“Best Practice” for Patient-Centered Communication: A Narrative Review

For decades, medical educators and patient advocates have stressed the importance of communicating with patients and their families during medical encounters, and communication has emerged as an important physician competency.1–5 There is a consensus about the essential elements of communication skills relevant to medical encounters and the need to teach those skills to medical trainees.6–11 Further, the assessment of communication skills has become an explicit component of medical education curricula and of formative and summative examinations for physicians in the United States and Canada.10,12–15 

At the same time, patients report that many of their informational and emotional needs remain unmet during encounters with their physicians.16–20 Currently, training and role modeling of communication and interpersonal skills in medical education is relatively brief, is placed early in the curriculum, and often is not reinforced in the latter stages of training.21–25 The decline in empathy and communication as trainees progress through programs has been well documented across multiple studies.26,27 If the medical education community is dedicated to renewing its commitment to teaching excellence in the communication skills of physicians, which some have called for,22 there is a need for better understanding of the evidence that supports the efficacy of good communication skills, and use of that evidence to define what should be taught to medical students and residents. In this era of milestone development, it seems timely to offer scientific support for the efficacy of good communication skills.

Early work to define communication skills relevant to medical practice used the terms physician-patient communication or excellence in communication. The term patient-centered communication has emerged in more recent writing on the subject. This usage followed the pronouncement from the Institute of Medicine in 2001 that medical care should become more patient-centered, that is, more responsive to patient needs and perspectives, with patient values guiding decision making.28 Although definitions of patient-centered communication vary,29,30 there are common core components. Epstein and Street31 offered an operational definition of patient-centered communication: (1) eliciting and understanding patient perspectives (concerns, ideas, expectations, needs, feelings, and functioning), (2) understanding the patient within his or her unique psychosocial and cultural contexts, and (3) reaching a shared understanding of patient problems and the treatments that are concordant with patient values.

More recent empirical studies and reviews couch their findings in patient-centered terminology. One way of conceptualizing the patient-centered method for encounters is to consider both parties' agendas: (1) the physician's focus on explaining the illness in terms of the taxonomy of disease, and (2) the patient's focus, which encompasses perspectives on his or her illness, the need for information and understanding, and the desire for partnership in management. The physician's task is to address both agendas, resolving any conflict between the 2 by further dialogue and negotiation.32 

Research that seeks to link physician communication skills during encounters to outcomes after those encounters has been conducted since the late 1960s, with many of these studies focusing on patient satisfaction after encounters with physicians.33–38 Although patients may be satisfied (or dissatisfied) because of factors unrelated to communication (eg, wait times and other attributes of the encounter), the available evidence indicates that patient satisfaction is strongly associated with the communication behaviors that occur during the physician-patient interaction. Several studies that focused on patient satisfaction were able to outline specific behaviors associated with satisfaction, grouping them into 2 categories: task-oriented behaviors (eg, drawing out patients with active listening responses and providing detailed information), and affective behaviors (ie, socioemotional exchanges, such as responding empathically, showing caring, and addressing the patient's main concerns).39,40 The Stewart and Roter41 and Roter et al42 studies involving interactional analysis showed that task behaviors were the more important contributors to patient satisfaction, but, to be effective, those behaviors needed to be couched in patient-centered terms, including not being overly directive while being attentive to the patient's receipt of information. Did the patient understand? Was the information congruent with the patient's beliefs and circumstances? Affective behaviors also were associated with satisfaction, but the relationship was less strong. Nonverbal communication behaviors, such as eye contact and listening attentively, are also linked to increased patient satisfaction.43 

Recent research on communication has sought to link clinician behaviors and skills to other important, but less immediate, clinical outcomes, such as patient recall, understanding treatment recommendations, and adherence to those recommendations.41,42,44–52 In these studies, the association of outcomes with clinicians' communication skills was less consistent and less strong than in studies that focused on patient satisfaction. The most consistent findings suggest that communication needs to be consistent with basic principles of information transfer. Successful communication should

• Be uncomplicated

• Be specific

• Use some repetition

• Minimize jargon

• Check patient understanding

In addition, communication should simultaneously employ a patient-centered approach and interpersonal interaction to promote patient satisfaction. Overly directive communication and teaching appeared to have negative consequences.41,42 

Several recent studies and reviews have found positive associations between communication skills and improved health outcomes for patients, such as physiologic measures (eg, blood pressure, blood glucose levels), health status (eg, headache frequency, depression), and measures of functional status, including less patient distress with the illness experiences.25,31,45,53–67 Most recently, Weiner and colleagues,68 in an observational study, showed improved health care outcomes in patients in primary care clinics who had their cues about psychosocial factors attended to. This data set must be viewed as intriguing and suggestive but incomplete. Many of these studies are small, and there is a lack of consistency in findings across studies. It is challenging to attribute the positive outcomes studied to the earlier clinician communication behaviors the patient experienced, and there are multiple influences and potential confounders, many of which reside outside the clinician's sphere of influence.

