Background
Description of the condition
The importance of effective communication between clinicians and patients was highlighted in the Institute of Medicine’s Crossing the quality chasm report (Richardson 2001). This report proposed free and open sharing of knowledge between patients and clinicians as 1 of 10 principles for redesigning the healthcare system to ensure delivery of optimal patient-centred care. Effective doctor-patient communication has been shown to affect patient satisfaction (Pollak 2011; Street 2009), decision making (NHS 2010), treatment adherence (Street 2009; Zolnierek 2009), a range of patient outcomes such as blood pressure and emotional health (Stewart 1995b), and doctors' job satisfaction (Maguire 2002). The information gathered as part of the medical history is critical to the establishment of an accurate diagnosis, with most information used to form a diagnosis gathered in this step (Peterson 1992). Subsequent steps of the medical consultation facilitate patient education and shared decision-making, with associations established between patient-centred communication, therapeutic alliance, and adherence to treatment (Pinto 2012; Thompson 2016). It has been estimated that a medical doctor will conduct approximately 200,000 medical consultations during his or her working career (Silverman 2013). Thus, it is incumbent upon educators that medical students are appropriately trained to ensure that, upon graduating, these consultations are conducted effectively using appropriate patient-centred communication (Simpson 1991).
Medical consultations should be a joint, collaborative effort between doctor and patient, using patient-centred communication to acknowledge and understand the patient's desires for information, shared decision-making, and discussions of care (Stewart 2001). While global agreement on definitions of the terms is elusive, there is broad agreement about the concepts and importance of patient-centred care and patient-centred consultation. These concepts are integral to the display of respect for patients, and thus, along with patient involvement in care systems, should be regarded as an ethical and democratic right (Gregory 2007). Definitions generally include elements of a biopsychosocial perspective of illness; consideration of each individual's personal meanings of illness; being sensitive to patients' preferences for information and shared decision-making; and developing a therapeutic relationship between doctor and patient (Mead 2002).
Patient-centred medical consultations call upon a set of skills that are considered both teachable and learnable by medical professionals at any point on their career trajectory (Aspegren 1999). We will use the term 'interpersonal communication' to refer to these skills which facilitate patient-centred communication and care. Interpersonal communication can be defined as communication that occurs from one individual to another (dyadic or small group), is non-mediated (face-to-face), and is shaped by the individual characteristics, social roles and relationships of the people involved (Hartley 1999). Interpersonal communication is the process by which we establish a communicative relationship and exchange messages to establish shared goals and understandings (Burleson 2010; Hargie 2011). We will refer to communication interventions as those aiming to improve the skills associated with such communication.
Significant progress has been made in the development and evaluation of formal curricula for interpersonal communication in medicine (Aspegren 1999; Smith 2007). Skills in interpersonal communication in clinical contexts are recognised as being different from everyday communication skills and should be developed through careful teaching and experiential learning (Benbassat 2009; Silverman 2013). Descriptive data suggest that students find acquiring skills in interpersonal communication challenging (Lumma-Sellenthin 2009; Royston 1997) for reasons including difficulty relearning or reconditioning engrained communication styles (Macdonald 2002), difficulty attending to medical and psychosocial needs simultaneously (Aper 2015), lack of exposure to models of patient-centred care (Thistlethwaite 1999), and wide variability between clinical and non-clinical role models (Rees 2002). With an absence of up-to-date, high-quality systematic reviews, evidence on the effects of communication curricula for improving medical students’ skills in conducting effective patient-centred consultations and in improving clinical practice is unclear.
The increasing demand for medical students to be specially trained to communicate effectively and efficiently has seen the emergence of skills associated with interpersonal communication as core graduate competencies in medical training around the world (Australian Medical Council 2012; General Medical Council 2015; Health Professions Council of South Africa 2014; Laidlaw 2009). New doctors are required to possess a range of skills for communicating in different formats such as face-to-face, online and by telephone, and different medical contexts, such as doctor-patient consultations, communicating about the patient (e.g. with other medical professionals), and communicating about medicine and science in general (e.g. lectures and conferences). In this review, we will focus on the medical consultation, referring to the verbal and non-verbal interaction between physician and patient that occurs during face-to-face encounters. The medical consultation has been described using a range of models, all of which summarise the process as including elements of relationship building, information gathering, information giving, and treatment planning which can occur in both initial and follow-up encounters (Keller 1994; Kurtz 1998; Makoul 1998; Novack 1992; Stewart 1995a). The overall goal of such encounters is a shared understanding of issues and plans, while the specific goals of any individual consultation can vary from diagnosis to an understanding of cause, risk, prognosis, the benefits and risks of various treatment options, health behaviour change, screening and any number of other therapeutic and health promoting activities.
