How to respond when patients invoke a diagnosis for themselves: Evidence from a nurse’s response practices in personality disorder interviews

ABSTRACT

What is going on when a psychiatric patient claims a psychiatric diagnosis for themselves which is different from the one a practitioner is investigating? We analyze cases from 10 interviews between psychiatric patients and a nurse using a formal interview schedule to assess whether the patient has a personality disorder. When the patient invokes (temporary) depression to explain some of their experiences or life circumstances, the nurse then has to handle that, while dispassionately pursuing an interview schedule that is, on the contrary, predicated on the diagnosis being a long-term personality disorder. We show how the nurse balances respect for the patient’s account while also performing her institutional duties. The data are in Finnish.

Somatic medicine is organized around the principle of distribution of tasks. Patients are considered to have knowledge about their symptom experiences, which they report to clinicians, who are in turn considered to have knowledge about diagnoses and treatments (Landmark et al., 2015). In our study, we focus on encounters in psychiatry, but the same basic idea holds: Clinicians interview patients about their experiences, which then leads to the clinician’s diagnostic conclusion by comparing the symptoms to the categories presented in the diagnostic manuals. However, during medical visits, patients may display their ability to recognize and weigh the evidence for common causes of their symptoms and, thus, express how they have made sense of their health problems (Gill et al., 2010). This is what can also happen in psychiatric assessment: The patients might take one step further from providing information on their symptoms to expressing their own understanding of their problems, thus, actively directing the diagnostic reasoning.

This article concerns moments when patients deploy psychiatric diagnoses through their own initiative (without being solicited by the clinician) to participate in constructing an image of their own personality during personality disorder (PD) interviews. Clinicians arguably face a challenge of balancing between two tasks: acknowledging the interactional work the patients do through diagnostic labels, on one hand, and advancing the information gathering on the other hand. In this article, we focus on one clinician, who is a nurse, facing one label, which is depression. We will analyze how she responds to two of her patients who invoke a diagnosis of depression while answering the questions of SCID-II (Structured Clinical Interview for DSM-IV Axis II Personality Disorders; First et al., 1997). Our aim is to highlight the means that the nurse utilizes in managing the task.

In psychiatric evaluation, the aims are to understand patients’ experienced problems within a psychiatric framework and to identify a suitable treatment. Usually this includes deciding on a psychiatric diagnosis. The psychiatric diagnosis can be defined as a semiotic act that offers a reinterpretation of patients’ experienced symptoms as signs of some disease. And since the diagnosis is an interpretation, it is open to different interpretations depending, for example, on the institutional setting, the participants’ cultural background, and the professional approach (Kleinman, 1988, pp. 7–8).

People invoke categories to accomplish social actions in their interactions with others (Schegloff, 2007a), including within psychiatric settings. Patients can thus invoke psychiatric labels in specific contexts to accomplish specific interactional effects, while concurrently expressing situated understandings of their own situation. In this article, we use conversation analysis (CA) as a method, because it allows us to investigate how psychiatric labels are used and responded to in the specific situation of PD interviews.

Patient uses of diagnostic labels

In our study, we found that the patients sometimes evoke depression to account for their experiences or behavior. This is in line with previous qualitative studies showing this tendency in patients’ talk regarding their diagnoses.

LaFrance (2007) investigated women’s talk about depression with discourse analytic methods and showed that by naming distress and using the pronoun “it,” the speakers objectify and construct depression as an independent entity and, in this way, separate it from their own character. Based on his participant observation fieldwork in an ADHD support group, Brinkmann (2014) showed that both patients and clinicians use the ADHD diagnosis to explain an experienced problem. This helps to reduce responsibility for the patients’ behavior. However, as Brinkmann (2014) points out, this is a circular process. First, the diagnosis is created based on the description of the patient’s symptoms, and then these same symptoms are explained by the diagnosis. In a similar vein, in their discourse analytic study of interviews with Finnish youth diagnosed with ADHD, Honkasilta et al. (2016) illustrated the typical tendency of many to attribute the responsibility for their culturally undesirable behaviors to their diagnosis. However, we must note that these studies did not address interaction between patients and clinicians in psychiatric institutions.

Closer to our research is a CA-based study about psychiatric assessment in an outpatient clinic by Weiste et al. (2018). The researchers demonstrated how patients might present a diagnosis as an external agent having an independent effect on their life. This way, patients can distance themselves from the problematic life experience. Some patients refer to their diagnosis while explaining which life circumstances had caused it, justifying their need for psychiatric care, and reducing their own responsibility for becoming ill. Patients can also express their disagreement with the diagnosis they have received—which can be seen as a way to distance their character from an unwanted diagnosis. In this study, the clinicians mostly did not affiliate with patients who used depression as the cause for their problems. Instead, they pointed out patients’ motivation for their behavior and presented them as active agents. In some cases, the clinicians did not comment on the patients’ invocations of their diagnoses.

Our study builds on this research by Weiste et al. (2018): Our research has many similarities with their setting as they also used CA to examine how patients invoke diagnoses and how clinicians respond to them. However, there are some differences between the settings. First of all, their analytic emphasis was on the patients’ use of diagnoses and, thus, they did not systematically focus on the clinicians’ responses—which is the gist of our analysis. In addition, in Weiste et al. (2018) there were two clinicians present (a psychiatrist with either a psychologist or a nurse), but in our setting nurses carried out the interviews alone. Therefore, the difference is both in training and the kinds of clinicians involved. Lastly, Weiste et al. (2018) studied the first visits of the assessment processes in which the conversation has a freer format. We focus on the structured interview format (at a later phase), which guides and restricts the discussion considerably as there is pressure in going through a long list of preset items. Therefore, we add to the current knowledge by offering a detailed analysis of a nurse’s practices of managing the assessment situation. In addition, as far as we are aware, this is the first study focusing on interaction in the context of SCID-II interviews.

