Comparing Companion Involvement in Clinical Telephone and Face-To-Face Consultations About Seizures

ABSTRACT

Companions (i.e., friends, family members, and other accompanying persons) play an important role in seizure clinic consultations, providing information that patients cannot. The COVID-19 pandemic has seen an increase in these consultations being conducted via telephone. Little is known, however, about how this shift might affect companion participation. Using conversation analysis applied to a set of recorded telephone neurologist-patient-companion consultations (n = 9) and comparable moments drawn from a set of face-to-face consultations (n = 37) (both collected in the UK), we aimed to explore this impact and to identify communication methods that clinicians can use to manage companion participation during telephone consultations. We identified four ways in which participation was observably affected by the telephone. Telephone consultations could make it unclear whether a companion was present and make it difficult for the companion to communicate directly with the neurologist. Passing the floor from one speaker to another was more complex remotely, which could also restrict the patient’s own participation once the companion had the floor. These issues are rooted in the limitations of the telephone as a communication medium. Based on the issues identified, we conclude our analysis by highlighting some of the ways in which neurologists and other health professionals can manage companion participation in telephone consultations. These include encouraging the use of speakerphone, checking whether a companion is present throughout the call, keeping track of who can hear what throughout the call, and directing questions using given names to avoid ambiguity.

Introduction

It is common for friends, family, and others (hereafter referred to as companions) to accompany patients to medical consultations. Research has shown how companions provide emotional support (Andrades et al., 2013; Huber et al., 2015; Schilling et al., 2002), aid in decision-making (Clayman et al., 2005; Pino et al., 2021), provide information about the patient’s condition (Ekberg et al., 2014; Wolff & Roter, 2008), and question the doctor (Eggly et al., 2006). Although these roles seem positive, reviews indicate both upsides and downsides to companion presence (Bracher et al., 2020; Laidsaar-Powell et al., 2013; Wolff & Roter, 2011), with Laidsaar-Powell et al. (2013) noting, for example, concerns about companions dominating the interaction, taking up time, or making the patient reluctant to share sensitive information.

One setting where companions are particularly important is the seizure clinic because, as “witnesses” to the patient’s seizures, they can provide information that the patient cannot (Hadjikoutis & Smith, 2005); not talking to witnesses has been shown increase the risk of misdiagnoses in this setting (Smith et al., 1999). Robson et al. (2013) examined the impact that companions had on the duration and structure of initial consultations in a seizure clinic. They showed how, even though accompanied and unaccompanied consultations did not significantly differ in length, companions occupied a mean of 16% of the discourse space in accompanied consultations. This indicates that companions do not simply “add words” to the seizure clinic consultations but fundamentally alter the existing doctor-patient dynamic.

The COVID-19 pandemic has seen an increase in epilepsy consultations conducted via telephone and over the internet (teleneurology) (McGinley et al., 2020). Stommel and Stommel (2021) have shown that, in primary care, most companions in video-mediated consultations acted as “bystanders” to the interaction, remaining largely off-screen and non-participatory. Given how important companions are in seizure clinic consultations, it is important to understand how the shift to remote consultations might affect their participation in this context as well.

The aim of this study is to address this topic. Using a set of newly recorded telephone consultations, we will highlight four ways in which the inherent limitations of the telephone affect companion participation. We will also highlight some of the ways in which these limitations can be overcome. Throughout, we will compare the new data with previously captured recordings of face-to-face consultations.

Methods

The telephone data for this study were collected between June and November 2021 as part of a wider comparative project on differences between face-to-face and remote consultations. Patient participants due to have new or follow up appointments in specialist seizure clinics at the Royal Hallamshire Hospital in Sheffield, United Kingdom, were identified and sent letters ahead of their appointments informing them about the research. Just prior to starting the telephone appointment, participating neurologists asked potential participants if they were happy for the conversation to be recorded so it could, potentially, be used for this research. Patients were made aware that recordings would be deleted and not used for the research if no consent for their use was provided subsequently.

Consultations were recorded by placing a camera in the consultation room and recording the doctor as they called the patient on speakerphone. This meant that we only had audio for the patient’s side of the call.

After the encounter, patients were invited to provide full written consent either by post or using an online form; the same was true of anyone else who participated in their appointment. Ethical permission was obtained for the collection of these data from the Yorkshire & The Humber – Bradford Leeds Research Ethics Committee (21/YH/0086).