Studies have shown that patients of physicians with high malpractice claims have twice as many complaints related to communications than do physicians with low malpractice claims.20 Similarly, Tamblyn et al69 reported a link between poor patient-physician communication, as measured by the Medical Council of Canada's clinical skills licensure examination, and subsequent rates of malpractice claims. Levinson and colleagues70 studied audiotapes of physicians with low and high malpractice claims and found that primary care physicians with no malpractice claims used more facilitation, encouraged the patient to talk, checked his or her understanding of explanations and instructions, and solicited his or her opinions. As a group, studies about malpractice claims show a link between poor communication and later complaints.71,72 This body of research, however, does not offer guidance on the specific communication skills that are most important in reducing the risk of malpractice.

Mazor and colleagues73 studied breakdowns in cancer care in 416 patients. Of the 93 patients who experienced problems with their treatment, 44 (47%) reported communication breakdowns, including fundamental problems in information sharing, emotional support, and care coordination. From the patients' perspectives, some of these were reported to be as harmful as traditional adverse events or errors.

Another negative outcome of poor physician communication skills may be the missed opportunities to improve self-management by patients with chronic disease. This area is less well studied. At the same time, given the burden of chronic disease from smoking, eating disorders, and alcohol consumption, combined with the evidence regarding positive outcomes attributable to physician communications about these behaviors, ineffective behaviors and missed opportunities have become an area of concern.17,74–77 Studies in this area will need to link specific clinician behaviors to positive and negative outcomes, as well as to disease-specific, self-management outcomes.

Patients are the recipients of health care services, making it relevant to identify what they want from their interactions with providers. In addition, if we consider the suffering and anxiety associated with illness and the difficulty seeking some types of care, patients' statements about what they wish for and are unhappy about in their interactions with physicians are worthy of consideration.

Two surveys75,76 concluded that patients strongly desire a patient-centered communication approach. In these studies, patients expressed the wish that their physicians would

• Explore the patient's ideas about the problem (thoughts, worries, feelings, expectations) and take the patient's input seriously

• Try to understand the whole person and family influences and how the problem affects the patient's life

• Tell the patient what is wrong in plain language

• Seek common ground and partnership: agree on the nature of the problem, on the priorities, and on the goals of treatment; make management decisions and clarify the respective roles of the physician and patient

• Strive for an enhanced physician-patient relationship: be approachable and friendly, share decision making, show genuine care, and be respectful

In 2005, Mazor and colleagues78 asked lay participants to serve as analog patients and rate the relative importance of a list of desired clinician behaviors. This study provided strong support for the same encounter behaviors that have emerged from empirical studies of patients: soliciting patient input on the agenda and expectations for the visit, allowing the patient to speak without interruption, presenting information clearly without jargon, providing specific advice and recommendations, asking for patient input about treatment plans, checking patient understanding, and acknowledging patient emotions. The patient-centered components of the Good Medical Practice—USA document79 echoed these same desired medical behaviors. It also emphasized the need for physicians to

• Discuss the costs of different tests, medications, and treatment options, and take into account what the patient's health care insurance will cover

• Understand and be responsive to the patient's living circumstances and support structure

• Offer involvement and support for other caregivers of the patient's choosing79 

Taken together, there is an impressive body of evidence that supports a positive association of physician communication behaviors with multiple outcomes of interest (table 1). Good communication skills clearly lead to more satisfied patients. Satisfaction is a desired outcome in its own right. The data suggest it is also a necessary (yet perhaps not a sufficient) condition for other patient outcomes, such as recall, patient understanding, and adherence to therapy. Although the evidence base is weaker, there are data to support the effect of good communication skills on “more distant” outcomes, including physiologic parameters and health status.

Based in part on the evidence that has emerged to date, there is a remarkable consensus about what constitutes competency in physician-patient communication in general and in patient-centered communication, in particular.7,9–11,31,80 This consensus is organized around 6 core functions (or goals) for medical encounters: (1) fostering the relationship, (2) gathering information, (3) providing information, (4) making decisions, (5) responding to emotions, and (6) enabling disease- and treatment-related behavior (self-management; table 2).

The framework above has, in turn, stimulated the outline of domains and subdomains of specific clinician behaviors, organized under the 6 key functions.61,81–88 Similar lists can be derived from consideration of instruments used to assess communication skills.89 

It is possible to create an amalgam of these skills, which can be considered as a best practice for communication in medical encounters (table 3). The concept of best practice acknowledges what de Haes and Bensing80(p293) have concluded: “Although in much of the literature on medical communication a compelling case is made about the relevance of adequate communication for high quality care, the evidence base of this argument is underdeveloped.” A true evidence base for communication will articulate “the hypothesized relationship between specified communication functions and elements and concrete endpoints (including the assumed mechanism and pathways).”80(p293) Smith and colleagues88 agree that better-described, behaviorally defined skills are a critical next step in advancing the field.