Interactional research on diagnostic talk

Our focus is on the interaction between the participants, and thus, the previous knowledge from CA-based studies regarding diagnostic conversations is relevant for our work. Different interactional activities are relevant to diagnosing, especially diagnostic questioning and delivering a diagnosis. Our study concerns the phase of diagnostic questioning. Previous CA research regarding diagnostic activities has mostly focused on somatic medicine. The interesting aspect of these studies for us is the delicate balance of agency and authority between the patient and the clinician.

Regarding the reception of a diagnosis in primary health care settings, Heath (1992) and Peräkylä (2006) have shown that through the design of their accounts, the patients systematically orient to the medical authority of the doctor. However, the patients have a chance to express their agency and knowledge about their symptoms, and sometimes they do present their disagreement with the offered diagnosis. Patients’ receptivity with respect to a diagnosis can vary depending on the context. White and Stubbe (2023) showed that in surgeon-patient consultations, the patients more frequently respond to diagnosis delivery and these responses are more likely to be extended, compared to responses in primary care settings. It is thus important to study diagnostic conversations in different contexts.

Even closer to our focus come the studies about patients’ expression of agency during the phase of diagnostic questioning. Gill et al. (2010) studied patients’ participation during the medical evaluation process and showed that patients may raise candidate explanations for their symptoms and assess their possibility during the assessment. As a conclusion, Gill et al. (2010) claimed that during medical clinic visits patients do not solely convey information about their symptoms; they may also express how they have made sense of their health problems and may press doctors to interpret their problems in certain ways. It has also been shown that the degree to which the patient displays personal commitment or certainty regarding their explanation influences the pressure they place on the doctor to evaluate their explanation (Gill, 1998).

Also worth mentioning is the study by Frankel (2001), who highlighted the importance of acknowledging the patients’ self-diagnosis during medical visits. He showed how interactional problems may ensue if the doctor fails to acknowledge the patients’ own understanding of their symptoms. This is noticeable at the end of the visits when the patients might express a rejection of the diagnostic news if there is a lack of agreement or alignment between their problem statement and the doctor’s proposed solution.

All these studies are valuable background for our study, as they prove that patients may take an active role in diagnostic reasoning and it is important for clinicians to acknowledge the patient’s own understanding. However, our focus is on interaction in a psychiatric setting, which differs from somatic medicine. Regarding psychiatric context, there are a few CA studies about diagnostic conversations. This includes the study by Weiste et al. (2018), which we reviewed in the previous section. Peräkylä et al. (2022) focused on a patient’s engagement during the diagnostic process and illustrated its fluctuation at different moments depending on the interactional context. Unlike Peräkylä et al. (2022), we focus on diagnoses invoked by patients. However, in both of these studies the central point is how the current interactional situation shapes the diagnostic talk and its effects.

Voutilainen and Peräkylä (2023) studied the ways in which clinicians anticipate a PD diagnosis at the early stages of psychiatric evaluation and communicate the actual diagnosis to the patient at the end of the process using face-work strategies. In their study, the focus was on the practices utilized by the clinicians. This, however, differs from our study. The function of evoking a diagnosis is different when it originates from a clinician whose job it is to diagnose the patient than when it is arrived at by a patient. To build on this research on diagnostic activities, we offer a unique contribution by offering a focus on a nurse’s response after a patient has invoked a diagnosis.

Psychiatric evaluation

Some basic information about psychiatric practices is required to help the reader follow the rest of the article effortlessly. The PD interview is a specific procedure for differential diagnostics that may take place during the psychiatric evaluation process. Diagnostic manuals shape the content of the interview as they define diagnostic criteria for each disorder. In Western countries, two psychiatric manuals are used: ICD (International Statistical Classification of Diseases and Related Health Problems) by the World Health Organization (WHO, 2018) and the DSM (American Diagnostic and Statistical Manual of Mental Disorders), published by the American Psychiatric Association (2013).

SCID-II is a semistructured interview format that is used to assess PDs according to the DSM (2000)-IV. The results of the SCID-II interview can be then adjusted to match up with diagnostic criteria of ICD. In the DSM-IV, a PD is defined as a long-lasting way of experiencing and behaving that causes distress or problems functioning and deviates from cultural expectations. It is worth mentioning that there are now newer versions of both the ICD and DSM (and associated personality questionnaire SCID-5-pd). However, during our data collection in 2019, ICD-10 (WHO, 1992) and SCID-II were still in use in Finland. Although our data are based on the older classification, we propose that our research is meaningful in addressing general interactional phenomena during PD interviews.

In clinical settings, clinicians other than doctors (who traditionally are responsible for diagnostic work) can be trained to use diagnostic tools. This happens in the setting of our study, where the nurses have been trained to use SCID interviews (they also use SCID-I for general psychiatric symptoms). In addition to the training, the interview manual guides the interviewer. For patients it may not be clear what type of PD is assessed at each moment because it is not explicated during the assessment.

