6Communication in bedside nursing handovers

Suzanne Eggins and Diana Slade

6.1Introduction

The previous chapters have described the challenges of emergency department clinical handovers when medical staff change. In this chapter the focus shifts to nursing staff shift changes in a general medical ward. We describe and analyze the practice of bedside handovers in a metropolitan teaching hospital, hospital B, acknowledging the challenges but also the benefits of this semi-public clinical handover mode for nurses and patients. In chapter 7 we then suggest strategies and resources to improve patient safety and to increase nurse and patient satisfaction with the practice.

6.2Setting the scene

It’s 1 pm in a ward at a busy Australian public hospital. The nurses who have worked the morning shift from 7 am will soon be going home. But first they must hand over care of their patients to the incoming afternoon shift. One of the two outgoing shift nurses, let’s call her Sarah, leads the incoming team into the four-bed room she has team-nursed with a second outgoing nurse (Louisa) who follows the handover team into the room. Sarah stands in the middle of the room and unfolds a crumpled ward sheet on which she has handwritten notes. The four incoming nurses include a senior nurse, Jenny, who is the team leader for the afternoon shift, and Jude, who will be taking over the patients in this room. The handover team stands in a semi-circle around Sarah, their backs to the three elderly patients (one with a visiting relative) sitting beside their bedside. As Sarah talks the incoming nurses add quick notes to their ward sheets.

The three-minute handover below was not unusual for shift handovers we recorded on site. But as we observed and later transcribed and analyzed it we were struck by the multiple communication problems it exposed. We noted that the patient was only peripherally involved in the handover and that the outgoing nurse seemed to struggle to present a concise, logically organized explanation of the patient’s care. We also noted the emergence of multiple discrepancies in information about the patient, which appeared to mean that potentially serious tasks had not yet been done or may have been done inappropriately. We return to the problems we see with this handover immediately after the transcript.

Text 6.1: ‘We haven’t been handed that over’6

As they interact to achieve this handover the nursing staff expose and confront two overt problems. The first is the fact that Sarah has not yet clarified an important discrepancy in the patient’s medication – should she be taking the disprin as charted or not? The second is the confusion over chemo precautions, which reveals possibly inappropriate management of the patient’s care. But as well as these explicitly discussed problems there are several implicit practices that we suggest should be questioned. Here we note that the nurses went ahead with the handover without waiting for the patient to be present. This reflects the assumption that the patient is not relevant to the handover. Louisa tries to manage the confusion over the chemo precautions by asking the patient directly, but note that this only happens as an aside to the main handover. The handover group as a whole does not take advantage of the patient’s return to her bed to involve her in the handover in any way, even though Louisa describes the patient as ‘pretty on to it’.

We also suggest the handover reveals an underlying confusion over accountability and responsibility for this patient’s care. Louisa seems to think that Sarah – the more junior nurse – is responsible for following up on the disprin question, and neither Louisa nor Sarah believe they are responsible for the confusion over the chemo precautions because ‘we haven’t been handed that over’, although Jude contradicts that. More worrying is the fact that at the end of this handover it is still not entirely clear who is taking responsibility and when for resolving the discrepancies over medication and precautions. The disprin issue is, presumably, still left with Sarah, but she is about to go home. On the precautions Jenny says ‘what we’ll do is … check with the RMO’, but who exactly does she mean will do this?

A final and underlying problem we note is that the information is handed over in an apparently unstructured sequence, without a clear indication of the different components that the nurse giving the handover will cover and in what order. This relatively ad hoc presentation means Jude, Louisa and Jenny all need to interpose questions to clarify or prompt Sarah. This creates a risky dynamic as each piece of information is not necessarily resolved before discussion moves on to another.

We now move from this specific example to the broader context, questions and problems of our research and training in effective bedside handover communication.

6.3Background: research questions and data

As earlier chapters have argued, clinical handover is one of the pivotal communicative genres in medical and nursing hospital practice. Research into handover identifies poor communication in handover as a major risk (Haig et al. 2006). Problems include failure to hand over all relevant content (medications, test results etc), the lack of structure, system and relevance in handover information, excessive reliance on memory without reference to written documentation (Wilson 2011), and ‘failure-prone communication processes’ (Arora et al. 2005: 11), such as the lack of face-to-face discussion between clinicians and doing handover away from patients and family (Wilson 2011).

