Health Communication

ISSN: 1041-0236 (Print) 1532-7027 (Online) Journal homepage: http://www.tandfonline.com/loi/hhth20

It Does Take Two to Tango: An Applied

Conversation Analysis of Interactions between
a Psychiatrist and Service-Users Discussing
Medication

Emma Kaminskiy & Mick Finlay

To cite this article: Emma Kaminskiy & Mick Finlay (2018): It Does Take Two to Tango: An
Applied Conversation Analysis of Interactions between a Psychiatrist and Service-Users Discussing
Medication, Health Communication, DOI: 10.1080/10410236.2018.1517633

To link to this article: https://doi.org/10.1080/10410236.2018.1517633

Published online: 12 Sep 2018.

Submit your article to this journal

Article views: 28

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=hhth20
HEALTH COMMUNICATION
https://doi.org/10.1080/10410236.2018.1517633

It Does Take Two to Tango: An Applied Conversation Analysis of Interactions

between a Psychiatrist and Service-Users Discussing Medication
Emma Kaminskiy and Mick Finlay
Department of Psychology, Anglia Ruskin University

ABSTRACT
Decisions concerning psychiatric medication are complex and often involve a protracted process of trial
and error. We examine three recorded meetings for power-sharing and power-taking discourse strate-
gies employed by both the psychiatrist and mental health service-user, when discussing psychiatric
medication. We identify examples of good practice, as well as missed opportunities to engage service-
users in co-constructed dialogue, and highlight that participation and active involvement in decisions is
not best seen as a fixed pattern, but is a complex interplay that changes both between and within
interactions.

Introduction highly contested (Moncrieff, 2009); and (c) the growing

awareness of long-lasting adverse effects associated with
Shared decision-making (SDM) is recommended in policy
long-term psychiatric medication use has provided more
and has been proposed to offer specific benefits to medication
doubt over the ‘cost–benefit’ ratio for psychiatric medica-
management practices in mental health services (National
tion use.
Institute of Clinical Excellence, 2009). Emerging from other
Research investigating communicative practices in mental
health-care arenas and the wider consumerisation of health-
health settings has identified different discursive strategies
care services, it has also been embraced in the transformation
employed by mental health practitioners, which can shut
of mental health service culture and practices. Here, SDM is
down or enable possibilities for SDM and service-user invol-
associated with increased emphasis on psychological and
vement. For example, research has shown how psychiatrists
social interventions, valuing service-users as experts by experi-
control topics that are open for discussion (McCabe, Heath,
ence and the promotion of recovery, social inclusion, and
Burns, & Priebe, 2002; McCabe, Khanom, Bailey, & Priebe,
empowerment (Deegan & Drake, 2006). SDM sits in the
2013) and often use closed and negatively framed questions
middle of a spectrum of patient participation, between patern-
thereby controlling how information on sensitive topics is
alism on the one end (derived from the biomedical paradigm
disclosed by service-users (McCabe, Sterno, Priebe, Barnes,
of health where the doctor makes the decision after evaluation
& Byng, 2017). In conversations concerning medication, psy-
of the diagnosis, treatment options, and risks and benefits)
chiatrists have been found to directly employ pressure and
and the informed model of decision-making on the other
persuasion in encounters with service-users, and often resist
(which assumes that the service-user has ultimate responsi-
engaging in service-users’ concerns about adverse effects
bility for the decision made and the professional acts as an
(Quirk, Chaplin, Lelliott, & Seale, 2012; Seale, Chaplin,
adviser). Within SDM, priority is placed on communication
Lelliott, & Quirk, 2007). Other research in a variety of psy-
and shared deliberation with a respect for service-users’
chiatric and non-psychiatric health-care settings finds that
experiential knowledge, values, and preferences (Edwards &
treatment proposals distinguish between pronouncements,
Elwyn, 2009). As such, the defining characteristic of SDM is
suggestions, proposals, offers, and assertions, and their use
its interactional nature. SDM presents specific opportunities
differs depending on the clinician’s perception of clinical risk,
for medication management practice in psychiatry where
the likelihood of acceptance/resistance, and the perceived
decisions are made over time, in a protracted process of trial
accountability of the decision by patients (Stivers et al.,
and error. Service-user knowledge and preferences need to be
2017; Thompson & McCabe, 2017). In addition, doctors use
at the centre of psychiatric consultation encounters for a
linguistic devices such as euphemisms and jargon which limit
number of reasons: (a) relations between mental health pro-
possibilities for discussing patient preferences and knowledge
fessionals and service-users often revolve around medication
and therefore reduce possibilities for SDM (Chou et al., 2017).
since the vast majority of people who receive mental health
Research has also shown how the patient can influence the
services in the UK are prescribed psychiatric drugs; (b) diag-
professional’s recommendations in a variety of ways, for
nosis, aetiology, and approaches to treatment are complex and