The nurses often asked the patients to fill out a self-administered questionnaire prior to the SCID-II interview. Questions in the questionnaire are in polar form. The interview starts with open-ended questions that are meant to offer an overall view of the patient’s personality. Then, it continues with questions regarding the answers in the questionnaire. The SCID-II interview thus differs from basic history taking (Heritage, 2010) as in our cases, it is used as a practice of revisiting checklist responses. The nurse typically concentrated on patients’ affirmative answers during the actual interview. In the end, a clinician can count the affirmative answers and determine whether the diagnostic criteria are met for some category (the diagnosis should still be based on the overall assessment of the patient).

We draw on previous literature on questioning in medicine, especially regarding how the questions embody certain agenda and presuppositions (Heritage, 2010). In the SCID-II interview, presuppositions are already embodied in the design of the questionnaire and then re-invoked during the interview. One notable feature of the SCID-II questions is that some of them are built in multilayered fashion (e.g., “Have you tried to hurt or kill yourself or threatened to do so?”). This means that answering them in the affirmative does not necessary clarify which presuppositions are accepted, which creates some complexity within the interview.

A crucial part of the PD assessment is to consider whether a certain experience or behavioral pattern is pervasive (has lasted at least from early adulthood). The logic is that long-lasting phenomena can be interpreted as a part of the patient’s personality. Therefore, the clinicians need to separate fluctuating mood disorders from stable aspects of personality. For example, to associate a certain trait with a PD, it cannot solely be present while the patient is depressed. The idea is that a person can have depression as a temporary symptom, which influences how they experience things while being affected by it. Personality, on the contrary, would be something more inherent to the person, a way of experiencing that persists across different emotional states.

To summarize, our research questions are, firstly, how does the nurse manage the delicate task of balancing recognition of the patients’ own situated understanding of their mental health and the institutional task of carrying out an assessment interview and, secondly, what is the structure and organization in the nurse’s follow-up questions. We propose that this exploration will offer new insight into understanding how participants in psychiatric assessment deal with psychiatric diagnoses in interaction, which relates to the epistemic terrains of the participants. Furthermore, we want to contribute to CA work on psychiatry, which has only recently begun to grow (to cite some recent studies, Bolden et al., 2019; Savander et al., 2021).

Data and Method

The data for this study were collected in a psychiatric outpatient clinic in Finland. They consist of video-recorded SCID-II interviews with 10 adult patients, male and female, and three psychiatric nurses who were all female. We have copies of the patient records, which offer us some background information about the patients.

In the clinic of our study, patients generally meet a nurse three to five times during the evaluation period. First, the nurse obtains an overall view of the patient’s situation and symptoms. Later, some patients are interviewed with SCID-II in cases of a suspected PD (this takes 1–2 meetings). Despite a nurse carrying out the diagnostic interviews, a psychiatrist is responsible for the actual diagnosis. Depending on the situation, some patients meet a psychiatrist, but others do not.

We received written informed consent to video-record the interviews from all the participants, including permission to publish pseudonymized transcripts and have copies of the patient records. Approval for the study was obtained from the ethical committee of the Hospital District of Helsinki and Uusimaa. To gain maximal information of the interactional situation, two GoPro cameras were set up in the room, one directed toward the patient and the other one toward the nurse. Only the SCID-II interviews were recorded (and two patients’ subsequent meetings with a psychiatrist) because they were the part of the evaluation process we focused on, and we did not want to disturb the process any more than necessary.

We compiled a collection of unsolicited mentions of a psychiatric diagnosis by patients to account for some aspect of their experience or behavior when answering a diagnostic question. However, these cases differed significantly. We followed established procedures in CA that led us to identify instances of nurse responses for inclusion in the collection that are homogeneous by sequential position and action. The collection features sequences in which a question is followed by a patient response invoking a diagnosis—the analytic focus being on the nurse’s turn in third position. This leaves us with a coherent set of cases in which patients invoke the same label (depression) and wherein they self-ascribe (as opposed to disavowing) that condition. As a result, we identified five cases from two different patients and one nurse.

An obvious limitation is the small collection and the inclusion of only one professional. However, these inclusion criteria offer us an internally coherent collection and, crucially, its analysis addresses the research questions we are raising. Our in-depth analyses of these cases advance our understandings of interviewing about PDs and nurse responses to patients’ agentic invocations of a diagnosis and this way pave the way for future studies.

We utilized CA as our method because it enabled us to track the turn-by-turn process of interaction. It is widely used to study naturally occurring interactions in medical visits (Maynard & Heritage, 2005). The first author of this article collected the data and also transcribed and translated the extracts. Both authors of this article were involved in the analysis and writing process. The third line for the transcripts can be found in a supplementary file online. The names and other potentially identifying information of the original extracts have been changed.

Data availability statement

Due to ethical reasons the recordings cannot be made freely available. The anonymized transcripts that support the findings of this study are available from the corresponding author upon reasonable request.

Results

In cases we examine, we will see Viivi ask questions according to SCID-II format. These questions raise particular presuppositions, which the patients manage by invoking depression in their answers. We start by illustrating this part of the interview. In Extract 1, Viivi interviews a patient, Jyri, who is a man in his 50s who has already suffered from depression for years. This extract starts with one of the open-ended questions in the beginning of SCID-II.

Extract 1: Jyri vid1 (7:00).

