A Study Of Power Relations In Doctor-Patient Interactions In Selected Hospitals In Lagos State, Nigeria

This paper explores power relations in clinical interactions in Nigeria. It seeks to investigate the use of power between doctors and patients during consultations on patient-centred approach to medicine in selected public and private hospitals in Lagos State, Nigeria. The objective is to establish how doctors' projection of power, using the discourse resources of transitivity, affects positively or negatively their relationship with patients. This study employs triangulation as its methodology. A judicious mix of quantitative and qualitative methods has been utilized to give the study a scientific shape. Proportionate stratified random sampling and purposive sampling procedures were employed. The study employs the theoretical and analytical paradigms of Systemic Functional Linguistics and Critical Discourse Analysis. The findings revealed that doctors predominantly use this process to the benefit of the patients. Minimal level of intrusive and cooperative interruptions was also observed. It is expected that this study will give more visibility on the best way patients can be empowered by lessening doctors’ use of polar interrogatives and completely avoiding interruptive discourse in clinical interactions.


Introduction
This study is an exploration of the use of the discourse resources of transitivity as deployed in the enactment of power in doctor-patient interactions in selected public and private hospitals in Lagos State, Nigeria. Issues of power in clinical encounters have been linked to the problem of imbalance in the discourse of the participants; hence, the study examines how the doctor's use of power supports or negates the principles of patient-centredness in medical consultations. Using the theoretical and analytical parameters of Systemic Functional Linguistics (SFL) and Critical Discourse Analysis (CDA), this study examines how the discourse roles of interactants reveal the dynamics of power projection in clinical encounters (transitivity).

Background to the Study
Studies on interpersonal communication in medical encounters confirm that doctor-patient rapport is key to effective health care delivery (Kurz et al, 2003). Patients' physical wellbeing is highly dependent on effective technical knowledge backed up by a robust and effective interpersonal communication (Ambady et al, 2002). Power relation is a pervasive phenomenon in social life and language, being an inalienable part of our social life, possesses multifarious ways of exhibiting power. As Kress (1982) observes '…language is entwined in social power in a number of ways: it indexes power, expresses power, and language is involved wherever there is contention over and challenge to power'. Again, studies have revealed that when doctors misuse their power within the doctor-patient encounter, patients' attitude towards doctors, health care, and in how they perceive themselves are adversely affected ( Gridley et al , cited in Polimeni, 2004). Clinical encounters have the fertile potential for social power abuse so also is patients consent to the doctor's authority and advice because of the doctor's gate-keeping monopoly over matters such as surgery, prescriptions, insurance, and sick leave (Friedson, 1970). Medicine is a social act and at the heart of effective diagnosis and treatment lies interaction between doctors and patients. A good rapport between doctors and patients is indispensable for a healthy society. This is evidenced by the fact that doctors conduct a mean of 120,000 to 160,000 interviews with patients in a practice lifetime (Lipkin et al, 1995). According to the American Medical Association (AMA) Code of Medical Ethics (1847), the life of a sick person can be shortened not only by the acts, but also by the words and the manner of a physician.

Objective of the Study
The objective of this study is to investigate the role of transitivity options of power projection in doctor-patient interactions in Lagos State, Nigeria.

