Communication in high risk ante-natal consultations: a direct observational study of interactions between patients and obstetricians

Data was collected between June and November 2016. Consent from clinic staff to video some of their consultations was obtained ahead of time. Staff identified patients attending the clinic for the first time and a female researcher approached them in the waiting room to seek their informed consent (full briefing and written consent was then carried out in a private room). Where consent was granted, a single small camera on a slim tripod was discreetly set up in a corner of the consultation room so that the video captured all participants’ faces and most of their bodies. A small audio recorder was also placed on the desk as a backup. This equipment was set up and turned on at the beginning of the consultation by the researcher who then left the room. Interviews after the consultation were audio-recorded. The patient and any accompanying adults (often the woman’s partner) were asked for a brief interview in a consultation room immediately following the consultation. Obstetricians were interviewed at the end of their clinic, or by email if they preferred. Medical notes for the consultation were obtained where consent was given to support analysis and interpretation of the consultation recordings and interviews. The field researcher wrote ethnographic field notes to provide additional background information. Participants were also asked for consent for their data to be added to a permanent corpus of health interactions for potential future ethically approved research.

The recordings of all consultations and interviews were fully transcribed and anonymised by the removal of identifying details such as names of people and places. A log of the activities in each consultation was created and proofed by a clinical member of the team. The inductive analysis followed an iterative process which alternated between individual analysis by the main field researcher (a discourse analyst and the team member with the most intimate knowledge of all the data) and group data sessions with the wider multi-disciplinary team of clinicians and applied linguists. The initial individual analysis combined a thematic analysis of the interview data (using NVivo software) in tandem with a structural discourse analysis of the consultations which identified macro-level features of the interactions including length of consultation, overall structure (including the RRS), how openings were managed and reasons for referral. This was supported by ethnographic information (field notes and medical notes). Validation (or otherwise) of these initial analyses was provided by the multi-disciplinary team which comprised an obstetrician, and general practitioner and a nurse and two non-clinical health communication researchers (experienced in interaction analysis and ethnography of communication). These data analysis sessions involved viewing and analysing video and transcript data against the analysis, and included attention to non-verbal features of the interaction including gaze direction, facial expressions, nods, body positioning and activities during silences. Later iterations of this two-stage process included searching for interactional evidence (or counter-evidence) in the consultation data for each of the themes emerging from the interview data. Any disagreements within the research team were resolved through discussion and consensus. Several further cycles of individual analysis and group data sessions resulted in a set of initial findings which were presented back to the clinical participants in the research, and the rest of the clinic staff. This provided an opportunity for further clinical feedback on our methods and findings, which then led to further revision and refinement.

All analysis deliberately followed an appreciative enquiry approach [50‐52]. The focus was thus principally on identifying positive features of the interaction as well as locating interactional evidence for the themes from the interviews which represented the participants’ perpectives. At the same time, other relevant features that came to light were noted and taken into account in the analysis and interpretation of the data.

Patients were not involved in the design or analysis of the study.

Thirty two patients were approached to participate, of whom 7 declined. A further 5 patient consultations were not recorded for logistic reasons. Recordings were made of 20 patients in consultations, 11 of whom were accompanied by a partner or other family member (18 video and 2 audio only). One consultation was only partially recorded at the doctor’s request. Sixteen of the 20 patients and those accompanying them were interviewed (the remaining patients declined due to time constraints). Two patients declined consent for their medical notes to be obtained. Thirteen obstetricians were filmed across the 20 consultations, all of whom were interviewed afterwards (one via email).

The patients were between 20 and 39 years old, with more in the older age bracket (12/20 aged 30–39); half were of European background (most of whom were New Zealand born) and most were well-educated. More than half (13/20) had given birth before. (See Table 1).

The 13 obstetricians in the study included 6 consultants (11 consultations recorded) and 7 registrars (9 consultations recorded). Most were female (9/13) including all of the registrars, and included overseas and locally trained doctors with a range of experience.

As background to and context for the analysis of the quality of the communication that follows, we initially report on several descriptive features of the consultations: length, structure, reasons for referral and the structure of the consultation openings.

The length of consultations ranged from about 15 min to nearly an hour. Of the 19 full length consultations recorded, 13 were 15–30 min in length and six were 30–60 min. The average length of consultation was 28 min, with registrars being more likely to have longer consultations.

Eleven of the consultations began with a discussion of the main referral issues while six began with the medical history (three had an unclear structure). There was a tendency for the more experienced obstetricians to deal with the main referral issue first and to back-fill the medical history later in the consultation, while registrars most often began with the medical history.

Most of the doctors (17) read the patient notes prior to beginning the consultation, although for three of the patients who saw a consultant, the notes were read during the consultation.