CONTACT Emma Kaminskiy emma.kaminskiy@anglia.ac.uk; emma.kaminskiy@anglia.ac.uk Department of Psychology, Anglia Ruskin University, East Road,
Cambridge, CB1 1PT Cambridgeshire, UK.
© 2018 Taylor & Francis Group, LLC
2 E. KAMINSKIY AND M. FINLAY
example by minimal agreements, silences, and counter pro-
posals (e.g. Costello & Roberts, 2001; Gill, Halkowski, &
Roberts, 2001; Roberts, 1999), and that practitioners employ
strategies to facilitate service-user involvement (e.g. Mishler,
1984). In mental health settings, professionals’ use of knowl-
edge displays and pseudo- and hypothetical questions can
assist service-users expand on topics and express preferences
on sensitive topics (Pino, 2016; Speer, 2010) and commonly
used question types (e.g. declarative statements requiring con-
firmation) are associated with improved perceptions of the
therapeutic relationship. In addition, psychiatrists have been
found to use two different types of accounts to justify treat-
ment recommendations: client attentive accounts and author-
ity-based accounts (Angell & Bolden, 2015).
These mixed findings on communicative practices point to
the importance of setting, context, clinical judgement of risk,
and the existence of pre-existing relationships for determining
the function of talk in psychiatric encounters. In addition, the
level of observed involvement, and patient resistance, will vary
depending on both psychiatrist and service-user characteris-
tics (Bergen et al., 2017; McCabe et al., 2013). Describing this
difference, Quirk et al. (2012) propose a typology of open,
directed and pressured shared decisions depending on the
type of pressure applied by the psychiatrist.
In short, research to date provides some insight in understand-
ing how SDM works in practice by identifying different general
approaches adopted by psychiatrists and by showing the types of
broad interactional behaviours which can shut down or enable
possibilities for service-user involvement. In this paper, we extend
this research by undertaking a detailed examination of the com-
municative patterns between psychiatrist and service-user in a
small set of examples. We explore how the service-user can influ-
ence what SDM looks like in practice, investigating the interac-
tions and communicative choices made by both participants when
discussing psychiatric medication. In particular, this paper aims to
examine how doctors’ behaviours adapt to the interactional turn of
the service-user, with a particular focus on the power-claiming
strategies employed by service-users in this context.
Data and method
We adopt a mixed-method case study approach to the analysis,
an applied conversation analysis (CA) supplemented with addi-
tional consideration as to how the meetings fit with the key
defined criteria of SDM. We investigate how similar and different
these appear to both what is already known about the interac-
tional dynamics of real-time encounters and also consider how
they fit with the key criteria of SDM models of patient involve-
ment (a deliberation and agreement, following a full weighing up
or the pros and cons and with full consideration of the service-
user experience, preferences, and values). Rather than identifying
typical patterns in such interactions, our aim was to explore in
detail the different ways that conversations concerning a person’s
psychiatric medication develop in a small set of examples.
Applied conversation analysis
The principles of CA have been employed as it allows for a
detailed inspection of the turn taking organisation of decision
sequences within naturally occurring talk. CA has been used
extensively in research into communication practices between
health-care professionals and patients. The analysis draws on
the literature presented above and what is known about
power-claiming and power-enabling strategies employed by
physicians as well as the active and power-claiming discursive
strategies that service-users employ. By adopting an applied
CA, the presentation of relevant background and contextual
data is afforded (ten Have, 2001).
Data
We present three case studies, based on interviews and recorded
meetings between a psychiatrist and service-users discussing med-
ication. The meetings took place in the East of England within an
NHS foundation trust. Participants were recruited from a com-
munity mental health team/pathway (CMHT) that served
approximately 260 people with diverse needs, who were broadly
considered to have severe and potentially enduring mental illness.
The data formed part of a primary research project that took place
in the CMHT between 2010 and 2014 which involved a series of
interviews with service-users, psychiatrists, and community psy-
chiatric nurses investigating views and experiences of SDM for
psychiatric medication management. The cases were selected for
inclusion because they illustrate three different patterns of power-
enabling and power-claiming strategies employed by both actors.
The first shows a more obviously successful process. The second
appears successful up to a point, but then changes. The third shows
a service-user only partly taking up opportunities to influence the
decision, and the psychiatrist needing to make more effort. In each
case, we see the actors sensitively designing their utterances for the
other, and in the second and third examples we see where oppor-
tunities for greater engagement are missed. The three recorded
meetings presented take place with the same psychiatrist.
Researchers were not present and meetings lasted between 13
and 38 minutes and recorded using a digital audio recorder.
Only those parts of meetings where medication is discussed were
transcribed and all data have been anonymised to protect the
identity of the speakers. The Jeffersonian approach to transcription
was employed to retain the characteristics of speech delivery (see
Atkinson & Heritage, 1999). The NHS National Research Ethics
Service granted ethical approval and participants provided full
informed consent.
Findings
The first example illustrates an interaction where the service-
user, Lara, has active involvement in the decision-making
process. The psychiatrist provides her opinion, but phrases it
in such a way that allows it to be challenged, and Lara takes
up this opportunity. In the second example, the service-user
(Rosie) is actively involved in the medication decision up to
the point it is agreed, but then is not engaged in the subse-
quent discussion of side-effect management. In the final
example, the service-user (Linda) is less assertive, and her
experiential knowledge less apparent. While she has opportu-
nities to contribute, she only partly takes these up. In this
example, we can see the paradox that more psychiatrist
 HEALTH COMMUNICATION 3

