Patients’ Perspectives On Ideal Physician Behaviors

What constitutes a good doctor? Is technical proficiency sufficient to be a good doctor? Clearly, a physician cannot lack necessary technical knowledge and skills and still be a good doctor. Less clear is whether a technically proficient physician can lack interpersonal skills necessary to relate well to patients and still be a good doctor. The answer to this question must incorporate the views of patients, the users of health care services.

The literature suggests that physicians’ interpersonal skills are critical to establishing strong, trust-based physician-patient relationships that offer multiple benefits. Most patients want a strong relationship with a primary care physician.^2–5  Not surprisingly, strong physician relationships appear to assume even greater importance during periods of serious illness.^4–8  The quality of a patient’s relationship with a physician can affect not only a patient’s emotional responses but also behavioral and medical outcomes such as compliance⁹  and recovery.¹⁰  Consequently, physicians have been urged to improve their communication¹¹  and patient education¹²  techniques, develop their empathetic abilities,¹³  encourage participative decision making,¹⁴  and convey respect¹⁵  and dignity.¹⁶  Barrier et al.¹⁷  stress patient-centered medical interviewing to encourage patients to disclose their true concerns and to do so early in the visit. They advise physicians to ask “What else?” questions. The pervasive interest in the topic of what a physician should be extends beyond the physician community and is reflected in the many patient-oriented websites and blogs dedicated to it, for example, doctorandpatient.blogspot.com, familydoctor.org and physicianreports.com.

Physicians seeking to improve the quality of their interactions with patients and medical schools seeking to best prepare new physicians would do well to focus on the specific behaviors that drive patient satisfaction. Focusing on behaviors is consistent with competency-based assessment, helps physicians judge whether they are living up to the ideals, and reflects how physicians actually practice medicine.¹⁸  Thus, a profile of ideal physician behaviors could serve as a training platform, an assessment model for health care professionals, a prototype for educator role-modeling skills and even an assessment tool for admission to medical school.

Numerous articles address aspects of the physician-patient relationship, but many are limited in guiding the behavior of practicing physicians. Limitations include segmentation, scope and solution issues. Segmentation issues occur when specific patient populations are studied, such as patients with depression,¹⁹  those facing end of life,²⁰  and physicians as patients,²¹  raising concerns of generalizability. Scope issues occur when such specific physician characteristics are studied as empathy²²  or participative decision-making style.²³  Although useful, such studies do not address the totality of behaviors considered by patients. Solution issues occur when practical steps that physicians can take to strengthen patients’ perceptions of their service are not identified.

Our goal in this research is to address these limitations and develop a basic profile of ideal physician behaviors from patients’ points of view that could be beneficial to physicians in meeting patient needs. In addition, we discuss the implications of our findings.

This study was part of a six-month investigation of the patient experience at Mayo Clinic in Scottsdale, Ariz., and Mayo Clinic in Rochester, Minn., during 2001 and 2002. The Mayo Foundation Institutional Review Board approved the research protocol. To strengthen generalizability, we selected a random sample of patients who had been recently served in one of 14 medical areas: cardiology, cardiac surgery, dermatology, emergency medicine, endocrinology, executive medicine, family medicine, gastroenterology, medical and radiation oncology, neurology, orthopedic surgery, transplant surgery, thoracic surgery and urology. The disciplines were selected to provide a broad representation of inpatient and outpatient services and various levels of acuity.

Within each discipline, equal numbers of potential respondents were selected at random and informed by mail that they might be called to participate in a telephone survey. The advance notification assured patients the survey was an attempt to improve patient service quality, all responses would be kept confidential, no Mayo Clinic physicians would be given the names and responses of any participant and they could decline participation without jeopardizing their relationship with their physicians at Mayo Clinic. Additionally, the notification assured participants that the interviewer did not have access to their medical records and would know only they had been seen within a particular specialty. The random selection of participants was to ensure that patients from all study disciplines would be represented and to increase confidence in the generalizability of patients’ responses.

