment plans from another physician’s or nurse’s notes.

Another class of unintended consequences is the loss of information that is fundamentally emotional in nature. It has been shown that the feelings of the nurse or doctor towards the patient or the patient’s course of care is an important aspect of information transfer during handoffs (Evans, Pereira, & Parker, 2008; Kerr, 2002;

Lamond, 2000; Philpin, 2006). Unfortunately, after the move from a paper-based system to a computerized system, psychosocial information such as the patient’s emotional state or personality-related concerns can disappear from written records due to concern regarding the permanence and far-reaching distribution of the electronic documentation (Zhou, Ackerman, & Zheng, 2009).

The move to a more computerized hospital environment as well as the introduction of new information technologies in the healthcare field may improve some aspects of information sharing and coordination. Unfortunately, it will also inevitably overlook aspects of coordination that are less formal (and thus difficult to represent as part of a persistent and far-reaching digital record), but are just as important as the more objective elements of patient care. A central premise of this dissertation project is that the expression of a medical worker’s emotions towards the work is one of these informal aspects of coordination that may be lost or mishandled in the move toward increased technology support. It is important to gain an understanding of the role of work-related emotion expressions in medical work in order to better design information systems or else plan for alternative mechanisms and processes in order to ensure the support of effective coordination and information sharing.

1.2     Research Motivation

The study of emotion expressions in the emergency room clearly has important real world implications on medical information system design and implementation. In addition, my primary research goals relate to the more general question of how the computer supported cooperative work (CSCW) and human computer interaction (HCI) communities can better design information systems to support the expressions of emotion in a collaborative environment.

1.2.1     Advance the Perception of Emotion in Collaborative System Design

Emotion has rarely been regarded as an important component in the design of information systems to support collaborative activities. Although often analyzed as part of designing enjoyable or fun technology experiences, for the most part, researchers have not considered the collaboration opportunities or challenges raised by emotion expression in work settings. Even when emotion has been investigated in work activities, the emotional content is typically viewed as external to the work activity itself. For instance, Nardi (2005) examined the creation and maintenance of socio-emotional connections within work-based networks. She suggested that people work to keep a social network at hand so that they can activate communication activities when needed to facilitate work goals. In one of her examples, an interviewee told of sending an instant message to a coworker to “just say hi” in order to indicate affinity as part of maintaining a social connection. This affinity could then be called upon for future work-specific communication. Although Nardi’s work is a first step toward the study of emotion in a computer-mediated work setting, it focused on the role of emotions and feelings as external, lubricating factors rather than content integrated within work activities.

The lack of attention to emotion in collaborative system design is consistent with the general orientation of cognitive and organizational behavior researchers, who have tended to view emotion as an unimportant and an inappropriate aspect of work (Ashforth

& Humphrey, 1995; Putnam & Mumby, 1993). Traditional theories of cognition and decision making position emotion as the antithesis of rational thought. Indeed, decisionmaking that weighs emotion over reason has often been blamed for poor decisions. These theorists believe that emotion does not and should not play a part in real work. Instead emotion is viewed as a disturbance – something that employees must work around or address, but not actively engage in.

Despite this traditional yet still pervasive absence of emotion in formal models of work, a growing number of studies over the last two decades have documented important roles for emotion in work (for a review, see Grandey, 2008). Emerging research in cognitive psychology has shown that emotion is not orthogonal to cognition and, quite to the contrary, can actually enhance cognitive processes (Isen, Daubman, & Nowicki, 1987). Emotion is now being analyzed as an integral supporting element of the cognitive system instead of as a peripheral and detrimental causal factor. For instance, researchers have found that positive group affective tone can improve group processes by increasing cooperativeness and group performance (Barsade, 2002; George, 1990). Thus, this growing collection of empirical studies and theory can be seen as an “affective revolution” (Barsade, Brief, & Spataro, 2003) that has spread from the cognitive sciences to the social sciences in the study of work. This is highlighted by Fineman (2004) in Understanding Emotion at Work, who regarded the typical study of organizational life as bland – that which does not allow the complexities of practice to be revealed. He argued for the use of an emotional lens (i.e. viewing emotions as an important aspect of all work and interactions), for he considers emotion to be the primary medium through which people interact. But, due to the fundamentally social nature of work, the expression of emotions are also regulated by display rules – emotional requirements of the job that may be explicitly stated in training or implicitly stated through on-the-job socialization (Rafaeli & Sutton, 1987). This is especially important when a job depends on effective interaction with customers.

Overall, there has been a scarcity of studies on the expression of emotion in work with the intention to inform the design of collaborative systems. More often, studies of emotion in CSCW have focused on non-work related text chat and blogs (Gill, French, Gergle, & Oberlander, 2008; Hancock, Gee, Ciaccio, & Lin, 2008). However, those that have been conducted have shown the positive effect of emotions on work outcomes as well as the usefulness on enhanced emotion expressions in computer-mediated communication. For instance, in computer-mediated groups, the mean level of group happiness leads to more cohesiveness, which subsequently leads to greater participation by those in the group (Rhoades & O'Connor, 1995); subjective assessment of computermediated communication by group members for a group decision making task was improved when the use of emoticons (emotional icons) were introduced (Rivera, Cooke, & Bauhs, 1996); and group process feedback increases the number of socio-emotional interactive sequences in a computer-supported work environment (Losada, Sánchez, & Noble, 1990).