Statement of the Problem
A catalogue of studies has been done on doctor-patient interactions. For instance, the use of questions to claim power (Frankel, 1992); problem presentation stage of medical talk (Ruusuvuori, 2000); diagnostic stage of medical interaction (Heath, 1992); opening stage of medical encounters (Robinson and Heritage, 2006); topic transition (Ainsworth-Vaughn, 2001); the structure of medical consultation (Byne and Long, 1976) and women powerlessness in medical interactions (Polimeni, 2004). Most of these studies were done outside the theoretical and analytical frameworks of Systemic-Functional Linguistics and Critical Discourse Analysis. This study hopes to fill in this gap by taking an indepth look at power relations in doctor-patient interactions using the theoretical lens of SFL and CDA. Moreover, transitivity analysis of institutional talk has been described as "a novel enterprise" because previous studies "tended to concentrate entirely on written texts" (Shimazumi, 1996) including fiction (Gregory et al, 1988) or non-fiction (Francis and Kramer-Dahl, 1992).
In Nigeria, studies on medical discourse are still in the infancy stage (Adegbite and Odebunmi, 2006). According to Odebunmi (2006), most studies on medical discourse in Nigeria have focused on the register, and the "pragmatics of the discourse" e.g. Ogunbode (1991) and Odebunmi (2003Odebunmi ( , 2006 and none of these works has adopted SFL analytical parameters nor have they employed an elaborate quantitative methodology. Hence this study attempts to fill in the theoretical and analytical gaps left by these studies by applying SFL and CDA paradigms in its exploration of power relations in doctor-patient interactions as well as deploy a mix of qualitative and quantitative methodology. Moreover, clinical consultations have often been criticized for being too paternalistic or doctor-centred. In this model, doctors see patients as cases or diagnosis rather than as persons. Doctors focus on symptoms but ignore the psychosocial context of the patients and their conditions. According to van Weel-Baumgarten (2010), "the doctor does most of the talking and takes decisions on diagnosis and treatment, without shared responsibility and decision-making" This strict bio-medical approach may breed negative outcomes such as low compliance by patients and social power abuse by doctors. Better outcomes can be achieved with a patient-centred clinical interaction. This study aims to critically analyze doctor-patient interactions in order to account for linguistic elements which negate or support patientcentredness in medical encounters in Lagos State, Nigeria.

Theoretical Framework
In this study, two theories were employed: Systemic Functional Linguistics (SFL) and Critical Discourse Analysis (CDA).

Systemic Functional Linguistics (SFL)
SFL is a tristratal construct of semantics (meaning), lexico-grammar (wording), and phonology (sound) which constitute a semiotic system. According to Eggins (2004), (Systemic), systemic linguists make four main theoretical claims about language: (a) That language use is functional; (b) That its function is to make meaning; (c) That meaning is influenced by social and cultural contexts; and (d) That the process of using language is a semiotic process in which people make meanings by making linguistic choices.

Critical Discourse Analysis (CDA)
CDA is a political school of discourse analysis whose major agenda is in "identifying how language contributes to power imbalances in society and how analysis of this might contribute to change in favour of the oppressed (Ward, 2004)). CDA aims to remove false or distorted consciousness and to render transparent what had been previously hidden from public consciousness by initiating a process of self-reflection in individuals and groups. CDA has a deconstructive agenda and is concerned with exposing language deployed in the service of power.

Triangulation
In this study, we adopted the concept of triangulation. Triangulation is the use of multiple methods in the study of the same object. According to Wolfram-Cox and Hassard (2005), the implicit assumption in using triangulation "is of developing a more effective method for capturing and fixing of social phenomena in order to realize a more accurate analysis and explanation". Since "different methods have different strengths and weaknesses" (Gillham, 2000) the "effectiveness of triangulation rests on the premise that the weakness in each single method will be compensated by the counter-balancing strengths of another" (Jick, 1979). Hence, triangulation can potentially elevate researchers above personal biases that emanate from a single methodology.

Types of Triangulation Used in this Study
In this study, three types of triangulation have been employed: (a) data triangulation, (b) theory triangulation, and (c) methodological triangulation.
(a) Data triangulation: This involves collecting and using data from male, female, young, old patients as well as male and female doctors in private and public hospitals in different parts of Lagos.
(b) Theory triangulation: SFL and CDA have been employed as theoretical tools.
(c) 'Methodological triangulation: "between" method triangulation is used whereby the following mix of qualitative and quantitative sampling procedures is employed: (i) Probability Sampling Technique (Proportionate Stratified Random Sampling); (ii) Non-Probability sampling Technique (Purposive Sampling)

Data Collection Procedure
The data which constitute the focus of this study were drawn from audio recordings of spoken interactions between doctors and patients in selected private and public hospitals in Lagos State. It is the interactional record of face-to-face conversation between patients and General Practitioners (GPs). Routine interactions between GPs and patients were audio-taped and transcribed.
Non-participant observation was adopted. This is often used to study focused aspects of a setting, to answer specific questions within a study. Hence, "it can thus be a powerful tool in triangulation" (Savenye & Robinson, 2004). Nonparticipant observation was chosen as the appropriate data collection method because the use of participant observation in close clinical setting may lead to what Stubbs (1983) calls observers' paradox, which simply means that when an observer is physically present at a location, people being observed may put up artificial behaviour. This is highly likely in a medical setting where sensitive matters are topical and of ethical concerns.