Viivi asks Jyri about close people in his life (lines 1–3). Jyri starts by commenting on the quantity of them (line 4). Viivi points out that Jyri’s answer is illustrative of his personality (line 5), but she does not explicate what kind of interpretations she would make based on the knowledge of someone having many close people in their lives. Then, Jyri starts answering the actual question by mentioning his family (line 6), followed by a three-second pause. After the pause, Jyri and Viivi start their turns at the same time. Jyri confirms his earlier account (line 8) and Viivi asks for confirmation for her interpretation that Jyri refers to his kids and wife. Jyri confirms this (line 10). He expands the topic by describing his loyalty to his family: he would be ready to give his life for any of his family members (lines 10–11 and 13–15). Apparently, Jyri treats this topic of family relations as morally charged. With this unprompted account he presents himself as a dedicated husband and father. Viivi acknowledges Jyri’s turn with a continuer (line 12).

When Jyri has stated his dedication for his family there is a point of possible completion (line 15), but Viivi only responds with another continuer (line 16). Another 3-second pause ensues (line 17). Jyri has mentioned earlier that he has many close people in his life, but so far he has only mentioned his family. Presumably, Viivi waits for him to come up with some more people. This is indeed what happens next: Jyri brings up his colleagues (lines 18–19 and 21). Viivi acknowledges this (line 20). Now, there is another point of a possible completion but Viivi does not take a turn so a 1-second pause ensues (line 22). Jyri thus continues his turn, but he takes a new approach in dealing with the topic.

Jyri brings up his lack of openness and unwillingness to discuss (lines 23–26), which are phenomena that could complicate close relationships. Jyri accounts for these problems with his depression. Here, Jyri is clearly answering more than the question was asking for, which could point to something problematic in his answer (Stivers & Heritage, 2001). We can make sense of this by noticing that Jyri is still dealing with the presuppositions of the original question (Heritage, 2010): “close people” as a category would refer to people with whom one shares connection and intimacy. Apparently, this is not currently the case in Jyri’s life. Hence, this addition in Jyri’s answer clarifies the issue about the nature of these relationships. His turn can be seen as a transformative answer (Stivers & Hayashi, 2010) in that it resists the agenda of the original question of only collecting a list of people; Jyri orients to the quality of those relationships by evoking his current challenges regarding them.

Extract 1 illustrates the general phenomenon in our collection: Patients do not limit themselves to factual reporting when answering the questions. They rather take an active role in shaping the inferences that can be made based on their answers. The ways in which the nurse designs questions (based on the questionnaire and the manual) matter for how the patients orient to them when answering.

But how is Viivi to answer to this turn? She is supposed to find out what Jyri’s personality is like. Now Jyri explains his problems with depression, which is a mood disorder. A key task in the SCID-II is to separate out personality from transitory mood disorders. Hence, when patients invoke depression, this creates a problem for the nurse concerning how to interpret their answers.

Overall, we will see how the distinction between depression and personality is a joint accomplishment. An orientation to this separation is already in some way present in the patients’ answers, but it can be implicit. Across the cases, Viivi expands the question-answer sequence with a third positioned solicitation for more information (Schegloff, 2007b, on sequence expansion). The clarity and certainty of the patient’s answer arguably guide the design of the Viivi’s follow-up questions. We start with two cases wherein only a small clarification seems sufficient for Viivi to ensure the distinction. Then, we move on to two cases in which Viivi expresses explicit validation for the depression diagnosis before asking the patients to evaluate their personality without it. Lastly, we present a single case in which Viivi does not offer any validation but requests the answer very directly.

Viivi never questions or challenges the patients’ ways of accounting for their experiences with depression. However, she does take a stance on the use of the diagnosis, and this is enabled through the use of membership categorization (Sacks, 1989). The patient is treated as “a person suffering from depression,” and this makes it possible for Viivi to make inferences about the patients’ situation based on her knowledge about the depression category.

From now on, our focus will be on Viivi’s responses. Extract 1 exemplified in detail the progression of talk that led the patient to evoke a depression diagnosis. We will shortly analyze how this happens in each subsequent case, but our analysis after the extracts will zoom in on how Viivi deals with these situations.

Viivi draws an inference about personality and invites confirmation

We now present two extracts wherein Viivi manages to easily accomplish her assessment task with a simple clarification. Jyri has already answered in such a way as to support this inference, which Viivi merely works to confirm. They jointly establish depression as something that supports differential diagnosis of PD by excluding that the experiences can be linked to a stable personality trait.

In Extract 2, we show how Viivi deals with the case above. After Jyri’s account (lines 23–26), Viivi knows he connects his relational challenges to his depression, and she can utilize this knowledge in her follow-up question.

Extract 2: Jyri, continuation from Extract 1 (lines 23-26 repeated).

Viivi approaches Jyri’s mention of his depression from the point of view of PD assessment—she orients to the personality description. Time reference “anymore” in Jyri’s turn (line 26) implies that he used to be open, before his depression. Jyri has also presented his view with a degree of certainty. This context makes it possible for Viivi to only ask for a confirmation that Jyri used to be more open. She does this through a candidate answer (lines 27–28), which is a useful method when the question recipient is wanted to respond with particular information and/or offer the information efficiently (Pomerantz, 1988). In her question design Viivi implicitly validates Jyri’s idea that lack of openness can be explained with depression. Jyri’s confirmatory response (line 29) is sufficient for Viivi, who shortly acknowledges the answer (line 30) and then moves on to the next question (not shown).

We find it notable that the topic of problems in close relationships is raised by Jyri. Viivi joins in to ask for a confirmation of her interpretation even though this discussion does not answer the original question (about naming Jyri’s close people). Therefore, Viivi’s follow-up question is adapted to Jyri’s telling: She now seeks information about his openness, which the original question was not addressing. This might be explained with the context of an open-ended question which aims to form an overall view of Jyri. Other than strictly the original interview questions may also help in reaching this aim.

