Hickson, Gerald B. MD; Pichert, James W. PhD; Webb, Lynn E. PhD; Gabbe, Steven G. MD
Consider the following scenario drawn from actual incidents (details have been changed to preserve confidentiality):
A resident leaves a conference to attend her weekly continuity clinic. As she enters the clinic, she observes an “event” involving an upper-level resident, the clinic receptionist, and a family. The family later reports the event to the medical center’s patient advocate or ombudsman. According to their complaint:
“The doctor my mom was to see entered the clinic acting agitated … talked down to the girl at the desk: ‘Answer my questions immediately with a yes or no … don’t need any extra conversation…. I’m here to see one of my patients.’ Receptionist replied, ‘it’s not a previous patient, but a new patient to be seen.’ Dr. became even more upset …
“Sensing the doctor was in a hurry, I said we’re ready to be seen. Dr. whirled toward me, made a T sign with his hands and barked, ‘Time out! It’s not your turn to talk!’ Turning back to the receptionist, he demanded, ‘Who scheduled me a new patient today?’
“Dr. yelled in my direction so the whole area could hear, ‘You didn’t do anything wrong. The staff did! … I don’t see new clinic patients on Wednesdays … it will be months before you can be seen in my clinic.’
“Then he turned and left us standing there. I don’t think that was very professional.”
Ironically, the conference that the resident observer had just left included cases with scenarios requiring review of core competencies, including patient-centered care and professionalism.1,2 Later in this article, we will refer to this event as the T-sign case because of the upper-level resident’s actions.
Other articles in this collection define and discuss strategies for educating physicians about professionalism. In this article we focus on an alternative, complementary approach that is designed to identify, measure, and address unprofessional behaviors. Our approach has merit for multiple reasons:
1. Most physicians and physicians-in-training behave professionally, so the imposition of mandatory education on professionalism may be seen as personally irrelevant—even insulting. In any case, mandating continuing medical education on any topic has not proved effective.3
2. Measuring professional behavior is challenging.4–6 Besides, what do we do when we see professional conduct except to recognize it as the norm? In contrast, identifying, quantifying, and addressing instances of unprofessional behaviors are both possible and desirable.7–12
3. Despite emphasizing professionalism in medical education, the hidden curriculum exposes trainees to instances of unprofessional behavior (e.g., patients pressured into agreeing with care plans, bias and discrimination, privacy and confidentiality violations, and slanderous comments about others’ failures to disclose harmful errors).13–17 Lest learners unthinkingly emulate their role models, we turn such instances into teaching cases designed to help trainees consider the implications and consequences of unprofessional behavior.18
4. Unprofessional behaviors are associated with poor adherence to practice guidelines, loss of patients, low staff morale and turnover, medical errors and adverse outcomes, and malpractice suits.9,19–23
5. The identification of unprofessional behavior challenges leaders to take action, but many have little or no training in how to address such problems, often allowing unprofessional behaviors to persist, escalate, and spread.7,9
6. Unprofessional behavior stresses any system and begets more unprofessional behavior,15,24 especially if disruptive providers have “protectors” or “enablers” shielding them. Such would-be benefactors provide administrative cover or create “workarounds” instead of constructively confronting the problems, ultimately leading to negative consequences for everyone. Addressing unprofessional behaviors aids in the identification of enablers and protectors and helps identify the problems that really need to be solved.
7. In addition to promoting professional behavior in our VUSM education programs, we have more than a decade of experience developing and evaluating initiatives to identify, assess, address, and teach one another about recognizing and dealing with allegations of unprofessional behaviors. We will share some of that experience in this article.11,12,18,19,25–27
If we consider how the T-sign case might be used in medical education, we can point out several options for the resident observer. As the reader ponders each action, it is helpful to consider certain factors about each choice, namely (1) why a resident (or peer physician or physician leader) might choose it (the potential “pros”), and (2) the down-side challenges associated with that choice (the potential “cons”). Given the case presented, the resident might choose to:
1. Continue walking by (while thinking, This senior resident has been working in the intensive care unit and must be fatigued. That receptionist has probably made a mistake);
2. Informally investigate before taking any action (I must not have all the facts, but something should be done);
3. Approach the parties and offer assistance (Everyone is watching me, including families and continuity clinic staff);
4. After deciding to help the patient or not, approach the fellow resident directly and declare that his behavior is inappropriate (Do you realize how you look and sound?) regardless of the event(s) that precipitated it;
5. After deciding to help the patient or not, follow up with the resident later in private (There must be two sides to the story, but what I observed does not seem consistent with your commitment to professionalism); or
6. Report the witnessed event to the continuity attending or residency director (It’s his/her job anyway, the scheduling problem has been mentioned to the attending before, and the clinic management never really does anything).