To assure patients of the confidentiality of their conversations, the interviews and analysis of transcripts were conducted by two of the authors who have no formal ties to Mayo Clinic (N.M.B. and L.L.B.). Three attempts were made to contact every third individual on the patient list before substituting another contact. Telephone surveys were completed with an average of 14 patients in each group for a total of 192 interviews, split almost evenly by sex. The telephone interviews lasted between 20 and 50 minutes and focused on the behavioral characteristics most desired in a physician. The questions included the following:

1. Tell me about the best experience that you had with a doctor in the Mayo system.

2. Tell me about the worst experience that you had with a doctor in the Mayo system.

3. Now think about all the people you interacted with at the Mayo Clinic other than doctors. Tell me about the best experience that you had with any one of them.

4. Tell me about the worst experience that you had with any one of them.

5. If you could make one change to improve the quality of patients’ service experience at Mayo Clinic, what would it be? Why?

To encourage rich, detailed descriptions of physician characteristics most salient to patients, respondents were asked to provide specifics on their best and worst experiences with a Mayo Clinic physician. They were not asked about physicians outside Mayo Clinic. Respondents ranged from long-time Mayo Clinic patients to first-time patients. They could refer to any Mayo Clinic physician experience and were not restricted to the most recent visit. Accordingly, respondents did not always refer to a physician practicing in the specific department from which the sampling was done.

After the completion of all interviews, two of the authors (N.M.B. and L.L.B.) independently reviewed each interview transcript. Next they developed a list of themes that captured the discrete behaviors mentioned by respondents. The two researchers then compared their lists to identify areas of agreement and resolve any disagreements. They reached consensus on all themes. The two researchers then separately coded each mention of a theme by a respondent. Comparison of coding decisions yielded inter-rater reliabilities of 0.92 and 0.94 for best and worst service descriptions, respectively. They discussed points of disagreement and reached consensus on all items.

To check the reliability of the themes that emerged, two additional judges, doctoral students unconnected with the research project, were recruited as coders. They were asked to review the transcripts, code responses by theme and identify any new themes. No new themes emerged. The additional judges were in almost perfect agreement in their coding and resolved the few discrepancies that surfaced.

Seven ideal physician behaviors (behavioral themes) were identified in the research: confident, empathetic, humane, personal, forthright, respectful and thorough. These behaviors are consistent with existing research,^6,24–30  offer a balance of breadth and specificity and hold promise for empirical investigation. Definitions of these behaviors and representative quotations are presented in Table 1. All definitions are from the patient’s perspective.

All 192 respondents could describe a “best physician” experience and frequently referred to more than one behavioral theme in their descriptions. Of the total sample, 43 mentioned one behavioral theme, 138 mentioned more than one, and 11 responded in a manner too general to allow the identification and coding of any specific behaviors. The most frequently mentioned theme was “thorough” and the least frequently mentioned was “empathetic.”

Only 89 respondents could describe a “worst physician” experience. Their responses reflected mirror opposites of desired physician behaviors, especially perceived insensitive, disrespectful behavior. The concerns expressed about poor service ranged from a physician’s arrogance in dismissing the patient’s input, disinterest in the patient as an individual, impatience in answering a patient’s questions to callousness in discussing the patient’s prognosis. Concerns also were expressed about physicians who seemed to have provided excellent service in the original visit but who failed to meet expectations they created about the speed or quality of follow-up service. A caveat in interpreting this finding is we report neither measures of the relative importance of various dimensions in affecting patients’ perceptions of physicians nor measures of the relative impact of positive vs negative perceptions. This latter issue is especially important because positive and negative perceptions need not be symmetrical in their effect on patients’ perceptions.

Medical services are different from other services. Unlike “want” services, such as telecommunications and entertainment, health care is a “need” service patients often dread. Serving a customer who arrives with some combination of illness, pain, anxiety and fear presents a distinct service challenge to physicians compared with other services. Medical customers are inherently under stress. Moreover, medical services are highly complex and technical. The patient is at a considerable knowledge disadvantage and has little choice but to trust the physician to perform the right service in the right way. Medical services are usually inseparable in that patients must be physically present to receive the service. For inpatient services, the patient not only visits the service “factory” but also lives in it. Few service industries provide beds for their customers; health care is one that does.

Medical services are inherently personal. Other services do not require customers to bare themselves physically and emotionally like medical services. Medical services also are personally important. Rarely are the stakes as high as they are for the medical services customer. A mistake in diagnosis, treatment plan or procedure can do great harm to the patient. The patient’s quality of life — and life itself — may be in the hands of the physician.