Next, Extract 3 is similar to Extract 2 in that Viivi orients to the task by asking for a confirmation. In the beginning, Viivi raises a question that Jyri had responded to in the affirmative in the self-administered questionnaire (lines 1–5). This question belongs to the category of borderline PD. As in Extract 2, Jyri clarifies that these symptoms started only after his depression, which implies that he has not had them all along (lines 6–7). He seems to account for the experience completely with depression. Jyri uses the timeframe “nowadays” in his answer (line 6), which could embody his understanding that the interview is concerned with stable traits. With this timeframe, Jyri indicates that he does not consider this experience as describing himself as a person but rather as a change brought on by depression.

Extract 3: Jyri vid4 (5:44).

When Jyri mentions depression in his response, Viivi takes a turn with an overlap (lines 8–9). In both Extracts 2 and 3, the time references Jyri uses (“anymore,” “nowadays”) offer Viivi a clear hint about how to interpret his responses. Viivi works to confirm what Jyri has already implied and only clarifies that Jyri does not experience this way without depression. She does this in a straightforward manner, without any explicit alignment with Jyri’s explanation. However, her turn shows an implicit acceptance of accounting for this complex phenomenon with depression label.

Viivi’s clarification statement (lines 8–9) makes confirmation or disconfirmation relevant. Jyri confirms Viivi’s interpretation, disclaiming strongly having this tendency in his personality (line 10). Jyri’s confirmatory response seems sufficient for Viivi as she does not pursue any elaboration on the matter. Viivi’s third-position receipt (“yes”) and assessment (“good”) closes off the local sequence by accepting the answer (line 11). In this vein, both Jyri and Viivi maintain a focus on questions about symptoms of PD, and invocation of depression is jointly treated and understood as accounting for the onset of symptoms that are not indicative of personality traits.

Despite the similarity of Extracts 2 and 3, they also differ from each other regarding Viivi’s actions. In Extract 2, she used a candidate answer for her confirmation check. In Extract 3, she uses a repeat (“depression”) and a negative formulation (“so it is not”). This way she articulates an inference, presenting it for confirmation (Heritage & Watson, 1979, on upshot formulations). These two formats nevertheless achieve the same outcome: seeking explicit confirmation of something that the patient has implied.

Viivi validates the invocation of depression and pursues information about personality

In the next two extracts, Viivi first explicitly validates the use of depression diagnosis. Afterward she pursues an answer better suited to the PD assessment.

In Extract 4, Viivi interviews another patient, Sara, about depressive PD. Sara is a woman in her 30s, also with a history of depression. Viivi points out that Sara has marked both “no” and “yes” to the question regarding despising oneself (lines 1–2). This would make it relevant for Sara give an account for this ambivalence and elaborate on the content of the affirmative answer. However, Sara does not straightforwardly do either. Instead, she evokes depression to explain the experience (lines 4–6).

Overall, there are a lot of tentative aspects to Sara’s answer. Her ambivalent answer in the questionnaire already refers to some complexity in her approach to the matter. Now, in the interview, Sara starts her response with a prolonged “yea:h” (line 4), as if she is still pondering about the matter. She then continues by saying that she has now started to think about this (line 4), framing her answer as revisiting the questionnaire answer. With this statement, Sara suggests a recent realization and something she has not yet reached clarity on. In addition, she uses “probably” before her evaluation (line 5) to express uncertainty about the matter. There is thus no clear indication that the symptom was absent before the onset of depression.

Extract 4: Sara vid1 12.50.

After Viivi has heard that Sara has brought up depression, Viivi takes a turn before Sara has reached a point of possible completion (line 7). At the start of Viivi’s turn, the point about depression is already available to her. Therefore, she can project what Sara is about to convey. Viivi’s first description “it relates to depression that one sees oneself” is left unfinished but with her next formulation “self-esteem decreases and things like that,” Viivi explicitly validates Sara’s claim (lines 8–9). She agrees with Sara that depression causes negative view on one’s value. Viivi’s statement is based on the logic that despising oneself is category-bound activity related to depression (Sacks, 1995).

Sara has framed this matter as something that she is wondering about. This can be heard as requesting information, as it is something that may lie in the epistemic domain of the nurse’s expertise. Viivi’s explicit validation in this case thus can result from the framing of the answer. She transforms the phrase “despising oneself” from the original question into a more neutral description “self-esteem decreases.” This is likely related to the sensitive topic they are dealing with—with the term transformation Viivi can tone down the harshness of “often despise yourself” and thus create a more neutral space for their discussion.

Concurrently, Viivi treats responding to the depression element as having the potential to start a trajectory of talk on it, which would divert from the agenda of the interview. She prevents this by retaining the turn with an in-breath at a transition-relevant place (line 9) and then quickly shifts the talk to meet the requirements of PD assessment (lines 9–12), thus not providing any more space for elaborative talk about depression. Viivi opts to reformulate the question entirely with a third-turn repair (Schegloff, 1997; from line 9), which fully unpacks the terms of the question, spelling out what exactly is required (specifically, leaving depression aside and evaluating oneself without it). Viivi uses two descriptions, “normal state” and “basic character” (lines 9–10) to emphasize the idea of evaluation that is not affected by depressed moods.