Description of the intervention
Models of the medical consultation and communication training
Teaching and assessment of interpersonal communication have been guided by frameworks and models evolving over several decades (reviewed in Boon 1998). To the core elements of relationship establishment, information gathering, and patient education (Lipkin 1995) have been added the need for the doctor to gather information about the patient's understanding of his or her own health (Makoul 2001a), recognition of the influence of clinicians’ personal experiences on their interpersonal communication (Windover 2014), and the need to incorporate the electronic health record (Duke 2013). While the most recent studies have incorporated the use of technology, the fundamental structure and content of the models has not changed since the establishment of the Kalamazoo Consensus Statement in 2001 (Makoul 2001b).
The Kalamazoo Consensus drew upon five contemporary models for doctor-patient communication; namely the Bayer Institute for Health Care Communication E4 Model (Keller 1994), Three Function Model/Brown Interview Checklist (Cole 2013; Novack 1992), the Calgary–Cambridge Observation Guide (Kurtz 1998), patient-centered clinical method (Stewart 1995b), and the SEGUE Framework for teaching and assessing communication skills (Makoul 1998). It brought together a comprehensive set of skill competencies (Makoul 2001b). Specific communication tasks as well as knowledge, skills, and attitudes associated with the following essential elements are listed in the consensus statement: build the relationship, open the discussion, gather information, understand the patient’s perspective, share information, reach agreement on problems and plans, and provide closure.
Subsequent to the synthesis of communication tasks in the Kalamazoo Consensus Statement, Kurtz and colleagues expanded their Calgary-Cambridge guide to more clearly connect with the process of doctor-patient consultations (Kurtz 2003). This clarified and expanded upon the specific skills used at each step of the medical consultation process. For example, when gathering information, necessary skills include using open and closed questions appropriately, structuring, clarifying and summarising information, picking up verbal and nonverbal cues from the patient, facilitating patients’ responses verbally and non-verbally, and listening attentively. The marriage of process and content proposed by Kurtz and colleagues represents a true amalgamation of the communication skills (performance of specific tasks and behaviours) with the interpersonal skills (relational, process-oriented skills such as respect, empathy, and considering the patient’s perspective) required to establish a therapeutic relationship (Kurtz 2003; Makoul 2001b).
In this review we will include all interventions for medical students that specifically target the skills associated with what we have defined as interpersonal communication. These skills are likely to include: the appropriate use of open and closed questions, active listening, picking up on verbal and nonverbal cues, facilitating patients’ responses, eliciting patients’ concerns, considering the patient’s ideas, concerns and expectations (gathering and understanding their perspective), working in partnership with the patient to explain and plan, and make shared decisions, maintaining structure of the consultation, clarifying and summarising information.
Training methods
Interperonal communication training for medical students takes a range of forms depending upon the resources available, the current training level of students (e.g. undergraduate degree, postgraduate training program), and the context of learning (on campus, clinical placement, online). Interventions can be categorised as being delivered face-to-face or in self-directed formats. Face-to-face interventions are typically delivered as lectures for large groups of students or workshops for small groups. The latter can provide opportunities for participants to practice communication with real or simulated patients or their peers. They also enable provision of feedback from peers, facilitators and/or patients (whether real or simulated). These interventions may be based around live role-plays or feedback on videotaped consultations (Deveugele 2005; Maguire 1986). Self-directed interventions are those where the learner receives individual training using written or audiovisual materials, either in hard copy or in online or e-learning format such as online video demonstrations (Cook 2010).
Assessment of interpersonal communication
The Kalamazoo Consensus outlines three methods of assessing interpersonal communication: checklists (observer ratings); patient surveys; and examinations (of knowledge and perceptions using traditional written questions or questions linked to stimulus such as a video vignette). Both checklists and patient surveys can be used in the assessment of interactions with real or simulated patients, can occur live or based on recorded interactions, and can be used in formative or summative assessments such as Objective Standardised Clinical Examinations (OSCE) (Duffy 2004). Given the variation in validity and reliability among techniques, the assessment method utilised affects the capacity to compare different intervention studies. In this review we will categorise each study based on the method of assessment (observer ratings or survey/examination scores) and the nature of the observed consultation (real patient, simulated patient, live, recorded). Given our focus on behaviour change, we will not include data obtained through student examinations or surveys. Where studies use more than one of the included methods of assessment, data for each outcome will be extracted and compared to other outcome data as appropriate.