persistence and direction might be necessary in order to
produce a more fully equitable encounter.
Lara: Actively claiming speaker rights
Lara (woman, 42) has received services from the community
team for a number of months. Extract 1 is taken from a review
meeting with Dr Kos, whom Lara has been seeing approxi-
mately once per month. Lara was seeing another psychiatrist
prior to Dr Kos, and this changed when the psychiatrist left
the team. A male friend/family member is present with Lara
in the meeting. Lara has had her psychiatric medication
changed in recent months, and is currently being prescribed
Lamotrigine, a mood stabiliser.
This meeting lasts 20 minutes, of which 10 minutes are
spent discussing medication. The extract below occurs 10 min-
utes into the meeting. Up to this point, Lara said that her
mood has improved since increasing the dose, although she is
having difficulty concentrating, is a bit disorganised, her
memory is worse than normal, and she has had a recent
breakout of eczema.
Here, Lara often claims speaker rights. She overlaps Dr Kos’s
speech to confirm (6) and then modify (7) the doctor’s suggestion
(see Drew, 2009 for discussion of overlaps in conversation), and
asks questions about the maximum dose (14–15, emboldened). Dr
Kos proposes leaving the medication dosage unchanged in line
23–24: ‘I would leave it to the three hundred with a view to
increasing if you feel that winter is coming. . .’. Note that Dr Kos
leads up to this opinion by first stating two considerations, Lara’s
skin condition and the fact that her mood has already started to
improve (for analysis of how doctors provide rationales before
treatment recommendations, see Roberts, 1999). Dr Kos does not
issue this as a directive, but in a less authoritative way: as her
personal preference (‘I would leave. . .’), and finally she leaves open
the possibility of increasing the dose in the future. This allows the
possibility for Lara to disagree.
Looking at this in more detail, Lara’s request to increase the
dose in line 40 is worked up to gradually, starting with the question
(line 14) ‘is it at the top end of how much you should take’ – a way
of testing the water (see Gill et al., 2001, for other examples). Dr
Kos appears to pick this up, explaining that the dose could be
increased, but then suggests the dosage should remain as it is,
albeit with the ‘get out’ of possibly increasing it in the winter if Lara
has a dip in her mood. Lara then declares she does not ‘normally
have much of a seasonal change’ (26–27), the use of ‘normally’ and
‘much of a’ acting to soften the challenge. Lara’s explicit proposal

Extract 1. P = Dr Kos L = Lara.

Outcome – increase the dose of Lamotrogine
01 P yes yes oh right yep yep (.) so u:::m but still I think because your mood
02 improved although they are (.) what you mentioned about duh being
03 disorganised and the problem with the concentration can be st:::ill uh the
04 symptoms of th::e of the of the mood so it still might be still just a bit u:h
05 under uh under normal and and [prese
06 L [it feels
07 (?) like that but its not as m::uch under normal as it was last time
08 [I saw you]
09 P [yes yes]
10 L [uh I’m sure the last time I saw you I am
11 probably coming across differently to when [I last saw you
12 P [yes yes absolutely
13 absolutely[
14 L [so yeah I think it is better but is it at t:he ^top end of how
15 much^ [you should take
16 P [No no its not no its its its still at least at least hundred
17 milligrams that we can increase to four hundred milligram is the
18 medically the dose that we recommend but you can go even high::er
19 [because there ar::e] differences between yeah how different people
20 L [O K :::: a y]
21 P tolerate the medication but at the moment considering the (.) uh the skin
22 condition and also that your mood is a::lready started to improve I would
23 leave to the three hundred u::h with a view of increasing if you feel that
24 winter is coming if the mood starts to dip and is is its again getting
25 u::nstable and you can increase again [by fifty milligrams
26 L [yeah ye::ah I don’t think I
27 normally have much of u::h much of seasonal [change particularly]
28 P [seasonal oh right]
29 L in that its u:h I’ve been re:ally really low in the summer [and
30 P [summer yeah
31 exactly when we met[
32 L [so but I don’t fe:::el that there’s a uh there might
33 be a first time use optimus but I find that I don’t have th::at much uh
34 awareness of it because again cuz like my memory is sh:::ot (.) um
35 P = but I think again its a vicious circle s::o once you eat properly you drink
36 properly your memory might simply just improve because you are not
37 you are h::ydrated and then uh an and and we[
38 L [ueu ((stutter)) yeah thats
39 possible I mean I fe:::el that I would qui::te like to take the risk of
40 the eczema and go up a bit [if your:::e OK with that
41 P [OH right OK that’s that’s fine yeah I agree
42 yeah I’m fine with it
4 E. KAMINSKIY AND M. FINLAY

to increase the dose in lines 38–40 begins with a minimal agree-
ment with the doctor’s suggestion (yeah that’s possible) followed
by an ‘I-mean-prefaced utterance’ (Maynard, 2013). This is a
device often used tie back to the speaker’s earlier utterance, and
and complete a course of action when there has been less than full
uptake of the speaker’s previous turns by the recipient. Here we see
it used when the more implicit ways of making the request are not
picked up by the doctor. Then she states she is willing to take the
risk of eczema, thus dealing with another of Dr Kos’s arguments
for leaving the medication unchanged. Her suggestion to increase
the medication doseis phrased as her own preference (‘I would
quite like to. . .’), and she seeks the approval of the doctor (‘If you’re
all right with that’). Lara therefore asserts her own medication
preference in contradiction to the doctor’s in a delicate way avoid-
ing threatening Dr Kos’s professional authority and identity (see
also Costello & Roberts, 2001; Gill et al., 2001). Dr Kos accepts
Lara’s proposal by saying ‘I agree with that’.
In summary, this extract shows evidence of a shared encoun-
ter. Dr Kos provides medical opinion and expertise about psy-
chiatric medication, but phrases this in a way that is open to
challenge by Lara. When faced with Lara’s challenge, Dr Kos does
not push on or try to persuade Lara to reconsider (e.g. see Quirk

Extract 2. P = Dr Kos; R = Rosie.