We argue that Viivi’s approach is related to the tentativeness of Sara’s answer. There were complexities both in the questionnaire answer (the reason for Sara marking both “no” and “yes” remains unknown) and her response in the interview. Thus, it seems reasonable to assume that the actions providing for confirmation of a state of affairs (which Viivi uses in Extracts 2 and 3) are not a good fit here—Sara has raised doubts about the impact of depression as opposed to making a statement about it. While Viivi proposes that the link to depression is sensible; she redirects Sara to self-assess by considering periods of time when she was not depressed. This reflexively suggests that Sara’s answer is treated as doing more than answering the question (just as Jyri does in Extracts 2 and 3). Viivi both validates Sara’s reflections on the effects of depression and concurrently establishes it as a matter that cannot be addressed here and now.

Viivi transforms the question from the original (lines 10–12). The questionnaire question was about despising oneself, but now Viivi asks about “blaming critical and derogatory attitude.” These qualities originate from the interview manual that includes instructions for the interviewer on how to interpret the patient’s answer. The fact that Viivi resorts to this guidance in the manual refers to the possibility that she is looking for a way to elaborate on the original question to Sara. Sara offers a negative response (line 13). Viivi comes in quickly to confirm Sara’s answer (line 14) and moves on to the next question (not shown). She clearly orients only to receive the distinction between Sara’s attitude in depressed and normal states—no elaboration on the matter is needed.

In Extract 5, we have another case in which Viivi offers some validation for the patient’s depression explanation. Viivi and Jyri are now going through questions regarding avoidant PD. A central part of this personality type is timidity with other people. One aspect of this timidity can be lack of self-worth when comparing oneself to others. In the self-administered questionnaire, Jyri answered this question affirmatively: “Do you believe that you’re not as good, as smart, or as attractive as most other people?” (Viivi refers to this in lines 1–4).

Viivi asks Jyri to elaborate on his answer (line 5). As Jyri does not start answering and passes the opportunity to provide details (emerging silence in line 6), Viivi continues (lines 7–9). She starts to ask a question about the origins of Jyri’s negative self-evaluations but abandons it in favor of a question about their duration. Jyri matches his answer (lines 10 and 12) with the structure of the last question (“has it been always,” “it has not been always”) to exclude that he has always had those thoughts. He once again links their origin to the onset of depression (on temporal references, see Raymond & White, 2017). The time reference Jyri uses here is in line with the agenda of the interview that attempts to establish whether a symptom is indicative of stable traits. This enables Jyri to address the question about time while concurrently accounting for his experienced lack of self-worth. Overlapping with Jyri’s turn, Viivi offers continuers (lines 11 and 13).

Extract 5: Jyri vid2 (2:40).

After Jyri has evoked depression in his response, Viivi offers affiliation. She even brings up a broader concept of “sense of worthlessness,” which was not a part of the original question, to upgrade the evaluation (lines 13–15). Despite the assessment task, Viivi at this point relaxes the agenda of the questioning and aligns Jyri’s approach to the diagnosis, supporting the view of depression causing Jyri’s lack of self-worth.

Jyri uses “probably” in his response (line 10), just like Sara did in the Extract 4, and thus they both present their answer with some uncertainty (in Finnish these words are different, “varmaa” and “kai,” but they are synonyms). Jyri’s previous nonresponse (line 6) seems to further evidence his uncertainty about the matter. A statement of lack of certainty can be treated as making offer of information relevant in a professional context, if the matter falls within the professional’s expertise. In Extracts 4 and 5, Viivi seems to depart from the task of gathering information to provide information that may help to resolve some of the patients’ conveyed doubts.

In her turn, Viivi uses Finnish clitic -han attached to the word depression (masennukseehan, line 14). There is no direct translation for this; it rather gives a special nuance to the message. In this case -han refers to a known fact. This message is added to the translation “as is known” in double brackets. With her turn and especially through the use of this clitic Viivi expresses some epistemic authority (Heritage & Raymond, 2005)—as a health-care professional she is qualified to assess Jyri’s statement as correct.

Jyri’s starts his next turn, overlapping with Viivi (line 16). At this point he has heard Viivi saying that something is strongly linked to depression, which means that Viivi’s turn is not yet complete. (Due to the different word order in Finnish the turn starts with “to depression” and ends with “sense of worthlessness.” When Jyri starts his turn, the part “sense of worthlessness” is thus not yet available to him.) There is a possibility that Jyri already projects that Viivi will complete her sentence with a symptom that he has reported, and this encourages him to further elaborate on the matter. However, another option is that Jyri actually disattends from mutual exchange at this point, and simply continues his own line of thought. His gaze is elsewhere and the link between his next turn and what Viivi has said is not completely clear. Overlapping with Viivi’s turn would also speak in favor of this option.

Jyri’s turn about exploring himself remains cryptic (lines 16–17). During a pause Viivi makes some notes (line 18), after which she acknowledges Jyri’s response and elaborates further on the effects of depression: “one accuses and incriminates oneself” (lines 19–22). She thus seems to arrive at treating Jyri’s turn of exploring himself as referring to self-critical thoughts and then offers some more validation for the depression experience. It is noteworthy that Viivi expands the topic with even more attributes than what was part of the original question (lines 13–15). As in Extract 5, Viivi utilizes membership categorization of depression label. She relates accusing and incriminating oneself to depression as category-bound actions (Sacks, 1995); these actions do not appear explicitly in the original question nor in Jyri’s answer, and thus Viivi’s inference is grounded in the understanding of depression as a category.