How the intervention might work
Interventions to improve the interpersonal communication of medical students aim to produce doctors capable of delivering effective, safe and patient-centred health care when they enter the workforce. Education-based interventions work by bringing about change in learners’ attitudes, increasing their knowledge, and importantly, increasing their competence in performing particular skills. In the case of medical consultation skills, educational interventions are likely to improve learners’ skills and knowledge through: modelling by and feedback from educators; and experiential learning, with opportunities to practice, reflect, and receive constructive feedback, draw upon knowledge and previous experience, and learn in a self-directed fashion (Kaufman 2003). While the highest level of evaluation of learning is the application of skills in clinical practice leading to improved patient outcomes, the only immediately measurable outcomes for undergraduate students are improvements in skill, knowledge, attitudes, and confidence (Kirkpatrick 1996; Naugle 2000; Smidt 2009). Student learning outcomes can be conceptualised hierarchically (Alliger 1989). At the very least, participation in a communication intervention should increase knowledge of patient-centred approaches to communication. Next, these interventions should increase confidence in undertaking effective doctor-patient consultations. However, the ultimate goal of communication interventions should always be to improve actual behavioural skills to undertake evidence-based doctor-patient consultations. These behavioural outcomes are assessable using the methods outlined in the Kalamazoo Consensus Statement and form the basis of the outcome measures assessed in this review.
Why it is important to do this review
Given community and professional concerns regarding the physical, emotional and financial impact of poor communication by medical practitioners, there is a critical need to evaluate communication training programs in medical education. To date, the effectiveness of interventions for improving medical students’ interpersonal communication has not been demonstrated unequivocally. Moreover, there is significant variability in communication curricula across medical schools (Hargie 2010; Hoffman 2004) and new methodologies have become increasingly popular since the publication of previous reviews (Lanken 2015). Given the rapid evolution of innovative teaching and learning approaches, it is timely to review the effectiveness of approaches which have been utilised. Thus there is a need to determine:
the evidence base for communication interventions for medical students;
the teaching and learning approaches associated with improvements in medical students' interpersonal communication;
the most effective approaches to teaching medical communication in the context of pragmatic limitations of medical curricula; and
gaps in knowledge about communication interventions for medical students to guide future teaching and research endeavours.
This review will aim to provide necessary guidance to medical educators and medical education accrediting bodies regarding the most effective communication programs in medical curricula, and identify the necessary resources for teaching these programs.
Determining the evidence base for communication interventions for medical students
A number of completed and ongoing reviews have sought to examine the effectiveness of communication training programs in medicine (Aspegren 1999; MacDonald-Wicks 2012; Smith 2007; Van Nuland 2005). Aspegren 1999 reviewed 83 randomised, quasi-randomised, and non-randomised trials and descriptive studies of communication training for medical students and concluded that teaching interpersonal communication to medical students can improve the students' ability to undertake doctor-patient consultations. However, the methodological quality of included studies was not adequately assessed, and the inclusion of non-randomised trials and descriptive studies limited the extent to which improvements in interpersonal communication could be attributed to the interventions described. Smith 2007 identified 24 randomised controlled trials (RCTs) that were available from 1977 to 2005 and conducted meta-analyses on 15 that met their inclusion criteria. Smith 2007 only included RCT interventions; however, given the settings in which these interventions are expected to be delivered (e.g. university classrooms, hospital clinics), it is possible that other study designs (such as cluster and quasi-RCTs) may also be relevant. In addition, a preliminary search of the research undertaken since 2005 suggests at least 30 additional RCTs of communication training for medical students have been published since the Smith 2007 review.
The authors of a planned Cochrane review (Van Nuland 2005) intend to assess the effects of communication training programs specifically for general practice (GP) trainees. GP trainees are completing their training for specialisation, and as such, have advanced in their training beyond the basic medical degree. Furthermore, Van Nuland 2005 will exclude studies that include undergraduate students. The authors of MacDonald-Wicks 2012 are reviewing the effectiveness of assessment tools and methods for teaching interpersonal communication to students in the health professions. Their review includes students from undergraduate and postgraduate medical, nursing, and allied health programs, including nutrition, dietetics, occupational therapy, physiotherapy, among others. Given the heterogeneous nature of professional roles and scope of practice, undergraduate training programs, and student cohorts, it is important to examine medical education separately.
The review by Smith 2007 demonstrated that providing structured feedback on participants' performance, and engaging in small group discussions were associated with larger improvements in skills compared to other methods (e.g. lectures, clerkship experience, assigned readings). An overview of systematic reviews of strategies for teaching communication skills to qualified doctors (Berkhof 2011) also reported little evidence for interventions based on lectures, or those based on modelling appropriate interpersonal communication to participants. Like the review by Smith and colleagues, stronger evidence was reported for interventions based on role-plays and feedback from educators, particularly when used in combination with self-directed didactic techniques (e.g. written information, reviewing videos).
Our review will differ from these in a number of ways. First, we will ensure methodological rigour by appropriately reviewing identified studies with careful consideration of research design and additional features of methodological quality. Second, we will focus on students completing an undergraduate or graduate-entry medical degree which will help to ensure that we know what works for these students, as compared with more refined samples completing a medical specialisation (e.g. Van Nuland 2005), or students in other allied health programs (e.g. MacDonald-Wicks 2012). Third, identifying the intervention characteristics associated with improvements in interpersonal communication will help to focus the development of future curricula.