Outcome – Change anti-psychotic medication from Aripriprozole to Olanzapine
01 P But at the moment you don’t fee::l that uh they are excluding you or or
02 they seem to be or=
03 R =I don:::t kno:::w I find it so::o difficult to get past my own [moo::d]
04 P [yes yes]
05 R and the voices as well [
06 P [uh hum um
07 R which leads me to my next que::stion (.) would it be::e possible to try
08 a different medication at sometime?
09 P Yes I think it might be a good idea to see whether you have any (.) uh (.)
10 I mean we switched to aripripozole simply because of the side-effects
11 you had on on a amisulpride and uum (.) and it might be worth to to try
12 something different because even the combination of the two when you
13 were taking a very low dose of amisulpride then it seemed to work (2.0)
14 because of the side effects=
15 R = because of the [side-effects
16 P [side-effects exactly (.) so u::m (.) in terms of other
17 medication I mean are you happy on Venlafaxine in terms of the mood do
18 you think that’s that working=
19 R =Yes=
20 P =YES yes we we will not change that so really the question is whether we
21 should change the aripriprozole uh the abilofy to something different (.)in
22 terms of uh (.) a group of medication which are the same group of
23 medication you you you’ve taken amisulpride and it has side-effects and
24 the other ones are olanzapine uh quetiapine risperidone have you had
25 any experience with any any ^of these^ =
26 R =Unfortat:::[ely
27 P [^ALL OF ^them=
28 R =Most of them
29 P [Most of them=
30 R =no:::ot [olanzapine] but all the rest I’ve tried risperidone I find quite
31 P [uh hum]
32 R str[a::nge] quetiapine I got(.) I felt quite ^[frag:::ile]^ with it and um
33 P [uh huh] [hhmm]
34 R (2.0) what was the other one?=
35 P =Uh (1.0) uh olanzapine I think olanzapine what you mentioned
36 R olanzapine I haven’t tried=
37 P =Yes yes[
38 R [Ive heard that makes you eat lo:::Ads[so
39 P [Yes::s I mean this this
40 can this can be a side [effect] yes[
41 R: [humm] [OKay (.) mayb::e a low dose would be
42 Okay=
43 P =Yes yes yes and in terms of the uh the problem you had with the
44 amisulpride I think it’s less likely with olan::zapine (.) so your prolactin
45 level was quite high even on the small dose of amisulpride I hope that
46 much less likely on olanzapine BUT I think what we should do we
47 should have a prolactin level tested now just to have a baseline level and
48 probabl::y ab::out a couple of weeks later just to see whether you have
49 any problems on olanzapine and we’ll keep an eye on your uh (.) on uh
50 your eating and on your weight as well. SO IF you if you would join the
51 gym no::w that would be a very good idea because then then uh (.) then
52 you might be able to to kind of control the the weight gain, the weight
53 gain unfortunately can be side effects with the olanzapi::ne=
54 R =would the la::dy in the social inclusi::on (.) help me with that?=
55 P =Yes I think I think what we should do just uh (.) because um (name) is
56 on sick leave and you don’t really need THAT kind of support that you
57 have been receiving from her (.) do you see anyone else (.) uh like (name
58 removed) (.) you’re not seeing anyone else alright
{conversation moves on to discuss support for housing and Rosie’s living situation}