We will consider the pros and cons of each choice later when we discuss selected training challenges we pose to our students, residents, and faculty colleagues.
Although episodes of disruptive behavior are not common, neither are they rare.9,11,12,15 Unfortunately, educational programs encouraging professionalism can be effectively torpedoed by disruptive/unprofessional physician models (“You guys are always talking about professionalism, but it is not what we see in this hospital” [personal communication from a fourth-year medical student]).
Promoting professional conduct is not possible without leadership’s commitment to addressing unprofessional/disruptive behaviors whenever they occur, regardless of the rank or title of the physician who behaved poorly.9,19,28 Such commitment therefore requires dedicated leadership, a model or framework for guiding intervention processes, appropriate institutional policies, surveillance tools, training in the “how to’s” for all members of the health care team, and accountability to one another.
Leadership Commitment at VUSM
Leadership’s commitment to professional behavior is exhibited during new faculty orientation sessions when the medical school dean welcomes the new faculty members and emphasizes the role they play in modeling professionalism for graduate and medical students. The Vanderbilt University Medical Center’s (VUMC) chief medical officer introduces credo behaviors for all faculty and staff, and faculty are asked to sign a commitment to know and exhibit these behaviors (http://www.mc.vanderbilt.edu/root/facts/mission.html). Also, the associate deans for undergraduate and graduate medical education promote the importance of providing feedback to medical students and residents about professionalism. Finally, the associate dean for clinical affairs presents Vanderbilt’s professional behavior policy and our systems for capturing, profiling, and providing feedback about allegations of unprofessional behavior and other complaints from patients or staff.
In 2004, recognizing the increasing complexity of the health care environment and the resulting pressures on academic leaders, the medical school dean initiated VUSM’s Academic Leadership Program (ALP) for new VUSM department chairs and, subsequently, all division chiefs and center directors. The ALP is focused at three levels: personal leadership style, team leadership, and the leader as an integral part of the VUSM leadership team. The ALP’s objectives include these specific actions related to addressing unprofessional behavior:
▪ Understanding the core management and oversight role of the leader, resources for information, planning, and problem solving
▪ Gaining insight into the impact of one’s own leadership style and interactions
▪ Understanding and applying leadership practices that align people, resources and partnerships, especially during times of change
▪ Applying problem-solving and decision-making methods to foster enhanced collaboration
▪ Actively valuing wellness, both as an individual and as a leader who supports others’ wellness
▪ Actively valuing, initiating, and supporting mentoring of faculty and students as a leadership practice.
To help fulfill these objectives, ALP training includes an introduction to our model for understanding disruptive behavior and ways to address it (Figure 1). Participant feedback has been extremely positive, suggesting that academic medical center (AMC) leaders respond well to training directly related to their ongoing leadership challenges, including how to deal effectively with unprofessional behavior.
A Process for Guiding Interventions
The disruptive behavior pyramid (Figure 1) puts the challenge of addressing unprofessional conduct in perspective and serves as the foundation for how we train team members to provide feedback to colleagues and subordinates. The base of the pyramid, which conceptually extends far below the illustration, is meant to convey that most health team members conduct themselves as professionals and never, or very rarely, exhibit behaviors that might be perceived as unprofessional. For example, about two thirds of physicians never or very rarely generate unsolicited patient complaints, which are markers of malpractice risk and sometimes unprofessional conduct.11,12 In fact, many of these outstanding physicians routinely exhibit the sort of professionalism that serves as a model for their colleagues and trainees. These exemplars should be recognized and rewarded.