Hand-in-hand with patient safety concerns health care is undergoing a cultural shift from ‘a passive view of patients as undiscriminating recipients of care defined by others (‘patient centerd’ or not)’ to an emerging view of patients as active partners and ‘co-producers’ in their health care (Iedema et al. 2008: 105). This cultural shift is supported by empirical evidence that suggests patient involvement improves clinical outcomes (Haynes et al. 1996; Kravitz & Melnikow 2001; Wong et al. 2008).

These parallel international pressures of safety concerns and the cultural shift towards inclusive, patient-centered care are reflected in the Australian Commission on Safety and Quality in Health Care National Standards (ACSQHC 2012a). Standard 6, the Australian clinical handover standard, was developed ‘to ensure there is timely, relevant and structured clinical handover that supports safe patient care’ (ACSQHC 2012b: 7). As explained in chapter 1, the clinical handover standard promotes the principle of ‘flexible standardization’ with an emphasis on involving patients as active participants:

Until recently, in Australian public hospitals nursing shift handovers were often performed in staff-only areas away from patients. In some contexts the shift handovers were verbal face-to-face interactions between the outgoing and incoming nurses or teams held in the staff tea room. In other contexts outgoing nurses taped their handovers for incoming shift nurses to listen to when they arrived. However, in response to international and national recommendations, including those of the 2008 inquiry into acute care services in one state in Australia (Garling 2008), an increasing number of public hospitals are now requiring that nursing shift handovers take place at the bedside. Bedside handover is recommended in the Australian clinical handover standard as a means of including patients. The Australian clinical handover standard also recommends that clinicians use standardized handover tools such as iSBAR to ensure the handovers cover the minimum dataset. The principle of flexible standardization means that each context (hospital, ward, team etc) is encouraged to tailor its bedside handovers to its local context without compromising the principles of systematic information transfer and patient inclusion.

In response to the Australian clinical handover standard hospital B, where we carried out our research, mandated bedside handover for shift-change nursing handovers from January 2012. Nurses did not receive any training in how to perform bedside handover even though for most nurses it was a new communicative practice that raised considerable challenges, discussed further below. iSBAR had already been promoted by the local health department as the recommended hospital-wide handover protocol from March 2012.

In this chapter we explore the question: How effectively are nurses managing bedside handover? We ask whether their bedside handovers are meeting or not meeting the Australian clinical handover standard’s criteria of greater patient involvement and the systematic transfer of a minimum dataset. We will suggest that examples like Text 6.1 and others from our data reveal safety and quality issues, with nurses struggling to know how to include patients and a lack of clarity around just what information to hand over and how to organize it.

In the next chapter we ask: How can we support nurses to hand over more effectively at the bedside? There we present communication tools that arise directly from our research.

This chapter draws on our analysis of handovers of patients across shift change handover rounds in two wards at hospital B in Australia – Wards M (general medical) and Ward W (acute geriatric). As part of the Effective Clinical Communication in Handover (ECCHo) project we audio-recorded and in some cases video-recorded more than 200 minutes of bedside nursing shift handovers, comprising handovers of more than 80 patients. We observed many more handovers and we interviewed and spoke informally with many nurses and managers across the three years of the ECCHo project. All participants gave informed consent and all transcripts have been de-identified, in accordance with guidelines approved by the human research ethics committees of the hospital and our university.

6.4Summary of interactional and information issues in bedside handovers

We approach our bedside handover data as discourse analysts, drawing on social– functional theories of language, including systemic functional linguistics (Halliday & Matthiessen 2004; Eggins 2004), ethnomethodology and conversation analysis (Garfinkel 1967; Schegloff 1981; Heritage & Clayman 2010), critical linguistics (Kress 1985) and critical discourse analysis (Fairclough 1995). We understand language as a systematic resource that is organized so that we can use it to express simultaneously two types of meanings. The first meanings we express are interpersonal or interactional meanings. These are about roles and relationships: how we’re talking to each other, including how we feel about each other. The second meanings we express are ideational or informational meanings. These are about content, actions and actors: what we’re talking about. We cannot avoid expressing these two types of meaning simultaneously – language is organized so that we are always indicating how we are relating to each other at the same time as we are always talking about something.