Jyri recognizes he has this experience “all the time” (line 23). Viivi shortly receipts his turn (line 24), after which Jyri further reinforces the conclusion (line 25). Viivi still adds some validation for the experience by formulating again the point about feeling worthless (line 26). Jyri continues to a story about a difficult situation at work (from line 27). He had to leave from work quickly and explain this to his boss. On the way to the interview, the boss called Jyri, thinking that he was still at work. Viivi treats the story as exemplifying the earlier points Jyri made about his depression. She proposes that Jyri felt guilty after the episode (lines 40–41), which Jyri confirms (line 42). This enables Viivi to treat Jyri’s feelings of guilt as a symptom of depression.

At this point, Viivi circles back to the task of confirming what he was like before depression (lines 43–45). She seeks confirmation that Jyri used to accuse himself less before it. “Accusing oneself” (line 45) is Viivi’s interpretation of what Jyri’s story was about. It is also continuation from the lines 21–22 where Viivi brought it up to the conversation. We thus notice that the question has transformed from the original version in the questionnaire, which involved beliefs about being less good, smart, or attractive than others. Viivi seems to have adapted her follow-up question to the current situation in their conversation. The difference from case 4 is presumably related to the distance between the original question and the follow-up question. In Extract 4, Viivi’s asks the follow-up question right after Sara’s response, in an actual third-turn position. In Extract 5, the participants end up in a quite long exchange before Viivi finally manages to get back to questioning, and this arguably makes her adapt her formulation to the moment at hand. Jyri confirms Viivi’s candidate answer (lines 46 and 48). Viivi briefly receives Jyri’s response (line 47). Her use of “okay” as part of her response seems to work here as a closure of a longer sequence of turns, projecting a change of topic (Gardner, 2001, p. 16).

We propose that this extract is comparable to Extract 4 regarding Viivi’s attempts. In both cases she receives the patient’s tentative response by validating the link between depression and the symptom. She then aims to clarify whether the patient experiences the same way outside of depression. With Sara (Extract 4) she manages to immediately move to the follow-up question. In Extract 5, it is arguable that Viivi would have gone on to request a self-assessment outside depression as well, but Jyri enters in overlap (line 16). Therefore, Viivi’s possibility to continue as she did with Sara is interdicted by Jyri’s overlapping turn expansion. Viivi cannot shift the focus back to the requirements of the questionnaire, as her validations (lines 13–15, 19–22, and 26) repeatedly prompt additional elaborations from Jyri. Eventually, Viivi nevertheless manages to move on to her question about how Jyri would evaluate himself before depression (lines 43–45). Jyri’s turns seem to defer Viivi’s project.

Viivi pursues information about personality without validation

The last extract differs from the earlier ones in its directness. Viivi does not offer any explicit validation for the depression label, and her follow-up question requests straightforwardly to leave it out of the evaluation.

In Extract 6, Viivi asks Sara to elaborate on her questionnaire answer by providing concrete details of self-harm (lines 1–9). Sara had answered affirmatively to the following question: “Have you tried to hurt or kill yourself or threatened to do so?” Self-destructive thoughts and acts are one criterion for borderline PD. Sara denies having engaged in actual self-destructive acts (line 10) and, next, instead of describing what she has done, Sara brings up her depression (lines 11–15). By stating that the depression “pulls her very deep,” she uses the diagnostic label as an account for some conduct related to self-harm. Despite Sara not offering details about her actions (perhaps does not have space for this due to Viivi’s overlap), it is notable that the depression label in this context has implications regarding differential diagnostics—that is, Sara’s self-destructiveness would not be a part of her personality but rather a symptom of her depression.

Extract 6: Sara vid2 (12:54).

After Sara has referred to depression and glossed over the details related to self-harming, Viivi interjects at a point where Sara has projected continuation (“but”). Viivi’s turn (lines 16 and 18) thus seems designed to halt progression of Sara’s turn to pursue the matter of threats. This happens despite Sara’s answer could be regarded as being relevant in assessing how much this behavior is descriptive of Sara’s personality. Viivi treats Sara’s “yes” answer as implying that there is something to account for. Inferentially, if they are not acts, perhaps her response refers to threats or ideation. Viivi goes on to pursue detailing of Sara’s conduct, thus orienting to the agenda of the original question and sidestepping Sara’s account. Her question can be seen as a third position repair.

Viivi’s follow up question (lines 16 and 18) seems to be designed to both distinguish threats from acts and working out whether these things are present independent of depression. Viivi starts with repeating a part of the original question “threatening with self-destruction” but she also adds a new element to the question: “toying with it.” We cannot say for certain whether it refers to having thoughts of self-destructive acts repeatedly in one’s mind or playing with the threats in social relationships. This new element in Viivi’s question is nevertheless interesting. “Toying” with self-destruction creates a new nuance to the question; it sounds less serious than threatening with it, but at the same time it emphasizes some form of irresponsibility and immaturity.

The reason for Viivi not expressing any validation for Sara’s account, as she did in Extracts 4 and 5, is not clear to us. However, we notice that this seems to be the only case wherein a patient does not fully answer the question. Sara’s answer targeted the last question posed at lines 8–9 (principle of contiguity; Sacks, 1987). She is not offered a chance to continue her response. Viivi clearly treats Sara’s response as incomplete because she did not respond to the topic of threats to self-harm (lines 7–8). We assume that this context orients Viivi more toward clarification rather than validation of depression. It is easily understandable why Viivi does not use confirmation checks in this extract. The original question in the questionnaire was not straightforward to begin with, asking more than one thing at the same time, and then Sara’s answer introduced an additional level of complexity. Sara’s answer thus does not support the simple inference that Viivi could use in Extracts 2 and 3.