The reasons for referral to the high-risk antenatal clinic were varied, and often there was more than one reason for referral. The most common reasons were previous caesarean delivery - often in combination with other reasons (7), large or small for dates (2), twins (2) and bleeding (3). Other reasons for referral were: recurrent miscarriage; high BMI³; threatened pre-term labour; previous infant deaths; hip dysplasia; recurrent herpes; low platelets; episode of dizziness/shortness of breath; low thyroid levels; schizophrenia management; malignant hyperthermia; latent tuberculosis; and previous HLH⁴ infant deaths.

A clear referral recognition sequence (RRS) was identified in 18 of the 20 consultations.⁵ Table 2 shows the patterns for referral recognition sequence (RRS) that were found.

In half of the consultations, the RRS was explicit and mentioned both who had made the referral and why, although in nearly half of these, there was some delay in relaying the reason for referral. This delay meant that patients were unclear about some aspects of the referral for a time and could result in some discomfort where patients explicitly asked for the reason for their referral (as in the example for RRS Type 1b), or in other cases showed increased anxiety. This happened when patient notes were being read during the consultation, which was more common among the more experienced consultants. Apart from this tendency, there were no other clear differences between consultants and registrars in the use of RRS patterns.

Analysis of participant interviews and of the consultation recordings highlighted a number of features that contributed to a sense of high quality communication and of patient satisfaction. Patient interview responses were almost universally positive, with all but one of the patients expressing satisfaction overall with the communication in the consultations. Comments on communication ranged from “good” (AN-SP34R-01 PT interview), to “he’s very good with his communication skills” (AN-SP29–01 PT interview), through to “amazing…. I think the best doctor I’ve seen.” (AN-SP36R-01 PT interview).

The following more specific positive features of the communication emerged from the mixed methods analysis. Tables 3 and 4 contain illustrative quotes from the interview data and excerpts from transcripts of the actual consultations that are referred to below.

Nearly every patient (and/or accompanying adult) who was interviewed (15/16) mentioned this in their evaluations of the communication, as shown in the selection of quotes in Table 3. One patient appreciated the way in which the thought processes of the doctor were made visible (Quote 1.1), while another specifically acknowledged the difficulty of good communication with someone you have not met before (Quote 1.2). Patients also explicitly mentioned thoroughness or described a thorough approach, as in Quotes 1.3 and 1.4. Within the consultations, this was evidenced by the coverage of multiple topics and the detailed discussions observed (the length of which make it impractical to reproduce an example here). An example of clear and informative communication is given in Consultation Excerpt 1.1 in which the registrar goes to some length to explain the reason for the extra concern with a baby that is small for dates, using an extended metaphor to make the point.

Informativeness and clarity was enhanced by clearly structured communication that patients reported experiencing in their consultations, as evidenced in Quotes 2.1 and 2.2.

Clear signalling of topics and agendas was directly observed in many of the consultations; this made the structure more obvious, as seen in Consultation Excerpts 2.1 and 2.2 in which the doctor first explicitly indicates they are about to deliver an ‘informing’, then numbers off the topics to be discussed in the consultation as a way of signposting the stages of the explanation.

While the question was not always specifically asked in the semi-structured interview (depending on how the conversation went), when 12 patients were explicitly asked if they felt they were listened to, all responded positively for example, in Quote 3.1. One patient in particular mentioned feeling that their anxiety had been legitimised (Quote 3.2).

Within the consultation, several doctors explicitly acknowledged the worry or anxiety patients may feel, as shown in Consultation Excerpt 3.1, and also in excerpt 6.2 (lines 13–20). In excerpt 6.2 (lines 15–16), the clinician explicitly claims to understand that the patient is and will be ‘anxious’ and ‘stressed out’, using informal language (such as ‘gonna’ and ‘aren’t ya’) that serve to minimise the social distance between clinician and patient, and quiet talk to index the sensitivity of this. The clinician also phrases their statement so that patient agreement is the ‘preferred response’ [53] by using a tag question (‘aren’t ya’). The patient does indeed agree (line 17).

A good example of a patient’s previous experience being acknowledged with empathy is shown in Consultation Excerpt 3.2, where in addition to the minimal responses (such as “yeah”), the doctor provides brief but effective acknowledgement of the impact of the experience on the patient at line 6 with a simple “wow”, and with an explicitly empathic statement at line 15–16.

Many patients (9/16) specifically mentioned in interviews that they felt comfortable to seek further information or explanation, as seen in Quotes 4.1 and 4.2. One illustration is seen in Consultation Excerpt 4.1 where the partner of the patient spontaneously asked the doctor to provide more information, which was responded to at length.