These two types of meanings in all communicative events map very closely onto the Australian clinical handover standard’s identification of two main requirements for effective clinical handover: patient inclusion (interactional meanings) and minimum datasets (informational meanings). In analyzing our bedside handover data we therefore found it both theoretically coherent and practically useful to differentiate between these two communicative dimensions:

Here now is a summary of the issues we identified with the bedside handovers we observed and transcribed, organized in terms of our two analytical dimensions. Following the summary we present and discuss examples from our data.

6.4.1Interactional issues

6.4.2Informational issues

We stress that while these problems showed up in the majority of bedside handovers we observed, there were of course exceptions. Some nurses demonstrated exemplary practice. These exceptions only served to highlight our first general finding, which was the variability between handovers. We were struck by the fact that even within a single ward with an apparently shared understanding of bedside handover protocols, patient handovers differed markedly according to which nurse was handing over. Different wards at the same hospital showed even greater variability. In Eggins & Slade (2016) we attempt to capture this variability by identifying different contrasting handover ‘styles’. However, in this chapter we concentrate on linking the communicative issues we observed to the Australian clinical handover standard’s clinical handover criteria and to the training we subsequently developed. In the next section we therefore discuss and exemplify first the interactional and then the information issues summarized above. In the following chapter we show how we designed a training intervention to better equip nurses to manage the challenges of bedside handover practice.

6.5Interactional issues in bedside handovers

In this chapter, and in our training sessions, we deliberately begin with the interactional dimension of handover. This is because we regard interaction as the missing dimension in handover research and discussion, where the focus has almost exclusively been on the information that clinicians should hand over. By definition clinical handover has never been and can never be a monologic achievement. Much of its complexity and risk lies in its interactivity – in the fact that handover can only be achieved through dialogic negotiation with others, ideally with patients as well as with other clinicians.

The critical interactional issue in our bedside handover data was, to put it simply, that almost no interaction with patients was taking place. In the minority of cases where it did happen, it was often inadequate. A secondary issue was that in one ward we studied almost no interaction occurred between the nurses during handover – the incoming nurses did not query, check or confirm the information they were being told, and the outgoing nurse did not check her handover had been understood.

Table 6.1 gives a snapshot of interaction in 57 patient handovers in Ward M, the first ward we studied. Table 6.1 shows that for 14 patients the incoming group did not go near the patients’ bedside but took the handover in the corridor, citing as reasons infection control or the fact that the patient had visitors or was sleeping. In a further eight cases, patients were not present at handover time (they were either having a cigarette downstairs or were off having tests), meaning that in 39 percent of cases, the handover was not in the presence of the patient. The statistics in Ward W, the second ward we studied, were even more striking: in one bedside round of 11 patients, the handover was not at the bedside for 9 of the patients. Text 6.1 (page 133 above) is an example of a handover that took place without the patient’s presence. Note that if the handover team had modified their schedule slightly, handing over another patient who was present first, then the patient they discuss in Text 6.1 would have returned from the toilet and could have been included in the handover. This option was not suggested by any team member.

Even when patients were present during handover, and when the handover group stood within auditory range, patients were rarely invited to contribute. In Ward M, 21 patients (37 percent) were not invited to contribute even though the handover took place in reasonable proximity to their bedside. In Text 6.1 from Ward W we see that even when the patient returns and is in direct proximity to the handover group, she is not invited to contribute.

Tab. 6.1: Patient inclusion and exclusion in bedside handover events

However, patients are persistently human – they do like to communicate. In Ward M’s sample of 57 handovers, 8 patients (14 percent) contributed even though they had not been invited to do so. But when patients chimed in nurses often did not acknowledge or respond to their contributions. Text 6.2 from Ward M is an example of this lack of response by the outgoing nurse.