Sara admits that there have been some threats but downgrades the certainty of this (“threatening maybe” in line 20) and does not provide further details. She makes an intriguing choice by calling her behavior attention seeking (line 20). Viivi affiliates with Sara’s turn, showing her view of a borderline PD patient as someone who uses self-destructive threats for attention seeking (lines 22–23) and continues exploring the subject by asking how much this has happened (line 24). This is the only case in these extracts in which Viivi’s questioning seems to identify a possible symptom of PD. In other cases PD was excluded, as the problem at hand was attributed to depression.

Discussion

In this paper, we have examined a nurse’s practices for pursuing an information-gathering task while concurrently acknowledging the patients’ own understanding of their situation when they invoke the diagnostic label of depression. The questions of SCID-II are often complex, raising multilayered presuppositions. They do not convey a very clear idea about the nature of PDs and how the experiences are expected to differ from other phenomena, such as depression. In Extract 4, for example, the participants attempted to separate depression and depressive PD from each other while the symptom (despising oneself) would fit with both options. We have shown how the participants embody these complexities during the interviews. One way for the patients to manage the presuppositions of the questions is to invoke depression and, in this way, take an active role in shaping the inferences that can be made based on their answers.

During the interviews, Viivi raises the questionnaire answers to pursue some elaboration on them while aiming to infer whether the patient’s answer points to a symptom of a PD. When the patients refer to some problem being related to depression, this implies they are not treating it as part of their personality; in this way, they fulfill the agenda of the question. However, in the cases we have examined, it is noteworthy that the patients already take one step further from factual reporting (describing their symptoms) and participate in diagnostic reasoning (proposing the reason for their symptoms). The patients’ answers thus embody some epistemic authority. This can be suitable for the interview as long as the participants manage to set personality traits and depression apart.

However, the patients’ agentic approach can create some complexities for Viivi, for example, when the patients do not clearly state but rather suggest the possibility that their symptoms are linked to depression. This establishes the relevance of exploring the matter more in the interview to clarify what is what. However, this also creates the potential for starting a conversation of depression at length, which is beyond the scope of the interview.

Across the cases, Viivi keeps her primary focus on progressing with the assessment. In this institutional interview setting, there still seems to be some flexibility in accomplishing this task. Viivi’s different choices of treating the patients’ responses are fitted to their specific qualities, especially the level of clarity and certainty in the patient’s response. This way, the follow-up questions embody recipient design (Heritage, 2010). Viivi balances her information-gathering orientation with recognition of the patients’ expressions of their own understanding of their situation. We can conclude that the latter is a secondary (but still significant) task in the setting.

It is noteworthy that Viivi never challenges the patients’ ways of explaining their experiences with depression. This means that she does not encourage the patients to ponder their experiences outside the diagnostic labels. It seems that expressing acceptance for depression as an explanation helps Viivi to move on with the assessment without conflicting with the patients. Validating the patients’ own views of their situation can thus be considered a means of interviewing: It facilitates a faster progression compared to having actual conversations about the topics that the patients bring up.

In the study of Weiste et al. (2018), clinicians could sidestep patients’ use of diagnoses for accounting for their conduct and highlight patients’ motivations and responsibility for their actions instead. This way, the agency offered to the patients was closer to being an active agent with different options to choose from in their lives. There is thus a difference between our findings and Weiste et al. (2018). In our data, the patients’ agency came to life in the statements in which they defined depression as the reason for their experiences.

This difference in the findings might be related to the structured interview guiding the interaction in our study. In the study of Weiste et al. (2018), the conversation had a less rigid format, while in our study there was pressure to go through a list of preset items. The training of clinicians might also have some effect. Future research is needed to compare how clinicians respond across different settings. We find it worthwhile for clinicians to be aware that even during these interview situations, they might have an effect on the patients’ approach to their problems; The clinicians’ follow-up questions and formulations can reinforce the patients’ own interpretations or open up new perspectives.

In our study, Viivi filters out the relevant aspects of the patients’ responses that are meaningful for the SCID-II assessment. This observation resonates with Antaki et al. (2005) regarding therapists’ formulations in psychotherapy. When the psychotherapy clients reply to questions about their history or symptoms, the therapists may treat their turns as incomplete or deficient and thus pursue clarification. In their formulations, the therapists use diagnostically relevant versions of what their clients say to advance the therapeutic or institutional agenda.

Overall, the patients’ self-initiated invocations of diagnoses were rare in our data. However, it is interesting to see that diagnoses were volunteered in response to questions about everyday conduct and experience. The patients seemed to claim their epistemic rights not only to their experienced symptoms but also to their knowledge of (and associated rights to invoke) diagnostic categories.

One limitation in this study is the small sample size, which limits the transferability of our findings. We hope to further inspire the continuation of sociological study on practices of psychiatric interviewing. After the strong tradition in the research of medical interviewing, interaction in psychiatry is worth investigating in its own right.

Acknowledgments

We wish to thank all the patients and clinicians who agreed to be recorded. We are grateful to three anonymous reviewers for their very helpful comments on an earlier version of this manuscript. Our gratitude also goes to Federica Ranzani for her insights in several data sessions in the early stages of this research. The first author would like to thank Martin Hildebrand and Antti Paakkari for their generous support during the process.

Disclosure statement

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

Supplementary Material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/08351813.2023.2235958.

Additional information

Funding

This work was supported by Kone Foundation under grant 201802533. The Foundation was not involved in the research process.