The next block up in the pyramid is labeled single unprofessional incidents. Our experience suggests that, on occasion, a professional will be alleged to have demonstrated such behaviors, and the validity of the allegation may not be immediately clear. To the extent that complaints are a proxy indicator, about 20% to 25% of medical professionals will allegedly behave in ways that dissatisfy their patients.11,12 Such allegations may represent anything from an observer’s misperception, to an isolated event unlikely to recur, to an observer’s first observation of a pattern of behavior. All medical professionals should be empowered and trained to acknowledge and address individual unprofessional incidents such as the T-sign case. In the absence of other information, almost all such events should be initially treated as anomalies that, although unlikely to recur, should nevertheless be the subject of an informal intervention, such as a “cup of coffee conversation,” described later. As the arrow on the right side of the pyramid indicates, however, important exceptions exist, such as when the law mandates reporting the event and/or provides sanctions for engaging in prohibited behavior. Claims of discriminatory behavior, or allegations of sexual boundary violations, substance abuse, or other impairment affecting a health professional’s ability to practice safely require prompt reporting to appropriate authorities for investigation and further action, if indicated.29,30
The next level up in our pyramid represents circumstances in which unprofessional or disruptive behaviors recur. The incidence of persistent disruptive behavior is unknown, but one survey estimates that 2% to 3% of medical staff display disruptive behavior sufficiently often to reach leadership’s attention.31 Pattern recognition can be challenging, so the figure is likely an underestimate. Assessing the actual rate requires surveillance systems such as those described below. Once identified, the pattern must be presented in what we refer to as an awareness intervention. The awareness interventions may be conducted by an authority figure, or, in certain circumstances, by a peer. In either case, awareness interventions should be based on sharing aggregated data that present the appearance of a pattern that sets the professional apart from his/her peers, similar to profiling a physician’s prescribing or utilization data. Our experience compiling and sharing unsolicited patient complaint data with physicians reveals that most respond in a professional manner and make practice and behavioral adjustments to reduce sources of patient dissatisfaction substantially.19,32
Unfortunately, a small proportion of professionals seem unable or unwilling to respond to an awareness intervention. These health care providers represent a threat to quality and safety, and they require what the pyramid terms an authority intervention. Such interventions require leaders to develop improvement and evaluation plans for these physicians with ongoing accountability. Finally, despite the sincere hope to “redeem” these providers via the foregoing series of graduated interventions, failure to respond to the authority intervention should lead to disciplinary action, the tip of the pyramid, including restriction or termination of privileges with appropriate reporting to government entities.
The Vanderbilt Policy
Leadership commitment and process models are obviously necessary but not sufficient to address unprofessional behavior. To aid consistent application of these tools, a professional conduct policy will help. Each institution will need to craft its own policy to create or promote leadership buy-in. For those interested in reviewing the Vanderbilt Professional Conduct Policy, it may be viewed at (http://vumcpolicies.mc.vanderbilt.edu/E-Manual/Hpolicy.nsf/AllDocs/63A7820DAB3B393C862571A8006975A4). Consistent with the disruptive behavior pyramid in Figure 1, the policy conveys expectations, establishes reporting lines, and, perhaps most important, demonstrates the commitment of our highest-level leaders to uphold the ideals of professionalism. The policy may also help encourage and guide appropriate interventions by providing pathways and describing the right things to do.
Policies, models, and even leadership commitment will not be effective, however, if unprofessional behavior goes unreported and unrecorded. Surveillance systems are required.
The most effective surveillance tools for detecting unprofessional behavior are the eyes and ears of patients, visitors, and health care team members. Most studies suggest, however, that both patients/families and health care professionals observe and experience frustration, but are hesitant to speak up or report it.5,7,27,33,34 This reluctance can be mitigated by leadership, policies, and an environment that promotes reporting. For example, at Vanderbilt, inpatients watch a video that encourages communication: “It’s always our intention to provide the highest-quality care, but sometimes we may fail to meet your expectations. When we do, we want you to let us know, so that we can attempt to make it right, as well as to address and reduce the chance that others will experience the same outcomes.” Vanderbilt’s staff employ a service recovery process called the HEART protocol (Hear the person, Empathize with the concern, Apologize if appropriate, Respond with a plan for assessing the allegation and closing loops, and Thank for bringing issue to our attention).35 Vanderbilt’s policy is that physician and resident-based complaints are to be centrally reported to an office of patient affairs (OPA), that managers receive case-based HEART protocol training,35 and that OPA personnel use an electronic database to store the patients’ observations of care—both compliments and complaints (such reporting systems are now available from a variety of vendors). Patient advocates record patient observations, ensure service recovery is attempted, and relay information to the parties involved. Patient advocates also alert leadership about incidents that warrant special attention, which helps promote ongoing accountability.27
In addition, patient/family complaints are routinely coded according to individuals named in the complaint, type of complaint, and location of the event in question. The complaints are then aggregated to identify attendings, residents, and medical center units that demonstrate disproportionate shares of complaints.11,12,19,27,32 Such aggregated complaint reports reliably indicate malpractice risk, and these profiles may be readily reviewed to identify patterns of alleged disruptive behavior involving patients and families.