In Consultation Excerpt 4.2, the consultant has been giving information at some length, and it is noteworthy that the patient, at line 6, begins an assertive bid to participate (“so”, just before the consultant has finished speaking. The patient here successfully gains the floor at line 8 and asks their question. This illustrates that even in more challenging interactional contexts such as this, where a specialist is engaged in an extended informing sequence (which patients typically do not interrupt), the patient here indeed had a level of comfort with active participation. The doctor, while completing their turn in the face of the patient’s bid for a turn, then gives the floor to the patient, maintaining mutual gaze and nodding as a ‘go-ahead’.

Five of the patients specifically mentioned the fact that they were given options and that they felt their wishes were sought and respected, as seen in Quotes 5.1 and 5.2. Again there was evidence of this occurring in the consultations. Consultation Excerpt 5.1 illustrates an obstetrician explicitly telling patients that they aim to inform them of their options. Consultation Excerpt 5.2 is from a Vaginal Birth After Caesarean (VBAC) clinic consultation with a non-native speaker of English. The clinician makes it clear that patients’ preferences for VBAC or caesarean will be considered, quickly checking on whether the patient has received information on the options. At line 5, the clinician asks a completely open question that is not tilted towards either of the available options. The patient expresses her preference for a caesarean section and the clinician explicitly affirms the importance of the patient’s preferences in lines 16–19, after noting the need for clinical assessments (lines 11–14).

In another consultation (for which no patient interview was conducted), there was a little more negotiation as to whose wishes might prevail (see Consultation Excerpt 5.3). When the obstetrician expressed an opinion that was at odds with the patient’s preferences (lines 4–5), there was push-back from the patient in line 6. The obstetrician went on to explain the risks if a herpes lesion was present in labour and that the patient may not be aware of a lesion, adding:

“as long as you're aware of that situation then you'd be better informed to make that decision, that's number one”

While emphasising the importance of medication and extra scans regarding small gestational size, the obstetrician also acknowledged that the patient had a “fair point” on several occasions during the consultation, which explicitly validated the patient’s perspective. By presenting information and options and acknowledging the patient’s perspective, even when being challenged, the obstetrician succeeded in keeping the interaction on positive terms and negotiations friendly and respectful.

Two patients particularly appreciated straightforward and realistic communication from the doctors (Quotes 6.1 and 6.2). Other patients reported positively on consultations in which open and realistic talk was observed, such as Consultation Excerpt 6.1 in which a registrar comments on the inherent uncertainty in this setting. Consultation Excerpt 6.2 is another example of plain talking that appeared to be appreciated. In line 5, the doctor uses very direct, colloquial language to talk about the possibility of the uterus ‘pulling apart’, albeit softening the words by lowering the volume of talk.

Several patients mentioned the way in which doctors made them feel relaxed and comfortable, as in Quotes 7.1 and 7.2.

Simple things like handshakes and small talk that may elicit laughter contributed to building rapport and making patients and those accompanying them feel welcome and comfortable. A simple example is shown in Consultation Excerpt 7.1 where the patient’s partner, who arrived late, was explicitly made welcome with introductions and small talk that elicited laughter.

In addition to the opportunities patients had to ask questions and participate, patients and those accompanying them were also able to display their knowledge, with many of them having experienced childbirth before. Such patients spontaneously used clinically appropriate technical terms such as “placenta praevia” (AN-SP31–02) or “breech” (AN-SP36R-01). The patient in Consultation Excerpt 4.2 (discussed above) displayed her confidence in her knowledge with an interruption (line 6) that treats the doctor’s partial utterance (‘if it’s not going well’) as sufficient (indicating that she doesn’t need to hear the rest of the explanation). Her following turn (lines 8–9) further displays her understanding by her use of the term ‘induction’ in a way that links back to the doctor’s discussion of triggering labour by breaking waters (i.e. displaying her understanding that this is a form of induction).

In interviews, the obstetricians explicitly valued patients being well-informed, especially in view of the limited consultation time available (see Quotes 8.1–8.3), and were observed in consultations giving patients opportunities to display their knowledge, thus also ascertaining their current understanding. Even those without previous childbirth experience were given opportunities to display recently acquired knowledge, as shown in Consultation Excerpt 8.1 in which the patient is a young first time mother. Here the doctor initially talked in non-technical terms (line 1) and asked a question at line 3 which opened up the floor to the patient to answer with a narrative that led up to her attempting to provide the technical term herself. The patient also later displayed her familiarity with the type of twins in line 11.

This study explores the quality of communication in routine obstetric ante-natal care using methodology that combines direct observation techniques with experiential data from patient interviews, and with a specific focus on identifying specific features of consultation discourse that contributed to patient-reported satisfaction and quality of communication.

Overall, patients’ reports on the communication in their consultations were very positive, and these reports were confirmed by researchers’ analysis of the recordings which indicates that the clinicians in this setting were able to meet women’s need for effective communication [54]. Some of the consultations were quite long (up to 1 hour), and the length of consultations may have contributed to high satisfaction levels.