Sources of Data
Two selected hospitals were used for this study: (a) University of Lagos Health Centre (Public) and (b) R-Jolad Hospital, Gbagada, Lagos (Private)

Role Projection
(1) Dominant processes uttered by the doctor and the patient The dominant processes in the data are material, relational, and mental. The doctor utters 80.3% of material process in all the data while the patient utters 19.7%. Of the 85 relational processes uttered by both doctor and patient across the five data, the doctor uttered 54 (63.5%) while the patient uttered 31 (36.5%). For mental processes, of the 39 used across the five data, the doctor used 30 representing 77% of the total while the patient uttered 9 representing 23% of the total. This shows again the doctor dominating the use of the major transitivity systems. The total number of the minor processes (verbal, behavioural and existential) amounts to 39 (13.4%) across the five data. The doctor also dominates them by uttering 30 processes representing 77% while the patient has 23%. Overall across the five data, 291 processes were uttered by both doctor and patient. The doctor used 216 processes representing 74.2% of the total while the patient uttered 75 processes representing 25.8% of the total. Hence, the percentage difference reveals that in terms of transitivity, the doctor apparently has the upper hand enabling him/her to control the course of the interaction.

(2) Difference in the dominant processes used by doctor and patient
The material processes uttered by the doctor (80.3%) differ significantly more than that used by the patient (19.7%). There is also a significant percentage difference between the relational processes uttered by the doctor (63.5%) and the patient (36.5%). However, there is no major difference between the mental processes used by doctors (76.9%) and those of patient (23.1 %). This shows that both the doctor and the patient do a lot of thinking exercise in the course of the encounter. The doctor also seems to have responded well to the patients' emotional needs in the encounter. Both interactants used most of the process types (material, mental, relational, verbal, and behavioural). From CDA standpoint, the patient is given a voice to express her views and concerns in the encounter.

(3) Mutual role projection by doctor and patient
This has to do with who projects whom in which roles? To answer the question, it is necessary to account for two types of role projections: projection by doctor and projection by patient. According to Table 2 below, the Doctor projects the Patient considerably more often (83.3%) than herself (16.7%). The Doctor projects the Patient more as Actor, Goal, Senser, Carrier, and Sayer. This is indicated by the percentage differences. As indicated in Table 3 below, the Patient projects himself (85.90%) significantly more than he projects the Doctor (14.10%). These figures show that the general trend of the interactions focuses on the Patient. The Patient is so much consumed by her own condition that most of the major transitivity roles of Actor, Goal, Senser, Sayer and Carrier are heavily allocated to himself leaving little for the Doctor.  It is important to show the difference in the way transitivity roles are projected for Doctor and Patient in individual interactant's utterances. Table 4 reveals the projection of Doctor in individual transitivity roles in the utterances of each discourse participant. Accordingly, the Doctor is projected by herself (88.8%) much more than by the Patient (7.5%). This difference is particularly significant for the roles of Actor, Goal and Senser.  Table 5 reveals that the projection of the patient significantly differs across interactants. The Patient is highly projected by the Doctor (78.10%) which is considerably different from the Patient's projection of himself (21.90%). This uneven distribution across speakers runs through all the major transitivity roles of Actor, Goal, Senser, Sayer, Carrier and Behaver. The implication of this in the medical practice is that the patient is put at the core of the consultation. The considerable difference in role projection for both interactants in each other's speech is in favour of the Patient and good for medical practice. We are therefore led to hypothesize, following Shimazumi's (1996) assertion that 'the Doctor will project roles for herself and patient, the patient will be projecting primarily her own role'.