et al., 2012). Instead, Lara brings her own experience, preferences
and openness to risk to bear on the decision but phrases her
challenge in a way that attends to issues of authority and face.
Rosie: A shared encounter, well to a point. . .
Rosie (woman, 24) has been in contact with Dr Kos for a
number of months, but has recently missed a few scheduled
routine appointments. Rosie has recently been in contact with
the out-of-hours crisis team. At the time of the meeting, Rosie
was being prescribed an anti-depressant, Venlafaxine and the
anti-psychotic medication, Aripriprozole. Rosie has been pre-
scribed various psychiatric medications for approximately
8 years and has previously experienced adverse effects of
different anti-psychotic medications (including problems
with weight gain and prolactin levels).
The key decision in this meeting is to change the type of
anti-psychotic medication. The following excerpt occurs at
approximately 12 minutes into the conversation. In the
Extract 2, below, Rosie asks explicitly about changing the
medication. She is subsequently actively involved in discus-
sion about medication alternatives, however, once a decision
has been made to try the new medication, the conversation
about adverse effects is brushed over.
In previous work, Gill et al. (2001) showed how patient requests
can be built subtly over long sequences. Before this extract, topics
related to medication were discussed at two separate points
(describing problems her voices are causing her, and problems of
drowsiness). In the extract above, after providing reasons why she
finds it hard to answer Dr Kos’s question due to her mood (03) and
the voices (05), Rosie moves onto the sequentially related topic of
changing her medication (07–08). Rosie requests this using a
question that has three features that attend to the delicacy of the
matter. She does not state which of her two medications she would
like to change (leaving this issue open for discussion), she uses
polite phrasing (‘would it possible’) which suggests deference to
the doctor’s authority, and she adds ‘at sometime’, which down-
plays the urgency (for another example of this, see Gill et al., 2001).
Dr Kos treats the question as a proposal to which she agrees (‘Yes, I
think it might be a good idea. . .’) (09) followed directly by a
description of the previous rationale for the prescribing aripripo-
zole (10–14). This protects Dr Kos’s ‘face’ since it defends against
the possibility that the previous decision was not properly justified.
Dr Kos then uses two gateway closed questions about whether
Rosie is happy with the Venlafaxine (17–18). Rosie responds with
the affirmative, which leads Dr Kos to move back to the issue of
whether to change the aripripozole (21). She does this with a
sequence of questions about Rosie’s previous experiences of the
same group of antipsychotic medications. This shows Dr Kos
respecting Rosie’s greater experiential knowledge of these medica-
tions (as opposed to her own professional knowledge – for a
discussion of epistemic status see Heritage, 2013), and can be
seen as serving an empowering and collaborative function (as
opposed to, for example, referring directly to Rosie’s medical
notes and making a treatment proposal) (see Thompson, Howes,
& McCabe, 2016).
Rosie responds to the questions about medication by stating
she has tried them and each had adverse effects, making her feel
HEALTH COMMUNICATION 5
‘strange’ and ‘fragile’ (32). The only one she has not tried is
Olanzapine (36). However, she expresses a concern in line 38 –
that ‘I have heard it makes you eat lo:::ads’. Again, she attends to
issues of voice: she attributes the potential problem to other
sources rather than expressing a direct opinion herself. Through
their use of voice, both parties acknowledge the epistemic gradient
between them (Heritage, 2013). Rosie does not claim direct knowl-
edge about the side effects (since she has not tired the medication
herself), while Dr Kos, with professional knowledge of such things,
does claim the knowledge (‘yes I mean this can be’). Note that
Rosie phrases the problem as definite: ‘it makes you. . .’. Dr Kos
only partly supports this – she modifies Rosie’s definite statement
by using the word ‘can’: ‘Yes I mean this can be a side-effect’. This
transforms the problem into only a possibility, which therefore
might not actually happen. In the next section, Rosie proposes
trying Olanzapine, but at a low dose (41–42).
Thus far, the dialogue is relatively shared and Rosie draws
on her own experience and knowledge to direct the decision,
taking an active role through ruling out some options and
expressing concern over a side-effect of the only option she
has not tried. Dr Kos responds positively to these turns,
asking questions to encourage Rosie to bring her knowledge
to bear in the encounter, while at the same time preserving
social identity rights and saving face.
However, once the option to try Olanzapine is agreed by
Rosie, Dr Kos immediately reinforces this as the preferred
option, and from this point, little attempt is made to involve
Rosie in exploring her concerns about adverse effects. Similar
to Quirk et al.’s (2012) findings, at this point, Dr Kos instead
appears to work towards securing the agreement and once
Rosie has suggested a low dose (41–42), Dr Kos agrees and
moves onto a monologue about a different side effect which
Rosie has not raised (prolactin levels) (lines 43–49) without
any invitation to Rosie for her involvement (‘I think what we
should do. . .’). Dr Kos finally returns to the concern Rosie had
expressed (50), but proposes a set of management strategies
for the possible weight gain (lines 50–53) without soliciting
Rosie’s participation.
Linda: An inability to assert speaker rights?
This is only the second meeting between Linda (woman, 26) and
Dr. Kos. Prior to this, Linda has seen psychiatrists in other teams.
At the time of this meeting, Linda was also seeing two other
psychiatrists, one in the eating disorder team. This is unusual for
this mental health team, and may be indicative of complexity of
the problems at that time. Linda was being prescribed
Venlafaxine, an anti-depressant and Gabapentin, a mood stabi-
liser. Other psychiatrists issued both prescriptions. Linda has
many years’ experience of psychiatric medication and has pre-
viously received electro-convulsive therapy (ECT). The meeting
in total lasts for 38 minutes, and medication-related conversa-
tion lasts for approximately 8 minutes. There is no direct change
to her prescription, however, the time of day the medication is
taken is changed in this meeting. Plans for future possible
changes are also discussed, including the future consideration
of the anti-psychotic, Quetiapine and/or a course of ECT.
This discourse is highly asymmetrical in speaker rights,
with Dr Kos asking all of the questions throughout the
6 E. KAMINSKIY AND M. FINLAY

Extract 3. P = Dr Kos; L = Linda.

No change to medication
01 P So at the moment you take (.) let’s just see if it’s in here yeah uh (,) you
02 are taking a Venlafaxine four hundred and fifty milligrams um and the
03 dose has just been increased about ten days ago a week ago=
04 L =yea::h=
05 P (.)= uh are you happy with the medication? Do you have any side effects
06 at all or?=
07 L = I haven’t noticed any (2.0)
08 P And you have been taking Venlafaxine in
09 the past so it’s actually the second time you are taking (.) ((sigh)) but in
10 the past you were also taking it with a combination of different
11 medication?=
12 L = um ye::s I was on it was (.) um (3.8)
13 P mirtazapine=
14 L = (.) yeah, and [um] and I think I was also on promatazine at the same
15 time so [yeah (.) so]
16 P [hum AND um um and uh] (2.0) if you you took different
17 combination different medication uh (.) do you remember which of those
18 you find the most beneficial at all?=
19 L =(.) um (6.2) from what I rem::ember um they a::ll (2.8) seemed to (2.0)
20 work for uh short while but then the effects just stopped and my mood
21 went down again (3.0) so
22 P ((sigh)) I mean (.) at the moment because you
23 just increased the dose I think we have to wait a little while to see to
24 assess how things are going. You are taking other medication as well (.)
25 gabapentin, why gabapentin was prescribed for you?=
26 L = Um (2.0) I had really um (1.0) bad (1.0) muscle pains
27 P: Yes=
28 L =in my legs and I couldn’t really do anything (1.0) so it <was for that=
29 P =And do you still take the gabapentin?=
30 L =yeah=
31 P =yeah and what dose do you take?=
32 L =a hundred=
33 P =a hundred=
34 L =daily=
35 P =yeah a hundred daily. You have it once a day (.) Um Once a day yes?