In evaluating a service as anxiety producing, complex, proximate, personal and important as medical care, patients are particularly attentive to what they can see and understand to interpret what they cannot see and understand. Technical quality is often difficult for patients to assess even after the service is performed. This helps explain why the physician’s technical competence was rarely mentioned by the patients we interviewed. Our open-ended questions, for example, “Tell me about the best (worst) experience that you had with a doctor in the Mayo system,” did not preclude respondents from referring to matters of technical proficiency, but they rarely did. How physicians provided the service (how they behaved) was not only important to patients but also easier for them to judge than technical quality. Patients can sense if the physician is rushed, preoccupied, tired, aloof, disinterested or alarmed just as they can sense a physician’s genuine interest, compassion, calmness and confidence.

The research by Fung et al.³¹  indicates if patients are forced to choose between technical quality and interpersonal quality in selecting a primary care physician, most will choose the more technically proficient physician. Their study design involved giving respondents information on physician technical and interpersonal quality in the form of simulated, computerized health care report cards. Even with this information available, a substantial proportion of the respondents (approximately one-third) selected physicians high in interpersonal quality. In reality, absent a report card patients often cannot conclusively determine the level of technical proficiency and appear to expect adequate proficiency as a given. They are most likely to judge what is important to them based on what they are able to judge. Our respondents relied on what they could see and understand in evaluating physicians, and most evaluated the physician’s behavior.

The nature of medical services turns patients into “detectives” looking for “clues” to reassure themselves of their caregiver’s competence and caring. Specific clues carry messages, and the clues and messages converge to tell a service’s story to the customer. In business parlance, managers are advised to coordinate the clues of service to establish a coherent quality image.³²  Such advice is no less relevant to physicians.

Clues are generally in three main categories³³ : functional, mechanic and humanic. Functional clues concern the technical quality of the offering. Anything that indicates or suggests the technical quality of the service — its presence or absence — is a functional clue. A missing laboratory report is a negative functional clue, and checking on a patient’s drug allergies before prescribing a drug is a positive one. Mechanic clues come from tangibles in the service experience and include sights, smells, sounds, tastes and textures. The comfort, orderliness, cleanliness, modernity and noise levels of medical facilities illustrate mechanic clues. Humanic clues come from the behavior and appearance of physicians — choice of words, tone of voice, level of enthusiasm, body language, neatness and appropriate dress. A physician who sits while conversing with a patient gives a different humanic clue than one who stands at the doorway.

Functional, mechanic and humanic clues play specific roles in creating the service experience. The clues interact, and perceptions of one type of clue can influence perceptions of another type. Functional clues are the “what” of the experience, carrying messages about the reliability and competence of the service. Mechanic and humanic clues are the “how” of the experience, revealing much about an organization’s and individual clinician’s commitment to genuinely being of service. Research outside medicine shows mechanic and humanic clues primarily influence customers’ emotional or affective perceptions of the service experience.³⁴  The research reported herein, restricted to patients’ perceptions of physicians, reveals the strong influence of humanic clues on patients’ memories of their best and worst physician experiences.

Although the technical ability of a physician is vital, it is important for physicians to learn and demonstrate interpersonal skills as well. Most service organizations invest in developing the interpersonal skills of their employees who interact with customers. It is difficult to imagine a service in which these skills are more important than medical service. Nonmedical service providers rarely perform the kind of extreme service roles many physicians commonly perform, for example, simultaneously caring for an acutely ill patient and communicating with anxious family members, informing a patient of a terminal disease diagnosis, finding the right way to advise patients sensitive about their weight they must lose weight, or telling a patient it is necessary he or she stop driving a vehicle.

The challenge of being interpersonally effective in performing medical services requires concerted attention in medical schools and suggests an important continuing education opportunity for professional medical societies and medical educators. Students, residents and physicians in practice should seek to understand how they are perceived by patients and what the literature reveals patients desire from their physicians. Then, they should hone their skills in demonstrating desired behaviors to their patients. Learning obtained from such courses as nonverbal communication, active listening and sharing bad news could improve patient perceptions of physician behaviors.