Role Allocation
The roles that social actors are given to play in representations play a major part in the studies of many critical linguists (e.g. Fowler et al, 1979 andvan Dijk, 2009). Role allocation refers to "who is represented as 'agent' (Actor), who is 'patient' (Goal) with respect to a given action (van Leeuwen, 1996). Activation and passivation are the key elements in role allocation. Activation happens when social actors are portrayed as the active, dynamic forces in an activity while passivation occurs when social actors are represented as 'undergoing' the activity or are at the receiving end of it (van Leeuween, 1996). Transitivity structures activate and passivate social actors by giving them grammatical participant roles coded as Actor in material processes, Behaver, in behavioural processes, Senser in mental processes, Sayer in verbal processes or Assigner in relational processes (Halliday, 2004). An example of activation from our data is: (a) Pt: Eh well, I just want to know my condition again (Hypertension, 4) An examples of passivation is: In (a), the active agent is the patient while in (b), it is the doctor. The patient is the passive agent in (b). Taking Table 2 into consideration, of the 107 allocations of Actor role in the doctor's utterances, 14% (15) is allocated to the doctor while 86% (92) is assigned to the patient. The role of Goal for the patient is 97.3% (71), while the doctor allocates (2.7%) to himself. This is because he is not the focus of the encounter. As for the role of Senser, the doctor allocated 55.9% to himself, while assigning 44.1% to the patient. 1t appears the doctor tends to assert his superior mental ability and activity in the encounter. Obviously, as the moderator-in-chief of the consultation, a lot of thinking and knowledge is needed to make it an optimal success. For the role of the Sayer, the doctor allocates 81.2% to the patient while 18.8% is assigned to himself. This is good news because it portrays the patient as someone who has 'said', 'told', 'asked' and 'explained' something in the interaction.
How does the patient allocate roles in her utterances? Readings from Table 3 show that the patient allocates 90.9% (20) of the role of Actor to himself, while giving 9.1 % to the doctor. The patient also makes himself the Goal of his utterances 90.9% (10) while assigning 9.1 % (1) to the doctor. As a Senser, the patient assigns 90% (9) of the utterances to himself, giving the doctor a paltry 10% (1). As a Sayer, the patient has 71.4% (5) while the doctor has 28.6% (2). These figures confirm that the most activated or active participant in the interaction is undoubtedly the patient.

Exclusion
According to Van Leeuwen (1996), exclusion is "an important aspect of Critical Discourse Analysis". Powerful social actors exclude the less powerful from discourse events or texts to serve their own parochial agenda especially if such exclusions leave no trace of their activities. He postulates two main types of exclusion -suppression and backgrounding. Suppression operates when "there is no reference to the social actor(s) in question anywhere in the text". In the case of backgrounding, "the excluded social actors may not be mentioned in relation to a given activity, but they are mentioned elsewhere in the text". Suppression is realized in a text through passive agent deletion. Perhaps because of the real time and face-to-face dialogic nature of doctor-patient interaction, there is no hard evidence of suppression in terms of passive agent deletion in the data. Rather, the patient is portrayed as active agent, responsible for their actions. E.g.: Have you been using your drugs regularly (Hypertension, 12) Doctors and patients have been using active verbs to describe situation in which they are agents. So suppression is not evidenced in the data. Since the patient and her condition are at the heart of the consultation, 1t is very difficult for her to be excluded from the discourse of the encounter. All agents of action (patient, doctor and illness e.g. you, 1, and it) are not excluded. Hence, agentive participants are not excluded from the interaction. Beneficiaries (social actors who benefit from an activity) have also not been excluded

Discussion and Conclusion
The use of material processes by the doctor shows that s/he is highly interested in "facts, events and information so as to assess the patient's condition. Doctors' use of material processes shows their interest in the physical conditions of the patient. The use of relational processes assists the doctor in effective diagnosis since relational process types serve to identify and characterize (Halliday and Matthiessen, 2004). The use of mental processes also shows that the doctor is more involved in the mental condition or feelings of the patient because when a patient describes how s/he feels, thinks or perceives a problem situation, it helps the doctor to present a more appropriate diagnosis, prescription and treatment. This is because mental processes involved a Senser (a human agent who feels and thinks) and a phenomenon (a thing felt, thought or perceived). By using material and mental processes predominantly, the doctor is able to delve into the outer (material) and inner (mental) worlds of her patients thereby achieving one of the golden principles of holistic medicine -"medical treatment based on the belief that the whole person must be treated, not just the part of their body that has a disease" (Longman Dictionary of Contemporary English, Sixth Printing, 2012). While the doctor is the powerful discourse participant, he uses his power to the benefit of his clients through the judicious use of transitivity processes.