Text 6.2: ‘Doris’

‘Alice’, the outgoing nurse, is handing over ‘Doris’ to four incoming nurses. Doris is lying on top of her bed and is alert and attentive.

Doris’s humorous remark in turn 2 is typical of unsought patient contributions. Eggins (2014) suggests hospital patients often use humor as a strategy to reassert their individuality in a context that threatens to take away their agency and autonomy. For Doris, though, it does not work – Alice ploughs on with the handover as if she had not spoken. She does not acknowledge Doris’s comment in turn 5 either.

This pattern – where nurses do not appear to see or hear patients – shows how acculturated clinicians can become to behaviors that they would in other contexts find unacceptably rude and inconsiderate, especially if others were talking about them.

One unfortunate correlate of failing to acknowledge and include patients is the tendency to talk about them in ways that lack sensitivity and diminish their agency and autonomy. One simple expression of this is the insistent use of ‘she’ or ‘he’ to refer to a patient is who present. Notice how in Text 6.1 even when the patient is walking right past them the nurses continue to talk about her as ‘she’ (‘she said …’ she’s written down for it; she needs the chair’). In Text 6.2 Alice uses Doris’s name only in the opening sentence. After that Doris is either ‘she’ (‘she’s afebrile; she’s on oral ABs; she’s for a Geris review’) or she is rendered completely invisible by the ellipsis of even the pronoun (‘Ah mobilizes with a wheelie walker; Set up assistance in the shower; Fluid balance chart.’).

As we see in Text 6.1, nurses who are not interacting directly with the patient also tend to describe patients as the passive recipient of others’ actions (‘I’ve started her on a fluid balance chart; her meds have been given; she’s had a wash’) or as body parts and processes (‘she’s scoring a MUSE of 2’; ‘her BP dropped’; ‘she has loose bowels’; ‘she’s had them open twice’). Similarly, in Text 6.2 Alice describes Doris as passive (‘she for a Geri’s review’) or as a set of physical attributes (QID obs are good; she’s afebrile). Neither outgoing nurse talks about their patients doing or feeling things themselves.

So how could nurses hand over differently? Text 6.3 is an example of a Ward M nurse directly inviting, and welcoming, the patient’s contribution.

Text 6.3: ‘Harry’

‘Jen’, is the outgoing nurse handing over ‘Harry’. ‘Jimmy’ is an incoming nurse who joins the round a bit late. ‘All’ means the four incoming nurses receiving the handover. Ward M’s medical registrar (‘Ted’) just happens to be nearby at the time of the handover.

We were struck by how this handover became a collaboratively constructed account, characterized by good humor, warmth and mutual respect between the patient and the nurse. The nurse talked to Harry, not about him, using the pronoun ‘you’, and sometimes ‘we’ to express the collaborative relationship. She described the patient as active and agentive in his own care, making him the doer of positive actions (‘Harry’s managed to sit out of bed after having a shower; he’s eating and drinking well’), rather than a set of attributes or someone to whom things are done.

But this inclusive style of bedside handover was the minority. In Ward M only 14 (25 percent) of patients were directly invited to contribute to the handover, with outgoing nurses using an expression that directly invited participation, such as ‘Anything you’d like to add, Stan?’ or ‘Have I missed anything, Maria?’ We did not consider that a simple announcement ‘This is the afternoon staff. We’ve come to do handover’ amounted to an invitation to participate.

Most striking of all, perhaps, is our finding that one hundred percent of the patients who were directly invited to contribute did contribute. Sometimes patients made only a minimal comment, often light-hearted and self-deprecating; at other times they offered information about their background or current state. But the 100 percent response rate clearly demonstrates the unsurprising fact that patients are eager to be spoken to, not just about, and welcome the chance to contribute to the discussion of their care.