The literature demonstrates that physicians can learn patient-centered communication skills.23–25,63,90–92 A review by Levinson, Lesser, and Epstein25 cited several studies that showed a strong association between physician training in, and later use of, patient-centered communication skills in medical encounters. High-intensity interventions (ie, multimethod, longer instructional time) had a stronger effect than did lower-intensity teaching. Health care institutions (hospitals, physician group practices) have developed training programs to teach patient-centered communication skills. Such programs have led to improved physician self-confidence in communication skills, as well as improvement in patient satisfaction.83,93,94 

Reports of patients as well as our own observations of performance on the United States Medical Licensing Examination Step 2 Clinical Skills Examination suggest that there remains a large gap between what is perceived as best practice by experts and what is actually used in clinical settings. This suggests that communication skills training needs to be increased, particularly in the later years of medical school and in residency. We also need to better understand why more physicians fail to embrace or prioritize patient-centered communications. Certainly, practice setting and problem urgency affect the scope of communication skills appropriate for any given encounter. There is a widespread perception by physicians that being patient-centered will take more time. At least 1 study95 has shown a weak association between visit length and use of patient-centered techniques. Levinson et al25 and educators assert that communication is like any other skill: it requires practice for efficient use. Another study supports this notion, showing that consultation time actually decreased once physicians developed better facility asking about and responding to patient questions.96 Thus, effective approaches for teaching communication skills require a much better understanding (at the individual and system levels) of the reasons why there has not been more robust adoption of patient-centered communication skills by physicians.

We assert that a compelling case can be made for a set of communication skills and behaviors; this emerges from the empirical evidence and the statements of experts and patients (table 3). This amalgam may be considered a best practice for communication in medical encounters, and at present, is our best guide. This assertion is supported by data outlining the prevalence of nonadherence to medications in US health care, by the growing use of patient satisfaction studies as a quality of care measure, and finally, by the attention of self-management in chronic disease.97,98 These issues touch all clinical disciplines and can be addressed by state-of-the-art communication between physician and patient. The specific skills listed in the right-hand column of table 3 form a basic skill set around which our collective curricular and assessment efforts should be directed.

Once learned, basic skills can be flexibly modified based on patient, disease, health care setting, and other differences that are inevitable in patient care. In some circumstances, the emotional reaction to illness is minimal and requires very little response; patients may have little to say about their current problem if it is minor and/or of recent onset, other than to express a desire to have it resolved. Not all patients desire a role in decision making, and not every clinical situation calls for a shared decision-making approach. All specialties have been challenged to develop curricula and assessment methods for interpersonal and communication skills as part of the Accreditation Council for Graduate Medical Education (ACGME) 6 competencies.99,100 Each specialty likely has key issues that require particular emphasis and/or some modification of a basic communication skills set. A focus on these discipline-specific needs for physician-patient communication is likely to emerge from the ACGME Milestone Project, currently being implemented in 7 specialties that entered the Next Accreditation System on July 1, 2013 (emergency medicine, internal medicine, neurological surgery, orthopedic surgery, pediatrics, diagnostic radiology, and urology).101 

A best practice approach allows us to move forward, while acknowledging—as have de Haes and Bensing80 and Smith et al88—that the evidence base underlying the efficacy of patient-centered communication is underdeveloped. Issues, such as inadequate conceptual grounding, small sample sizes, varying definitions and assessment methods, and inconsistent results, have been noted by various investigators.30,61,80,102 Many of the studies assert causation of certain outcomes by selected behaviors, when the research designs only allow conclusions about associations of behaviors with outcomes. Future research should address these weaknesses and address less-studied areas, such as shared decision making and patient self-management.30 In addition, there are areas important to modern practice for which we have very little evidence: the ever-shortening time available for medical encounters and the use of electronic health records. Further, there is a need for a much better understanding of how the needs and concerns of patients with specific diseases and conditions interact with the communication behaviors outlined in table 3. What behaviors by clinicians result in the best patient understanding of an explanation of a complex disease, such as rheumatoid arthritis? What skills result in the best acceptance by patients offered a choice of therapies for cancer? What communication works best to optimize self-management in chronic diseases, such as asthma or depression?

The field of physician-patient communication has come a long way since the pioneering work of Korsch et al,33 Lipkin et al,11 Stiles et al,34 Kaplan et al,54 Roter et al,42 and Francis et al.47 The scope of communication skills important for successful clinical encounters has been broadened and better defined.103 It has been demonstrated that these behaviors can be taught, and strategies for both instruction and assessment are developing. There is also a growing body of medical educators and researchers contributing to the field as well as national (American Academy on Communication in Healthcare)96 and international (European Association on Communication in Health Care)104 academies, each with their own periodicals (Medical Encounter,105 ,Patient Education and Counseling103).

An effort to make significant enhancements to medical trainees' communication skills likely will require significant changes in instruction at both the undergraduate and graduate levels of training. It will also require changes in the approaches for assessing communication skills during residency. The challenges of assessing communication skills should not be underestimated.95 At the very least, assessment should use well-established instruments for measurement of trainees' communication skills in patient encounters.89,103 A prerequisite for ultimate success will be faculty acceptance of the importance of communication skills and a commitment to development. In some settings, implementation of learned skills will require system changes to allow sufficient time in encounters for patient-centered approaches to be used. The time to begin is now. As Levinson and Pizzo22 ^(p1803) have stated, “If the medical profession wishes to maintain or perhaps regain trust and respect from the public, it must meet patients' needs with a renewed commitment to excellence in the communication skills of physicians. It is time to make this commitment.”