Members of the medical team may similarly observe and report their perceptions of unprofessional conduct to authority figures directly, or they may use a Web-based electronic risk event-reporting system. The Web-based system allows individuals to report events anonymously or not, but in either case the observer is asked to provide essential details about the event. Reports are reviewed and shared with leadership to promote assessment of the “whys” of the event, to promote overall accountability to the professional conduct policy, and to assist with pattern recognition.
Policies and surveillance systems have value only if they support meaningful action such as those suggested by the graduated intervention process depicted in Figure 1. Because professionals often lack training in the how-to’s of managing unprofessional behavior,36 this section begins a general description of our approach to such training and briefly describes a process we teach for deciding whether and how to proceed with initial interventions. This section concludes by describing a specific intervention program.
Intervention skills training
Table 1 identifies selected training programs we offer, primary training objectives, and target audiences (students, residents, attendings, leaders). In every case, the method of training is interactive, incorporating role play and use of audience response technology, small group learning, case- and problem-based sessions, and participation by those in leadership roles. Because participants may come to training with different experiences in dealing with unprofessional behavior, they commonly challenge their colleagues’ responses to these scenarios. Our training programs, then, emphasize a balance beam approach to problem solving, in which every approach may have pros and cons that must be weighed before taking action. For example Table 2 lists selected pros and cons associated with alternative strategies for dealing with the resident who flashed the T-sign. Training goals therefore include identifying and considering the pros and cons of each approach. Consider how the balance beam approach might be employed in the T-sign case, which we have used in training our leaders and residents. We resist the temptation to offer a one-size-fits-all solution to handling the T-sign situation, because each alternative does have pros and cons that will be differentially weighted by various stakeholders. The important thing is to consider each alternative, then choose. In the T-sign case, assume the upper-level resident had no previous history of such behavior and was otherwise well regarded. In that case, many persons would choose to perform an informal intervention.
Informal intervention: A “cup of coffee conversation”
It seems reasonable that, whenever possible, disruptive behaviors should be addressed, even informally. Training in having a “cup of coffee conversation” is based on principles of sharing bad news.37,38 The training encourages participants to choose a private setting for this informal intervention where there can be a brief review of the observations of the unprofessional event(s) followed by a pause to allow the individual to respond. The colleague or subordinate who has behaved unprofessionally should be invited to offer his or her view of what happened and should be encouraged to reflect on the event and why it may have occurred. Such encounters must be followed, if possible, by an expressed appreciation for the colleague’s contribution to the team. The colleague’s reaction to this discussion may include anger, denial, or rationalization. In our own experience, confronted individuals commonly respond by implying that their disruptive conduct was somehow justified because of someone else’s mistake; in the T-sign example, this is represented by the upper-level resident’s charge that the receptionist is always making mistakes. Trainees are taught to be sensitive to such attempts by the colleague to minimize his or her role and to remind the colleague that, in spite of others’ mistakes, there are professional and unprofessional ways to respond. Training in having a cup of coffee conversation is essential for all other levels of intervention as outlined on the disruptive behavior pyramid. We believe that many professionals who exhibit unprofessional conduct may fail to recognize their own bad behaviors, and a lack of early intervention, like the cup of coffee conversation, allows patterns of behavior to become routine. Because few physicians have been trained in such interventions, they may talk about their colleagues’ unacceptable behaviors to others, but they are often unwilling or unable to sit down and talk with their colleagues directly as a first step in promoting accountability.
In our view, most cup of coffee conversations do not need to be documented. Supervisors, for example, regularly provide informal feedback to those under their charge. Deciding to document informal interventions, however, may be influenced by the nature of the event in question or reporting requirements related to such an event (e.g., when reporting is part of the patient complaint process, CMS guidelines require an institutional response).