The overall structure of the consultations varied according to the level of experience of the doctors, with more experienced consultants able to “cut to the chase” by not exhaustively following medical checklists. While most of the existing literature about consultation length comes from General Practice [55], this study aligns with hospital outpatient findings in which registrars took more time with patients than consultants [56]. Registrars sometimes completed a medical history before moving to the referral and this may have contributed to their longer consultations. The importance of the opening of the consultation and of the referral recognition sequence (RRS) in specialist practice, as reported elsewhere, was confirmed in this data set; when there were delays in the full RRS being completed, the smooth progression of the consultation was disrupted with patients sometimes asking explicitly why they were there or showing signs of increased anxiety in the face of this uncertainty.

Generally patients reported that their feelings were heard, anxieties acknowledged and questions answered. Again this was reflected in observations of the recorded consultations, which also showed high levels of participation and invitations for shared decision making. Many patients displayed fluency in many aspects of the discussion in the consultations, including familiarity with clinical terminology and a biomedical framing of the conditions contributing to the high risk state, especially where they had previous experience of complications in pregnancy and childbirth. This is not to say that the communication observed was uniformly positive, but our appreciative inquiry stance provides a constructive basis for recommending strategies and practices that are likely to enhance quality of communication.

We are not aware of other studies that have directly observed the detailed interactional processes at work in a generic high risk ante-natal clinic, and that correlate these direct observations with interview data from both patients and clinicians. Limitations of the study are the small size of the data set, possible skewing in the patient sample towards older and more highly educated women, and the observer effect (the communication in the consultations may have been affected by the presence of the recording devices, with participants perhaps inclined to show themselves in the best light). The interviews were also limited by their short length (to minimise imposition on participants). Some patient responses in interviews may have been affected by a reluctance to criticise their clinicians.

Our results show that communication in the high risk ante-natal context studied here differs from that documented in other health settings in a number of ways, notably the high levels of patient participation and the ways in which risk is discussed.

Given the challenges that clinicians in all disciplines often face in achieving effective interaction with patients, the results portray a positive picture of doctor patient communication and a relatively empowered and well-informed cohort of patients [57‐59]. Our results accord with survey studies that found an association between ‘high quality’ interactions with realistic and clear information and reassurance [60, 61]. The high level of patient satisfaction is consistent with a 2008 review of maternity services in this region of New Zealand which found low levels of dissatisfaction among 115 participants [62]. The specific elements that were highly rated in our interviews and observed in the consultation data are largely consistent with more high-level descriptions in other studies of what patients want from maternity care generally [10‐12, 54].

Previous research on surgeon communication has emphasised the need to make sure that patients are enabled to present their problems fully, despite the surgeon already having another source of information (achieved through the RRS) [47]. This ensures that patient and doctor are aligned about the purpose of the consultation. Attending to this by getting the topic of ‘risk’ on the table early in the consultation is also a way to reduce anxiety for patients who, by definition, know they have been referred for some reason. Our analysis of the ways in which the specialists in our data achieve the referral recognition sequence provides detail about the communication that underpins overall perceptions of effective communication, and this can inform training in medical communication.

The high level of patient participation observed in this study contrasts with earlier studies which found that many women undergoing antenatal care did not actively participate and were uncomfortable with decision making responsibility [15], rarely asked questions and were not encouraged to do so [63], and often did not share in decision making [34]. The increased ‘agency’ that we observed may be due to recent efforts to develop more ‘patient-centred’ care [64‐66] and shared decision making [67, 68]. The high participation levels may also have been influenced by the patients feeling that they were being given options and their choices respected, an endorsement of the communication styles of the study doctors. The fact that patients felt that the doctors were informative, clear and realistic with the information they provided also contributes to shared decision making.

Ante-natal and maternity care is a clinical environment where health literacy can have an important impact on outcomes [69]. We observed examples where patients displayed knowledge and were willing to ask questions in often complex areas of care such as choices regarding operative intervention or the type and timing of imaging, indicative of high levels of health literacy. The presence of the unborn child as an unseen and unheard additional patient who needs to be cared for by the mother (and partner) may provide an added incentive for patients to further develop their own health literacy and to be an active participant in the consultation on behalf of their child.

The explicit way that risk is talked about was noticeable in the interactions and reported by patients as being appreciated. Risk was framed in ways that were clearly articulated and understood. Explicit orientation to risk seems acceptable in this context where patients have been referred to a high risk antenatal clinic, a finding that is congruent with other research [70]. For example, in this study doctors opted to manage the potential for causing anxiety by explicitly addressing clinical risks, rather than avoiding such talk which may then prolong uncertainties and worry [70]. Focus group studies with women on topics such as gestational weight gain also indicate that women prefer these issues to be addressed explicitly but sensitively [71‐74].