These recordings were compared with face-to-face consultations recorded for an earlier project (Toerien et al., 2013; Wiseman et al., 2016). These recordings were collected at the Southern General Hospital in Glasgow and the Royal Hallamshire Hospital in Sheffield in 2012.

The data were transcribed in detail using a simplified version of the Jefferson transcription conventions (Jefferson, 2003), with timed pauses and gaps, overlaps (indicated by square brackets), and inbreaths (e.g. “.hhh”) included.The data were analyzed using conversation analysis (CA), a microanalytic method used in the study of interaction in both everyday and institutional settings (Sidnell, 2010). As applied to medical interaction, CA has been used to understand, among other things, how treatments are recommended (Stivers, 2007), and how diagnoses are made (Peräkylä, 1998) (see Heritage and Maynard (2006) and Beach (2012) for collections of work on CA in medicine).

In this case, we used CA to identify the impact that telephone consultations had on companion participation. We searched the cases featuring companions and extracted moments where the telephone was made relevant (or potentially made relevant) in some way. These included moments where the telephone was overtly remarked upon or managed by interactants (e.g., moments where the phone was passed from one person to another) as well as subtler moments that could have been attributable to the impact of the telephone. We then analyzed these moments in depth, looking for commonalities and patterns in the way that the telephone was impacting the interaction. Comparable moments from the face-to-face consultations were used to aid in this analysis.

Results

Sample

In total, 39 patients provided consent to have their telephone consultation recorded. Of these encounters, 9 featured companions and were thus eligible for the present study. The 5 neurologists who were involved in these encounters also provided written informed consent. These were compared to consultations from a set of 37 recordings featuring companions (of 56 total) from the existing face-to-face data. In a forthcoming publication (Ford & Reuber, in preparation) we will show that face-to-face consultations were significantly more likely to feature a companion (66% to 26%, respectively).

Limitations of telephone consultations

This analysis is divided into two sub-sections. The first outlines four ways in which participation was observably affected by the telephone. The second discusses some of the ways in which interactions in the telephone consultations could, to some extent, overcome these limitations. In some cases, the extracts highlight practices that were common across multiple consultations (e.g., Extracts 3 and 5); in other cases (e.g., Extract 12), they highlight instances which, while representative of a wider limitation, were unique within the data.

Limitation 1: Uncertainty about companion presence

In face-to-face consultations, the presence or absence of a companion is easily determined. Even if they are not formally introduced, the medical professional can see that they are present and, where necessary, bring them into the interaction by addressing them. Due to the lack visual information, this is not necessarily the case in telephone consultations. Consider Extract 1, which comes from a consultation where the patient is having difficulty communicating due to her condition.

Extract 1. [Telephone/Consultation09/Neur_D/00m17s].

After several failed attempts at engaging her, the neurologist asks outright at line 12 if she is “by [her]self at the moment.” The uncertainty implied by this question highlights how, in the absence of shared physical space and any audible contributions from a third party, “it may not always be self-evident from the onset of the consultation whether a companion is present at all” (Stommel & Stommel, 2021, p. 184).

It transpires that there is indeed a companion present (the patient’s partner). The patient hands the phone to him, and he ends up acting as the primary speaker for the rest of the consultation. Yet because the companion’s presence was not determined at the beginning of the consultation, there is no way for the neurologist to know how aware he is of what was discussed before he came onto the phone. As can be seen in Extract 2, therefore, the neurologist must bring him up to speed. Again, this highlights an ambiguity about just how involved the patient’s partner had been in the consultation before he started to interact with the neurologist directly.

Extract 2. [Telephone/Consultation09/Neur_D/02m48s].

Limitation 2: Lack of direct interaction between companion and neurologist

By design, the telephone is better suited to one-on-one interactions. The impact of this in the consultations that we recorded can be seen in Extract 3.

Extract 3. [Telephone/Consultation26/Neur_F/04m59s].

At line 1, the neurologist asks the patient how long he has “been having these head pains for.” The patient audibly struggles to answer this question at line 3 (“Uh ”), which leads to the companion (his wife) stepping in to help at lines 4 and 7–8. However, because she is further away from the phone, her contributions are noticeably quieter than those of the patient. He therefore has to repeat them for the neurologist at lines 6 (“S’been on and off”) and 9 (“Just before I went on the duloxetine”).

For comparison, consider Extract 4, which comes from a similar moment in a face-to-face consultation.

Extract 4. [Face-to-face/G03304/04m22s].

Again, the doctor here has asked a question that the patient is audibly struggling with (lines 1–2), and again the companion has stepped in to help. Because of the shared presence, however, her response (line 3) is conveyed directly to the neurologist. This avoids the redundancy of the patient having to repeat what has already been said.