Based on our observations and interviews, we interpreted the reluctance of nurses to hand over at the bedside and to invite patient inclusion as indicating both institutional and skill factors. Firstly, it suggested to us an institutional hesitation over the redefinition of nurse–patient roles. A serious commitment to patient inclusion might lead to wards displaying written notices advising patients of the time of the bedside round and encouraging them to be present to contribute – a step that Ward M’s nurse manager instituted in light of our findings. It could also mean handover teams varying their handover sequence to allow patients temporarily out of the room to return to their bedside, as could have happened in Text 6.1. Hospitals might also review the rationales and benefits of bedside handovers in nurse professional development sessions. As described later, we built such a review of rationales into our bedside handover training.

Secondly, interviews, observations and recorded data indicated that many nurses lacked the confidence and communication skills needed for the new interactional context of bedside handover. Many were not sure how to engage patients and how to react to patient contributions (particularly humor). Nurses for whom English is a second language sometimes did not appear to understand what patients said, although a non-English speaking background did not automatically correlate with a nurse’s exclusion of patients from the handover. We judged that a practical, skillsbased workshop on bedside handover communication could give nurses the confidence they needed to invite patients into the handover interaction. We describe the design of our workshop in chapter 7.

6.6Informational issues in bedside handovers

6.6.1Structure and protocols

Bedside nursing handover was a new practice at hospital B, and so offered new interactional challenges for staff in both the wards we studied. However, when we turn to the informational side of handovers, the two wards had very different local experiences. In 2011 Ward M had been involved in a one-year pilot study on the implementation of iSBAR, a handover protocol designed by and for medical practitioners rather than nurses. Table 6.2 below summarizes the iSBAR components as they are traditionally explained, although in the next chapter we show how we modified this to better suit the nursing context.

Tab. 6.2: The iSBAR communication protocol for clinical handover

McGregor et al. (2011) studied how the Ward M nurses adapted and incorporated iSBAR into their tea-room shift-change handovers. In fact, the nurses had redesigned their ward sheets to reflect the iSBAR stages. So several Ward M nurses had already been observed and audio recorded and all were aware that handovers should be systematic and organized, preferably using iSBAR.

Ward W, on the other hand, had no previous experience as a research site. iSBAR was not ward policy, was not reflected on the ward sheets and most nurses were not familiar with it at all. There was no agreed protocol or sequence for handover. From formal and informal discussions with staff it seemed to us that clinical handover was not an issue on Ward W’s radar.

The lack of exposure to iSBAR or a similar systematic protocol is evident in the way Sarah gives and the other nurses receive the handover in Text 6.1. As Tab. 6.3 shows, several iSBAR stages are not covered (Situation and Background) or are mentioned only minimally (Introduce, Recommendations). Ward W nurses appear to assume that they can rely on the abbreviated notes on the ward sheet. But the ward sheets in use when we recorded Text 6.1 had only one column that listed all the following, with items separated only by commas: current presentation, relevant past medical history and MRSA7 status. A nurse new to the ward could not tell from the ward sheet or the verbal handover which of the listed conditions were the presenting ones and which the background ones. And as the issue of chemo precautions in Text 6.1 shows, information on the ward sheet was sometimes confusing or perhaps wrong.

Tab. 6.3: The iSBAR communication stages in Text 6.1

Non-use of a handover protocol or checklist does not in itself mean that a handover will not be organized and systematic. We have observed many impressively concise, comprehensive and well-structured handovers by clinicians who do not follow conventional protocols (Eggins & Slade 2012). However, without a framework, clinicians giving handovers are at risk of forgetting to pass on essential content and/ or of failing to develop a coherent account of a patient’s situation, rather than listing disparate facts. Those receiving handovers when there is no shared framework are never quite sure what to expect. As they do not know whether or when particular information will be provided they may be more inclined to interrupt the giver to ask for information, with the risk that the giver’s train of thought is disturbed and information is left unresolved or is skipped. We believe Text 6.1 shows these problems.

6.6.2Redefining the ‘minimum dataset’

Although our data and observations suggest that a protocol like iSBAR can be useful in providing an overall structure for handover, we suggest that to make the informational exchange effective we need to redefine current understandings of a ‘minimum dataset’. The term ‘minimum dataset’ suggests that the only informational issue in handover is whether certain information is present or not. An effective handover would be one where all the minimum dataset items had been mentioned. But our research suggests that information presented as an unordered list of facts is less effective than information presented as a cumulative and causal explanation of patient care.