The nascent movement of emotion studies in work shows a growing recognition that emotion can play an important role in coordinated work. However, although emotion may be expressed in a workplace and may be an important aspect of the work itself, it is rarely included in formal models of work and thus has not played a prominent role in the design of information systems intended to support communication and coordination. This perception must be overcome. On the one hand, emotions may be providing critical information that will be lost as work environments come to rely more on information technology; on the other, emotion may be part of a rich social context for action that will be unable to operate as work interactions become increasingly technology-supported. Given the pervasive existence and impacts of emotion, it is not wise to simply ignore emotions or assume they will ‘find a suitable way’ to persist with the introduction of new collaborative work systems. As efforts expand, the discussion and investigation of emotion will become even more widespread and the implications for the design of information systems will begin to amass. Thus a high-level goal of the work reported here is to contribute to this awakening of interest in emotion in coordinated work and system design by articulating the emotion expression in support of emotion awareness.

1.2.2    Advance the Design Paradigms for Emotion in Human Computer Interaction

When considering emotion from the perspective of the human-computer interaction field, two prevailing schools of thought encompass the research and writings on the subject. The first began with Rosalind Picard, a MIT Media Lab professor and author of the seminal work, Affective Computing (1997). Picard, trained as an electrical engineer, began her career in the area of signal processing which she then applied to the detection of physiological and later, behavioral markers of emotion. Thus, her approach has focused attention on the reliable sensing of emotion. In addition to the hundreds of papers by Picard’s Affective Computing Research Group, this perspective on information system design and development has inspired a variety of research projects focused on sensing various instantiations of emotion during human computer interaction. Other work from this tradition has then been to translate the sensed emotions into displays and outputs that are reliably understood by another.

Years later, Donald Norman has helped to popularize the role of affect in user interface design through his book titled Emotional Design (2004). In the model he presents, human reaction to design exists on three levels: visceral (appearance), behavioral (how the item performs) and reflective (what the product evokes in relation to the self and previous experiences). However, Norman’s design focus is relatively narrow, aimed primarily at display and beauty in visual design. Although Norman’s book has spurred the field of human computer interaction to consider emotions more thoughtfully, it has left many designers and researchers with little more than a ‘pretty is more useable’ mantra. In fact, I would argue that this book was a disservice to those interested in affect in information system design, because it created the impression that affect need only be considered in the look-and-feel elements.

Much like the emotional design perspectives, researchers working in information science have attempted to raise positive emotions through the design of information systems. Recently, Diane Nahl and Dania Bilal (2007) published an edited book of work entitled Information and Emotion: The Emergent Affective Paradigm in Information

Behavior Research and Theory. Much of the work in this area, including the work by

Nahl and Bilal, has focused on the emotions felt during the information seeking process. In general, the field of information science tends to focus on understanding the emotions raised by an individual using an information system as opposed to collaborators.

More recently, a social, interactional perspective was developed in contrast to the approach of the previous, informational efforts. This view surfaced initially with the Affective Presence Group, a group of international researchers who are “interested in creating affective interactive systems or devices that trigger social and emotional engagement and reflection” (Höök, ND). Out of this group came the seminal treatise,

How Emotion is Made and Measured (Boehner, DePaula, Dourish, & Sengers, 2007). This paper directly challenged the affective computing and emotional design approaches by claiming that emotions are not individual states that occur in a vacuum but are rather developed in conjunction with their environment and through interactions with others. Thus, they termed the previous efforts as part of the ‘informational approach’ – i.e. they view emotion as an “internal, individual, and delineable phenomenon, which operates in concert with and in the context of traditional cognitive behavior” and thus they have “difficulty in accounting for and adequately incorporating an understanding of everyday action as situated in social and cultural contexts that give them meaning” (p. 275). In contrast, the ‘interactional approach’ asserts that emotion is more than “information units that are internally constructed and subsequently delivered” (p. 275). Thus, researchers in this field have developed systems and methods for measuring emotion that encourage reflection on and creation of emotional experiences. This perspective has inspired a whole new line of emotion research within the field of human-computer interaction.

All of these approaches indicate the growing importance and interest in emotion in information system design. However, all of the design perspectives begin with a focus on explicitly designing for as opposed to beginning with an understanding of socialemotional phenomenon and then determining the best course for designing with emotions in mind. Thus, the work reported here begins with an eye on how emotion is currently initiated, developed, and expressed during collaborative work tasks. This then puts the emphasis on the social and allows the technical to naturally evolve from real world findings. From this technical evolution a reflection on the two prevailing design perspectives (information processing and interactional) may enlighten how to negotiate between these two perspectives.