It also means that the accompanying person’s contributions fit more naturally into the ongoing flow of the interaction. Indeed, another characteristic of companions having their contributions mediated by the patient was that the patient could treat these contributions as something disruptive that needed to be accounted for. An example can be seen in Extract 5, which comes from a consultation with a patient and his wife who, like the patient’s wife in Extract 3, is sitting away from the phone.

Extract 5. [Telephone/Consultation19/Neur_B/14m03s].

At line 1, the neurologist asks the patient what causes his sickness. The patient begins to answer this question on his own but reaches a detail that he needs to consult his wife on. However, before consulting his wife at lines 10-11, he apologizes to the neurologist for doing so: “bear with me please for a second” (line 10). He also, as in Extract 4, repeats her contribution at lines 17–18.

Compare this to Extract 6, which comes from a face-to-face consultation.

Extract 6. [Face-to-face/S07603/06m30s].

As in Extract 5, the patient here recruits the companion to help him describe his condition. However, because all parties are in the same room, he does not need to account for this. Instead, he simply turns to her at lines 12-13 and invites her to contribute with a question: “also I have a bad week don’t I?”

Of course, there is no reason why the patient in the telephone consultation could not just turn to his wife to solicit her contribution; indeed, he may have done so. The point is that, due to the limitations of the telephone, the patient treated his wife’s contributions as a disruptive side activity rather than a natural, taken-for-granted part of the interaction.

Limitation 3: Difficult transitions between speakers

As noted in our introduction, there are topics on which companions are more qualified to speak than patients. This is especially true in neurology, where conditions often involve the patient losing consciousness or being otherwise unaware of what has happened. This meant that, in addition to the one-off contributions seen in the previous section, there were times in our data when companions had to contribute at length. However, as we have also seen, the telephone is inherently more suited to one-on-one interaction. Transitions between patient and companion could thus take time and be fraught with ambiguity.

An example of this can be seen in Extract 7, which comes from around 1 minute into a consultation with a patient and her father. The patient’s father has not contributed at all until this point, with the patient having answered the neurologist’s questions on her own. As the extract begins, however, the neurologist is asking about the patient’s recent seizures – a topic that the patient’s father, as a witness to these seizures, is better suited to speak on.

Extract 7. [Telephone/Consultation03/Neur_A/01m07s].

As we have already seen (e.g., Extract 6), patients in face-to-face consultations could give the floor to companions simply by turning to them and inviting them to contribute. Here, in contrast, the transition has to be verbally initiated and accounted for. First, the patient must announce that her father is going to speak (line 3); then she must hand over the phone (line 4); then her father must introduce himself (lines 7); then, finally, the consultation can proceed (lines 10-11).

The patient’s father remained the primary speaker for 5 minutes after this point, answering the neurologist’s questions on those topics that he was qualified to speak on. These included the nature of the patient’s seizures and whether the patient should be given clobazam while her seizures are in progress (he being the one who would have to administer it).

With these topics dealt with, the neurologist went on to ask about changing the patient’s daily medication, as can be seen in Extract 8: “the other question then was … whether we want to try the perampanel um … that was the switch from zonisamide to- to perampanel. That’s what I mentioned to you last time” (lines 2–4).

Extract 8. [Telephone/Consultation03/Neur_A/05m35s].

Unlike the preceding topics, this is one that the patient herself is better suited to speak on, given that she will take the medication and have to deal with any side effects. However, because the neurologist uses the pronouns “we” (“whether we want to try perampanel”) and “you” (“That’s what I mentioned to you last time”) in his question, it is unclear at first who the question is directed at.

In a face-to-face consultation, such a confusion would not exist because there would be other indications of who was being addressed. An example can be seen in Extract 9, where a neurologist addresses both a patient and a companion with the pronoun “you” by shifting his gaze between them in turn (lines 14–16).

Extract 9. [Face-to-face/S01001/01m30s].

In Extract 8, on the other hand, the patient’s father has to disclaim knowledge and authority on this topic and pass the floor back to the patient: “((Patient Name)) had better answer that” (lines 6-7). Just as the first transition between speakers had to be verbally initiated and accounted for, so too does the transition back to the patient.

Limitation 4: Patient participation limited

So far, we have seen examples where companions have played an information-providing role, taking the floor either briefly (e.g., Extracts 3 and 5) or at length (e.g., Extract 7) to speak on topics that the patient cannot. There were cases, however, where companions spoke on the patient’s behalf for most of the consultation because the patient was not able or found it difficult to do so.