dialogue and, on a few occasions, making direct treatment
proposals. This example portrays an interaction where Linda
is offered opportunities by the psychiatrist to contribute but
only partly takes these up.
In Extract 3, Dr Kos asks a total of nine topic directing,
closed or wh’ questions (see Thompson et al., 2016). The
frequency of questions and short duration of turns is notice-
ably different from the previous excerpts presented. Notably,
this was only Dr Kos’s second meeting with Linda, which
may also indicate that formulation and assessment were
priorities.
There are, however, also many demonstrations of Dr
Kos attempting to engage Linda by creating slots for
Linda to assess the effectiveness of the treatments she has
tried before and bring her experience to bear in the
encounter. For example, in lines 31, 33, and 35, Dr Kos
repeats Linda’s responses, which illustrates she is being
heard and encourages additional narrative. In addition,
Dr Kos does not attempt to repair Linda’s turn during a
number of long pauses and silences (lines 19, 21, 26). In
the CA literature, transitions (from one turn to a next)
with no or only a slight gap and make up the vast majority
of transitions (Sacks, Schegloff, & Jefferson, 1974) and, as
such, the noticeable pauses in this excerpt are unusual. In
line 19, Linda pauses three times in speech (in response to
a direct question which of those you find the most benefi-
cial?). It is striking that at no point in Linda’s turn does
Dr Kos interject or attempt to repair the sequence by
filling the silence, even during a very long pause (6.2
seconds). On Linda’s part, it could be that these extended
pauses represent a form of resistance (common in institu-
tional interactions), or are allowing time to cognitively
prepare for talk. Avoiding the temptation to fill silences
in conversation can be empowering, particularly when the
speaker has less institutional power or needs longer to
formulate their speech.
However, at other points, we see opportunities for
shared exploration missed. For example, in line 5, Dr
Kos asks two questions about whether Linda is happy
with Venlafaxine and if she has any side effects. Linda
Extract
4. P = Dr Kos L = Linda.
36 P (2) uh do you have any uh (.) plan at the moment what you would
37 prefer to take if you have to take something in addition to the (1)
38 L um Dr
39 Shara um suggested um (2.0) either addi::ng quetiapine (1.0) or um
40 (3.0) um (1.0) uh (.) a ^course of ECT [again]^ cuz that’s helped in
41 P [yes]
42 L the [past] so=
43 P [hum]
44 P =How would you fe::el about uh (.) ECT again?
45 L (9.0) I’m not really (.) sort of (.) worried (.) about ECT cuz I’ve had it
46 quite a few times already=
47 P = And did you find it really helpful actually or did you =
48 L = Um (.) yeah and like the people I was seeing and stuff and Dr
Shara
49 has (2) like said it definitely improved [my] mood so
50 P [yes, uh hum]
51 P = Have you tried medication ((cough)) called lamotragine in the past?
52 L = NO