Physicians need to be clue conscious in how they provide service, and this, too, represents an educational opportunity. The management of customer experience is being taught more frequently in business schools, and it could be taught to physicians and medical students as well. Table 2 illustrates how physicians can effectively manage humanic clues in support of the ideal physician behaviors. The clue examples described are from the interview transcripts, from patient focus groups also conducted at Mayo Clinic, and from the personal observations of the research team. Educational sessions that incorporate role playing and participant feedback can be built around the ideal physician behaviors and clue management concepts.

Recently, there has been an increased focus in medical education on key competencies necessary for physicians. The Project Professionalism of the American Board of Internal Medicine has outlined specific values, including humanistic and communication behaviors, that are expected of their membership.³⁵  The Outcome Project of the Accreditation Council for Graduate Medical Education now requires all accredited residency programs to address the training of physicians in six core competency domains: patient care, medical knowledge, practice-based learning and improvement, interpersonal and communication skills, professionalism and systems-based practice. Residency programs must define the specific knowledge, skills and attitudes required in the domains, provide educational experiences as needed for their residents to demonstrate competency in each, and assess resident performance throughout the training experience.³⁶  Three of these required competency sets — patient care, interpersonal and communication skills and professionalism — are directly related to the ideal physician behaviors described in this article.

Patient feedback can clarify which physician behaviors have the greatest impact on patient satisfaction. Medical practices should consider periodically surveying patients to identify opportunities for strengthening patient relationships.³⁷  The surveying need not be elaborate or expensive. Much can be learned from an anonymous one-page survey including salient demographic categories and three openended questions: “What do you like most about your main doctor?” “How can your main doctor improve service to you?” “How can our medical practice improve service to you?” Survey results can be the basis for organizing physician (and staff) educational sessions on behavioral skills most desired by patients.

Finally, these seven ideal behaviors can be a useful tool for physician executives working with physician performance problems. The usual level of feedback a supervising physician has to work with is vague, such as “he/she has a poor bedside manner” or “this doctor is not listening to me.” The behavioral categories and illustrative humanic clues in Table 2 give the physician coach a set of specific factors to assess performance and then subsequently guide a remedial educational plan for the identified physician. Trends of deficiency among several physicians in the medical group, health plan, or hospital medical staff can provide opportunities for continuing medical education to improve team performance.

Interviews with 192 patients receiving medical care within 14 medical specialties reveal a profile of seven ideal physician behaviors: confident, empathetic, humane, personal, forthright, respectful and thorough. These behaviors are not hypothetical; they are from the voices of patients who had a recent medical service experience at Mayo Clinic and were asked to describe their best and worst experiences with a Mayo Clinic physician. That patients discussed their physicians’ behavior rather than their physicians’ technical ability does not suggest technical skills are less important than interpersonal skills, but it does suggest the former are more difficult for patients to judge. The findings may also reflect patients’ inclination to assume a physician is competent unless the absence of competence is demonstrated.

The educational implications for preparing practicing physicians and physicians in training to become and remain interpersonally effective under the most challenging of circumstances are far-reaching. The ideal physician behavioral profile presented in this article is comprehensive and yet reasonably definitive; it can be used as a training and assessment framework. The concept of clue management may be a useful tool for transforming the broader behavioral themes into discrete behaviors positively affecting patient perceptions of the service experience.

That our data are qualitative and were gathered from patients at one institution (the main Mayo Clinic campuses in Minnesota and Arizona) are limitations of the research. Several aspects of our study ameliorate these limitations to a degree. First, the study is exploratory and in such designs it is conventional to use qualitative methods because the purpose is to reveal key themes and concerns rather than to make statements about prevalence of phenomena. Second, we took care to include 14 different practice areas, representing varying degrees of medical severity to incorporate a diverse patient base. Finally, the patients came from various parts of the United States and typically had gone to other health care institutions before seeing a physician at Mayo Clinic. Thus, our respondents drew from a rich experience base when commenting on Mayo Clinic physicians.

All the behaviors identified in our research were discussed by multiple patients; however, we did not assess the relative importance of each dimension beyond frequency counts or investigate whether there is variance in the ideal dimensions based on demographic variables. This was beyond the scope of our study, but it is an important avenue for future research in the area.

After one of the patient focus group interviews conducted at Mayo Clinic, a participant sent a handwritten note. This patient with breast cancer captures the powerful role of humanic clues in medicine and, in so doing, offers a fitting ending to our article.

Direct research expenses for this study were provided by Mayo Foundation.