When we look at the components of information covered in the Assessment stage of the handover in Text 6.1 (Tab. 6.3 above), we see that there is no particular reason for the sequencing of the chunks of information. They are offered as a list of separate bits of information that could just as well have been delivered in a different sequence. They do not amount to an explanation of the patient’s care; only to an enumeration of aspects of it from the nurse’s point of view. The same is true of the information in Text 6.2: we are told a list of facts about Doris but not how those bits of information are connected or what motivates them. When incoming staff are not told why tasks have been or need to be done, we believe the rationale that should guide continuity of care is missing.

By contrast, in Text 6.3 Jen offers not just a list of facts about Harry but a collaboratively constructed explanation of his behavior and situation. She makes causal links between Harry’s care plan and his rational, agentive response to his situation: he can’t eat a full diet because of the pain; he’s got to be pushed to have pain relief because he doesn’t want to be a zombie; he’s a bit dizzy because there’s been so much going on around him. This explanatory type of handover not only restores agency and respect to the patient but also demonstrates a fundamental principle of nursing care: that all nursing tasks should be performed for a reason. By making explicit the causal relationships between Harry’s situation and the tasks that nursing staff need to do with and for him, Jen makes it clear to the incoming team why and how to achieve Harry’s continuity of care.

6.6.3Unstated assumptions: responsibility and accountability

The final informational issue we will discuss is that in most of the bedside handovers we observed there was no overt transfer of responsibility and accountability for the patient’s ongoing care, even though this transfer is the defining rationale for handover. In our data, either future tasks are left implicit or the tasks are mentioned but the person who is to do them and the timeframe are not clearly stated. For example, in Text 6.1 it is not clear who is going to find out about the disprin and ask the medical officer about the chemo precautions – Jenny’s use of ‘we’ in ‘we’re just going to check with the RMO’ is ambiguous. In Text 6.2, Alice does not make a single statement that indicates what the incoming staff should do or watch for to care for Doris. It is implied that they need to provide stand-by assistance when she moves around and perhaps that they should follow up to find out the outcome of the Geris review. But nothing is stated explicitly.

Even in Text 6.3 future care tasks are implied rather than stated. Jen implies the incoming team needs to follow up with Ted about the variable dose and offer pain relief to Harry. They also need to offer him fluids. Her explanatory formulation of the handover makes it easier to see what the future tasks must be, but even so we wonder at the lack of explicitness. In handovers in other high risk domains, such as aviation and nuclear power, those receiving the handover must give a readback to confirm they have understood their tasks. In our training we encouraged nurses to do this and our colleagues in other jurisdictions have added an interpretation of the R in iSBAR as readback (see chapter 9).

6.7Conclusion: improving quality and safety in bedside handover

In this chapter we have drawn on qualitative research that studies how clinicians and patients interact in real contexts to help us understand current nursing handover practice. We have shown how naturally occurring hospital events offers insight into the challenges some nurses may face in managing the new communicative practice of bedside handover. We have suggested that for bedside nursing handover to meet the safety and quality goals, such as those set out in the Australian clinical handover standard, nurses need to be supported to manage more effectively the interactional and informational demands of the practice.

In follow-up training we offered to the hospital we incorporated two new communication protocols, directly derived from our analysis of actual handovers. To support nurses in the interactional dimensions of handover we developed the CARE protocol, where the keywords Connect–Ask–Respond–Empathize encapsulate the communicative skills both outgoing and incoming nurses can use to include the patient meaningfully in the handover. To improve the informational dimension of handover, we first reinforced the use of iSBAR to organize information on the worksheet. Second, we introduced the sequence Past–Present–Future as a way for nurses to sequence their verbal handover and to ensure they clearly handed over responsibility for specified future tasks. Chapter 7 briefly explains and exemplifies these protocols. Slade et al. (under review) describes in more detail the training and the highly positive qualitative and quantitative evaluations of its outcomes. In the next chapter we present resources nurses can use to improve their communication in bedside nursing handovers.