It was not unusual for a companion to do most of the speaking even in a face-to-face consultation. However, whereas a patient who “steps back” in a face-to-face consultation is still physically present and party to the ongoing interaction, the telephone made it observably difficult to determine just how involved a patient was once the companion had taken the floor.

An illustration of this point can be seen in Extract 10, which comes from the same consultation as Extract 1. As we have already seen, the patient here has found it difficult to speak and her partner has taken over. However, because she is no longer contributing to the interaction and because of the lack of visual information, it is unclear if she is still present. This leads the neurologist, at lines 13–14, to ask the companion directly if the patient is still “at [his] side” and “listening in.”

Extract 10. [Telephone/Consultation09/Neur_D/03m23s].

Some of the points made in previous sections also applied when the patient was the one “off phone.” For example, we saw several examples above of patients repeating companions’ contributions back to the neurologist due to their low volume; in Extract 11, below, this is reversed, with the companion repeating the patient’s contribution back to the neurologist (line 8).

Extract 11. [Telephone/Consultation22/Neu_E/12m17s].

A patient’s attempted contributions after the companion had taken over could give an insight into how engaged they were with the ongoing interaction. Extract 12 comes from earlier in the same consultation as Extract 11, with a patient’s mother speaking on the former’s behalf.

Extract 12. [Telephone/Consultation22/Neu_E/05m18s].

As the extract begins, the neurologist is asking the patient’s mother about the duration of her absence seizures (lines 1–2, 4, and 6). The patient’s mother attempts to answer this question but expresses uncertainty at line 8 (“oh I don’t know”). The patient steps in at lines 9-10 to attempt to provide an answer. However, as her mother tells her at lines 12-13, they are talking about a different type of seizure, indicating that the patient has not been closely following the interaction. Of course, there might be other reasons for the patient’s misunderstanding here (including, as her mother suggests at lines 15–16, overall confusion), but being away from the phone could, based on what we have seen, have contributed.

The suggestion here is not that companions were deliberately excluding patients from the interaction, nor that neurologists were ignoring them; indeed, in the two cases that we have seen, the handover to the companion was initiated by the patient themselves. As in previous sub-sections, the issues that we have identified here are merely products of the telephone and the lack of shared space.

Addressing the limitations

In the previous section, we saw four ways in which the limitations of the telephone had an impact on companion contributions. Although addressing these problems fully would involve changes that go beyond the confines of the consultation itself, even within our recordings, we did observe some potential solutions. Consider Extract 13, which comes from early in a consultation with a patient and his partner.

Extract 13. [Telephone/Consultation07/Neu_B/00m05s].

At line 9, the neurologist tells the patient that she is going to “talk to [his] partner in a minute,” acknowledging the companion’s presence while precluding her participation at this stage. Around three minutes in, the neurologist indeed brings the patient’s partner into the consultation, as can be seen in Extract 14, line 14: “Could I ask your partner about them?” The companion then comes onto the phone and speaks to the neurologist for around 8 minutes, during which the patient himself does not speak at all.

Extract 14. [Telephone/Consultation07/Neu_B/03m12s].

Once she had finished speaking to his partner, the neurologist initiated the transfer back to the patient: “Do you want to put um your partner back on the phone?” The patient then spoke for the remainder of the consultation (17 minutes), with no further contribution from his partner.

In some ways, the neurologist in this extract has addressed the limitations outlined above. She has acknowledged the presence of the companion from the start, avoiding the uncertainty seen in Extract 1. She has also formalized the allocation of time to each speaker, avoiding indirect, accounted-for contributions (Extracts 3, 5, and 11), ambiguous questions (Extract 8), and impromptu handovers (Extracts 7 and 8).

It is noteworthy that the patient and companion contributed exclusively within the time that had been allocated to them, suggesting that this approach could have inhibited them making more spontaneous, collaborative contributions of the kind that often occurred in the face-to-face data (see Extract 15, below). This is speculative, however. What is clear is that the neurologist’s allocation of time in Extracts 13 and 14 has allowed for a more structured approach to companion contribution in a telephone consultation.

Extract 15. [Face-to-face/S07403/00m57s].

Another possible solution observed in the data was the use of speakerphone on the patient’s end of the call. Because this is more speculative and difficult to illustrate with data extracts, however, we will leave it for the discussion.