only answers the latter question, saying she hasn’t ‘noticed
any’, but fails to confirm that she is happy with the
medication. This might indicate that Linda has some mis-
givings, particularly given her less than complete denial of
side-effects (saying she hasn’t noticed any leaves open the
possibility that there might be some side-effects of which
she is not yet aware). However, when Linda doesn’t fill the
short silence, Dr Kos then moves the topic on in line 8
rather than returning to the unanswered question of
whether Linda is happy with the medication. In lines
20–21, Linda reports that a potential problematic experi-
ence with her past use of the same medication along with
others she was taking at the time, namely that ‘then the
effects just stopped and my mood went down again’. In
this case, Dr Kos suggests they cannot change the medica-
tion given it has just recently been increased in dosage
(23). While this may indeed be clinically justified, it also
does not provide Linda further opportunity to explore her
possible misgivings.
In Extract 4, which occurs shortly after, Linda does exert
more speaker rights.
Here, Dr Kos seeks Linda’s opinion on future medication
plans (36–37) and Linda responds by suggesting trying
Quetiapine/ECT (39–40). Linda voices the proposal through a
third person who would be assumed to have expert knowledge
and institutional authority: ‘Dr Shara um suggested. . .’. In
Roberts’ (1999) analysis of interactions in oncology consulta-
tions, patients sometimes voiced objections to the doctor’s
recommendations and/or rationale by attributing them to
other doctors or to something they had read. The difference
here is that Linda is orienting to the controversial nature of
what she herself is proposing, and is anticipating a possible
objection by doctor. In response, Dr Kos asks for Linda’s opinion
directly (how would you feel about ECT again?). Linda replies
using her own voice that she is not worried about it (perhaps
orienting to the controversial nature of the treatment). Dr Kos
next asks about Linda’s experience, using an either/or question
which signals that Linda can disagree with the treatment sug-
gested by Dr Shara. Linda replies by again using Dr Shara’s voice
to claim it had positive effects (48–49). At this point, rather than
pursuing the topic to find out whether Linda really felt about the
two suggestions, Dr Kos changes the topic and introduces a new
possible medication option ‘Lamotrogine’ (51).
In summary, Linda appears only partially takes up opportu-
nities to claim her speaker rights and exert experiential knowledge,
and Dr Kos misses opportunities to encourage her to discuss her
possible misgivings further. In Extract 3, Dr Kos does ask specifi-
cally about how she feels about her medication, but Linda replies in
a limited way. She does provide some clues that she might not be
satisfied, but these are not actively followed up by Dr Kos. When
Linda does express an alternative course of treatment in Extract 4,
she voices it through an authority-figure. Voicing this through
another person with institutional authority might be particularly
important given the content of what is being proposed by Linda at
these points (both Quetiapine and ECT are associated with serious
side effects and ECT is a controversial and rare therapy in this
HEALTH COMMUNICATION 7
context). Throughout, there appears to be no consideration of the
possible adverse effects. While Dr Kos uses discursive techniques
to engage Linda a close inspection of the interaction reveals that
more opportunities were probably available.
Discussion
This article has examined the interactional and communicative
choices that both participants (mental health service-user and
psychiatrist) make when discussing psychiatric medication. The
meetings took place in an institutional setting where structural-
and system-level constraints to SDM exist. Some of these struc-
tural factors are specific to the mental health treatment context
(e.g. judgement of insight, capacity, fear of coercion, diagnosis,
labelling, and stigma). These structural features present particular
challenges for different forms of knowledge having equal status in
psychiatric encounters concerning medication. The analysis
demonstrates the importance of viewing the practicalities of
implementing SDM in context, as involving two participants in
a complex interplay. Previous research in other settings have
found that patients can influence the professional’s recommenda-
tions and that practitioners also employ strategies to facilitate
service-user involvement in encounters (Costello & Roberts,
2001; Gill et al., 2001; Mishler, 1984; Roberts, 1999).
All extracts presented in this paper involve the same psy-
chiatrist, Dr Kos, and all occur in the same treatment setting.
Given the contextual similarities of these meetings, we might
expect a level of similarity between these dialogues in terms of
how the meetings conform to the requirements of SDM, and
the discourse cues employed by the actors involved. On the one
hand, aspects of the SDM process identified in the literature
(Charles, Gafni, & Whelan, 1999; Edwards & Elwyn, 2009) are
not fully adhered to. Dr Kos does not present a full or complete
set of options in any of these meetings and this represents one
way in which power is enacted (Kaminskiy, 2015). The weigh-
ing up of pros and cons and, in particular, consideration of
potential adverse effects, and how these relate to the service-
users’ ‘life world’, is also not fully considered (Mishler, 1984).
On the other hand, experiential knowledge is both solicited and
attended to by Dr Kos, and the idea of a ‘meeting of experts’
applies to some of the excerpts analysed (e.g. Lara and Rosie).
This mixed pattern supports Seale, Chaplin, Lelliott, & Quirk’s
(2006) study into psychiatrists’ experiences where juxtaposed dis-
courses of doctor influence, professional judgement, and expert
knowledge were found alongside a genuine desire to support
greater service-user autonomy and ownership of decisions.
In Quirk et al.’s (2012) study of anti-psychotic prescribing
decisions between psychiatrists and mental health service-users,
three decision types: ‘open’, ‘directed’, and ‘pressured’ decisions
were proposed. Directed decisions, where ‘decisions are pro-
duced collaboratively, by the diplomacy of the doctor and by
the absence of resistance by the patient’ (p. 105), and, open
decisions, where the decision is made with little or no little or
no pressure being exerted by the psychiatrist, occurred most
frequently, and these are reflected in our findings. However,
our analysis also points to the changing nature of the
8 E. KAMINSKIY AND M. FINLAY
interaction types over time, e.g. with Rosie’s interaction moving
between features of open and directed decisions throughout the
dialogue. However, Lara’s dialogue, at times, is less well
described by the typology and represents, potentially, a more
empowering portrayal of active service-user participation in
practice. Further research could explore these complexities. In
particular, further research that moves beyond investigations of
how service-users respond to the interactional cues of the
doctor, towards a more sophisticated understanding of how
the doctor responds and supports the active and power-claim-
ing strategies of the service-user is warranted.
Finally, our findings also suggest that missed opportunities to
involve people in decisions need further investigation (see Linda).
This is an important practical point, and suggests that the point at
which the psychiatrist ‘gives up’ needs attention. With more
reticent service-users, the psychiatrist might have to take more
time and use more encouragement strategies in order to involve
them. Additional research exploring the interactional cues of these
more challenging interactions would be fruitful.
In summary, this article adds to our understanding of SDM in
mental health, an under-researched context, by exploring in
detail the different ways that conversations concerning a per-
son’s psychiatric medication develop in a small set of examples.
Much of the existing literature on treatment recommendations is
prescriptive; by adopting an exemplar approach, a more detailed
examination in the context of the history of each patient is
afforded. Future research triangulating the scale and scope of
this variation to other work is recommended.
Nonetheless, the findings provide important illustrations of
how a physician encourages SDM, as well as where opportu-
nities to engage service-users fails. The findings also provide a
detailed examination of how patients do and do not take part
in the process. Findings suggest that some service-users do
not readily respond to doctors attempt to engage them while
others take a much more proactive approach. Service-users
make proposals and suggestions in complex ways that attend
to issues of power and authority, and within a single encoun-
ter doctors can vary from encouraging SDM, to shutting
down topics emphasised as important by service-users. At
times, we suggest there may be missed opportunities to
involve people in decisions. However, in circumstances
where the patient actively brings their experiential knowledge
to bear, opportunities emerge for meaningful and shared
dialogue. This is particularly relevant for decisions concerning
psychiatric medication where people often have extensive
experience and strong preferences towards treatment and
where the cost–benefit ratio is unclear.
References
Angell, B., & Bolden, G. B. (2015). Justifying medication decisions in
mental health care: Psychiatrists’ accounts for treatment recommen-
dations. Social Science & Medicine, 138, 44–56. doi:10.1016/j.
socscimed.2015.04.029
Atkinson, J. M., & Heritage, J. (1999). Jefferson’s transcript notation. In
A. Jaworski & N. Coupland (Eds.), The discourse reader (pp. 158–166).
New York, NY: Routledge.
Bergen, C., Stivers, T., Barnes, R. K., Heritage, J., McCabe, R., Thompson,
L., & Toerien, M. (2017). Closing the deal: A cross-cultural
comparison of treatment resistance. Health Communication, 33,
1377–1388. doi:10.1080/10410236.2017.1350917
Charles, C., Gafni, A., & Whelan, T. (1999). Decision-making in the
physician–patient encounter: Revisiting the shared treatment deci-
sion-making model. Social Science & Medicine, 49, 651–661.
doi:10.1016/S0277-9536(99)00145-8
Chou, W. Y. S., Hamel, L. M., Thai, C. L., Debono, D., Chapman, R. A.,
Albrecht, T. L., . . . Eggly, S. (2017). Discussing prognosis and treatment
goals with patients with advanced cancer: A qualitative analysis of oncol-
ogists’ language. Health Expectations, 20, 1073–1080. doi:10.1111/
hex.12549
Costello, B. A., & Roberts, F. (2001). Medical recommendations as joint
social practice. Health Communication, 13, 241–260. doi:10.1207/
S15327027HC1303_2
Deegan, P., & Drake, R. (2006). Shared decision-making and medication
management in the recovery process. Psychiatric Services, 57, 1636–
1639. doi:10.1176/ps.2006.57.11.1636
Drew, P. (2009). Quit talking while I’m interrupting: A comparison
between positions of overlap onset in conversation. In M. Haakana,
M. Laakso, & J. Lindström (Eds.), Comparative aspects of conversation
analysis (pp. 70–93). Helsinki, Finland: Finnish Literature Society.
Edwards, A., & Elwyn, G. (2009). Shared decision-making in health care:
Achieving evidence-based patient choice. Oxford, UK: Oxford University
Press.
Gill, V. T., Halkowski, T., & Roberts, F. (2001). Accomplishing a request
without making one: A single case analysis of a primary care visit.
Text, 21, 55–81. doi:10.1515/text.1.21.1-2.55
Heritage, J. (2013). Epistemics in conversation. In J. Sidnell & T. Stivers
(Eds.), Handbook of conversation analysis (pp. 370–394). Boston, MS:
Wiley-Blackwell.
Kaminskiy, E. (2015). The elephant in the room: A theoretical examina-
tion of power for shared decision making in psychiatric medication
management. Intersectionalities, 4, 19–38.
Maynard, D. W. (2013). Defensive mechanisms: I-mean prefaced utter-
ances in complaint and other conversational sequences. In M.
Hayashi, G. Raymond, & J. Siddell (Eds.), Conversational repair and
human understanding (pp. 198–233). Cambridge, UK: Cambridge
University Press.
McCabe, R., Heath, C., Burns, T., & Priebe, S. (2002). Engagement of
patients with psychosis in the consultation: Conversation analytic
study. British Medical Journal, 325, 1148–1151. doi:10.1136/
bmj.325.7373.1148
McCabe, R., Khanom, H., Bailey, P., & Priebe, S. (2013). Shared decision-
making in ongoing outpatient psychiatric treatment. Patient Education
and Counseling, 91, 326–328. doi:10.1016/j.pec.2012.12.020
McCabe, R., Sterno, I., Priebe, S., Barnes, R., & Byng, R. (2017). How do
healthcare professionals interview patients to assess suicide risk? BMC
Psychiatry, 17, 122. doi:10.1186/s12888-017-1489-6
Mishler, E. G. (1984). The discourse of medicine: Dialectics of medical
interviews. New Jersey, NY: Greenwood Publishing Group.
Moncrieff, J. (2009). The myth of the chemical cure: A critique of psychia-
tric drug treatment. Basingstoke, UK: Palgrave MacMillan.
National Institute of Clinical Excellence. (2009). Medicines adherence:
Involving patients in decisions about prescribed medicines and support-
ing adherence. London, UK: Author.
Pino, M. (2016). Knowledge displays: Soliciting clients to fill knowledge
gaps and to reconcile knowledge discrepancies in therapeutic interac-
tion. Patient Education and Counseling, 99, 897–904. doi:10.1016/j.
pec.2015.10.006
Quirk, A., Chaplin, R., Lelliott, P., & Seale, C. (2012). How pressure is
applied in shared decisions about antipsychotic medication: A con-
versation analytic study of psychiatric outpatient consultations.
Sociology of Health & Illness, 34, 95–113. doi:10.1111/j.1467-
9566.2011.01363.x
Roberts, F. D. (1999). Talking about treatment: Recommendations for
breast cancer adjuvant therapy. Oxford, UK: Oxford University Press.
Sacks, H., Schegloff, E. A., & Jefferson, G. (1974). A simplest systematics
for the organization of turn-taking for conversation. Language, 50,
696–735. doi:10.1353/lan.1974.0010