1.3     Research Questions

In order to address the aforementioned problem and research motivations, I have chosen to study emotion expressions in the information-intensive, tightly coordinated emergency room (ER) of a large urban academic hospital. The ER is a rich site for the preliminary study of emotion in collaboration because of the type of work that takes place in such a unit: it is a fast-paced environment where patient conditions can change rapidly and personnel must often make split-second decisions that can literally mean life or death. Emotion expressions are frequent and may influence decision-making; this means that the analysis of emotion is both feasible (sufficient frequency and variety of episodes) and of high impact. Therefore, to better understand work-related emotion expressions in an emergency room, I have conducted a study of medical personnel and their expressions of emotion in information sharing and coordination practices.

In service of my problem motivation – to understand the role of emotion expression in medical work in order to design effective collaborative medical information systems – I have sought to answer the following research questions:

RQ1: How does the expression of emotions support the coordinated work of an ER?

RQ2: In what form are emotions expressed in the ER?

These questions address the more practical aspects of this dissertation. The findings of this study aim to answer these questions in order to provide meaningful guidance to further medical information system design directions. For instance, the first question addresses why emotion expressions should continue to be supported with the introduction of information systems in the medical domain, whereas the second research question aims to specifically identify what emotion expressions could possibly disappear in the move to electronic medical documentation in an ER.

In addition to my problem motivation, I aim to address two research motivations. My first research motivation was to further the CSCW field’s understanding of the occurrence of emotion expressions in work environments as well as the relationship between the occurrence of emotion expressions and emotion awareness in work. To that end, I have specified the following research questions:

RQ3: How does the expression of emotion play a part in the development of emotion awareness in coordinated work?

RQ4: How does the social context shape the expression of emotion in coordinated work?

Because I will be investigating a real world collaborative context, I expect to go beyond the individual and lab studies documented in the existing research literature and uncover critical contextual factors associated with emotion expressions towards work goals. Knowing why emotion expressions have been integrated into work practices informs our understanding of the differing and evolving ways people share information to coordinate work. In addition, the direct expression of emotion is one such mechanism for expressing emotions in the ER. However, social and contextual factors in the work environment of the ER may modify the mechanisms of emotion expression. For instance, in contrast to the direct expression of emotion, some emotions may be surreptitiously expressed. I want to discover what factors may be important in how emotions are expressed – and perhaps not expressed – in the ER. This analysis can be used to determine what considerations must be made for the design of systems that enable emotion awareness.

My second research motivation was to further the affective computing field’s understanding of the social nature of emotion expressions in a work environment and, thus, to provide a counterbalance to the two prevailing interventionist-oriented design approaches. To that end, I address the following research question:

RQ5: How can we design information systems for the socially manifested emotion expressions of a coordinated work environment?

This might mean actively designing for emotions or it may mean being cognizant of the occurrence and functionality of emotions in order to allow for their continued appropriation. Thus, in addressing this research question, I will utilize what I have learned from answering the previous questions in considering the application of each design perspective to the development of emotion awareness technologies for an emergency room. In addition, I will address the unintended consequences that may arise from the introduction of new information technology into an effective socio-affective work environment.

1.4     Dissertation Overview

The remainder of the dissertation is organized in the following manner:

Chapter 2: Emotion Theories, the Workplace, and Design. This chapter provides an overview of the theories and models of emotion as well as an overview of related studies of emotion in the workplace. Following that is a more comprehensive discussion of the two prevailing perspectives on emotion in human-computer interaction that I previewed earlier.

Chapter 3: Study Design and Methodology. This chapter outlines the epistemological basis for the ethnographic approach used as well as provides details on the observation and interview methods.

Chapter 4: Research Setting and Information Flow. In this chapter, I present the context in which the data was collected. This includes the personnel who were observed and interviewed as well as the structure and media used to support emergency room activities.

Chapter 5: Emotion Expression in the Emergency Room. This chapter

describes the occurrence of work-related emotion expressions in the emergency room. It provides a description of the source of emotion as well as the outcome of the emotion expressions. I then discuss how the expression of emotion serves a critical role in the work performed in the emergency room.

Chapter 6: Moderators of Emotion Expression. I discuss in the second results chapter how various moderators influence the expression of emotion. Moderators take the form of either constraining emotion expressions or promoting them within the workplace.

Chapter 7: Expressions in Emotion Awareness. In this chapter I relate the findings from chapters 5 and 6 to the concept of awareness, articulation work, and common ground to form a framework for understanding emotion awareness and the expression of emotion in work.

Chapter 8: Design For and Around Emotion Expressions. In this chapter I consider the implications for collaborative information system design by situating the findings between the two prevailing perspectives on designing for emotion in HCI.

Chapter 9: Conclusion. This chapter revisits the research questions and presents the major contributions of the study along with plans for future work.

[1] GSW = Gun Shot Wound

[2] O.R. = Operating Room

EMOTION AWARENESS AND INVISIBILITY IN AN EMERGENCY ROOM:  A SOCIO-TECHNICAL DILEMMA

 

Sharing is caring!