Discussion

The aim of this study was to explore how companions participate in seizure clinic consultations via the telephone as compared to face-to-face. Using recorded consultations, we have shown four ways in which the inherent limitations of the telephone can impact on companion participation: by making it ambiguous as to whether a companion was even present, by making it difficult for the companion to contribute without going via the patient, by making it difficult for the floor to be passed from one speaker to another, and by making the patient’s own participation ambiguous once the companion had the floor. General problems arising from these issues were redundancy, with information needing to be repeated for the party who may not have heard it (e.g., Extracts 3 and 5); uncertainty about who could hear what (e.g., Extracts 1 and 10); and misunderstandings about who questions were directed at (e.g., Extract 8).

We have also highlighted some of the ways in which these limitations could be overcome. In Extracts 13 and 14, for example, we saw a neurologist formally allocate time to the patient and companion early in the consultation, giving the consultation a clearer structure and avoiding some of the interactional problems seen in preceding extracts. We also noted briefly the potential benefits of speakerphone; certainly, there were rare cases where patients and companions contributed in a collaborative way more akin to a face-to-face consultation (see Extract 15) and these cases were characterized by both parties speaking at a roughly equal volume. However, we can only speculate as to whether speakerphone was being used in these cases because, again, we did not have video of the patient’s end of the call.

Yet such solutions are ameliorative at best. The absence of shared physical space and visual information means that many of the taken-for-granted fundamentals of face-to-face consultations (e.g., a shared understanding of who is present from the start of the consultation) are simply missing over the phone, and additional work will always be required to make up for their absence (see below). And while the speakerphone could clearly help, we would be reluctant to propose it is a catch-all solution. In Extract 7, for example, the patient declares that they are using speakerphone, yet we still see the long-winded transition between her and her father. Other extracts suggest that speakerphone might be being used (e.g., Extract 12, where the patient attempts to contribute) but, again, similar issues arise.

As for whether video consultations might be an improvement, we would point to Stommel and Stommel (2021), who suggest that the low levels of companion participation that they observed in video-mediated primary care consultations “are related to the dominant set-up of the talking head position in video-interaction” (p. 197). As they note, “it could be a medium factor that makes extensive participation of the companion less likely and more problematic in video-consultations” (p. 197, emphasis added). It appears that many of the issues we have identified with telephone consultations can also be present in video consultations.

At the beginning of the results section, we mentioned findings from a forthcoming article showing that the face-to-face consultations in our dataset were significantly more likely to feature a companion (Ford & Reuber, in preparation). Although this finding must be taken with caution due to the disparities between the datasets, it is noteworthy that it is supported and complemented by the present study; if companion participation is indeed limited by the telephone, it follows that they would be less inclined to be present during a call. However, there are other possible explanations; it might be, for example, that some patients appreciate the privacy of a telephone call or that companions were more present in face-to-face consultations simply because they had driven the patient to the appointment (seizure disorders often prevent patients from being able to drive themselves). Regardless of the reason, given that companions can provide important information in seizure clinic interactions (Robson et al., 2016), their lower levels of involvement in telephone consultations could be consequential.

Based on our analysis, we can make five recommendations for neurologists and others conducting telephone consultations where a companion is present.

1. Establish early in the consultation if there is a companion (see also Stommel & Stommel, 2021).

2. Encourage the use of speakerphone where possible.

3. Ensure that both the patient and the companion are audible and at roughly the same volume.

4. If the patient asks the companion to speak on their behalf, ensure that the patient can still hear the ongoing interaction and contribute where necessary.

5. Direct questions using given names to avoid ambiguity.

While such steps would not fully overcome the limitations of the telephone as a communication medium, they would avoid at least some of the interactional problems around companion involvement that we have outlined in this study.

A key weakness of this study is, again, the lack of video on the patient side of the encounter. While the differences that we have identified are striking, further research will be needed to determine their prevalence and comprehensiveness. There are also the disparities between the face-to-face and telephone datasets, especially in when they were collected, the neurologists featured in them, and the number of consultations featuring companions. While the issues that we have identified did seem to be linked to the communication modality, then, we cannot rule out the influence of these other factors.

Acknowledgements

The authors gratefully acknowledge the funders of this project: Epilepsy Research UK and Epilepsy Action. They are grateful for Dr Merran Toerien for her help with the access to data previously collected in the context of a project funded by the UK National Institute for Health Research Health Services and Delivery Research Programme (NIHR HS&DR project number 10/2000/61).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The work was supported by the Epilepsy Action Epilepsy Research UK through an Innovations in Healthcare Urgent Research grant.