Seale, C., Chaplin, R., Lelliott, P., & Quirk, A. (2006). Sharing decisions
in consultations involving anti-psychotic medication: A qualitative
study of psychiatrists’ experiences. Social Science & Medicine, 62,
2861–2873. doi:10.1016/j.socscimed.2005.11.002
Seale, C., Chaplin, R., Lelliott, P., & Quirk, A. (2007). Antipsychotic
medication, sedation and mental clouding: An observational study
of psychiatric consultations. Social Science & Medicine, 65, 698–711.
doi:10.1016/j.socscimed.2007.03.047
Speer, S. (2010). Pursuing views and testing commitments.
Hypothetical questions in the psychiatric assessment of transsex-
ual patients. In A. F. Freed & S. Ehrlich (Eds.), Why do you ask?
The function of questions in institutional discourse (pp. 133–158).
Oxford, UK: Oxford University Press.
HEALTH COMMUNICATION 9
Stivers, T., Heritage, J., Barnes, R. K., McCabe, R., Thompson, L., &
Toerien, M. (2017). Treatment recommendations as actions. Health
Communication, 33, 1335–1344. doi:10.1080/10410236.2017.1350913
ten Have, P. (2001). Applied conversation analysis. In A. McHoul & M.
Rapley (Eds.), How to analyse talk in institutional settings: A casebook
of methods (pp. 3–11). London, UK: Continuum.
Thompson, L., Howes, C., & McCabe, R. (2016). Effect of questions
used by psychiatrists on therapeutic alliance and adherence. The
British Journal of Psychiatry, 209(1), 40–47. doi:10.1192/bjp.
bp.114.151910
Thompson, L., & McCabe, R. (2017). How psychiatrists recommend
treatment and its relationship with patient uptake. Health
Communication, 33, 1345–1354. doi:10.1080/10410236.2017.1350916