When patterns emerge, then persist
Suppose the resident who flashed the T-sign (or any physician at any level of training) exhibits several more instances of unprofessional behavior in the following months. What then? In addition to the principles suggested in the previous section, awareness and authority interventions rely on data demonstrating that the provider’s behavior warrants increasing levels of concern.
First, we teach the importance of considering why a health care provider might continue to exhibit unprofessional behaviors. Understanding the differential diagnosis is essential for effective interventions to be reasonably devised and applied. Too often, the disruptive professional’s behavior is dismissed as a function of personality. Research suggests that personality tests do not identify disruptive professionals,39 and experience and research have demonstrated that the etiologies of unprofessional and disruptive behavior include at least the following six drivers24,28–30,40,41:
1. Substance abuse, psychological issues;
2. Narcissism, perfectionism (inappropriate handling of one’s feelings of guilt or shame when things don’t go perfectly), or selfishness;
3. Spillover of chronic or acute family/home problems;
4. Poorly controlled anger—especially under heightened stress—perhaps attributable to
a. Poor clinical/administrative/ systems support,
b. Poor practice management skills or challenges related to poor performing colleagues, or
c. Care providers (any profession) whose constant criticisms of others create poor practice environments;
5. Bad behavior gets desired results, so the behavior is rewarded; and
6. Clinical and administrative inertia (no one addresses the unprofessional behavior earlier and others in the environment may behave similarly, so the behavior becomes “normal” and accepted for the individual).
The proportions of disruptive professionals whose behavior may be accounted for by each factor, or a combination of these, is unknown, but, regardless of their driver(s), these physicians represent a threat to quality and safety, and they require what the pyramid terms an authority intervention.
The Patient Advocacy Reporting System (PARSSM) process.
The PARS process begins with systematic, reliable analysis and profiling of patient/family complaints. It is designed to recognize patterns, and it goes beyond the informal coffee cup conversation described earlier. In addition:
▪ A Patient Complaint Monitoring Committee (PCMC) composed of physician peers or, if preferred, authority figures, is established under the applicable peer review and quality-assurance statutes to provide protection from legal discovery. Confidentiality, respectful attitudes, and supportive behaviors are essential in the PARS process.
▪ PCMC members receive six to eight hours of training (see Table 1) and practice, including role play exercises, sharing sensitive information with colleagues who have a record of many complaints against them, and responding to their colleagues’ varied reactions.
▪ Most institutions adopt three levels of interventions in the PARS process: level one describes a confidentially delivered, nonpunitive peer awareness intervention; level two is the authority intervention and requires development of a specific action plan; and level three is a disciplinary intervention that occurs after review by the institution’s highest level of administration.
▪ Using patient complaint records drawn from the institution’s database, VUMC creates a patient complaint profile for each “high-complaint” physician, and the profile depicts each physician’s complaint score relative to other group members. Assembled data portray the types of complaints that stand out (see Figure 2 and Table 3 for hypothetical “Dr. X”).
About 60% of physicians demonstrate improved complaint scores after level one awareness interventions.19 Recidivism exists, but it is rare: less than 2%. Other high-complaint physicians are mobile: about 20% leave the institution because of relocation or retirement. Another 20% require additional help at level two authority interventions to improve.19
Selected results of interventions on attending physicians.
How might AMC leaders recognize when a physician’s disruptive behavior recurs? Patient complaints recorded by an institution’s office of patient affairs can be analyzed systematically to identify accurately physicians at increased risk of being sued.11,12,19 Only a fraction of patient complaints point to disruptive behavior as defined in VUSM’s policy, but all of them reflect a disruption (or potential disruption) in the physician–patient relationship, thus increasing the risk potential. Resulting complaint profiles support interventions designed to reduce that risk. Our PARS team’s experience with more than 800 initial and follow-up interventions since 1997, in both community and academic medical centers, shows reductions in patient complaints and malpractice claims rates.19
Selected results of interventions on resident physicians.
The PARS process may also be applied to residents at any AMC whose patient-relations representatives accurately identify and record the residents associated with patient complaints. For example, figures similar to Figure 2, but based on the distribution of residents’ complaints, have been used to provide nonpunitive, information-only, individualized presentations to a small number of residents by respected attending physician faculty members at one AMC. These resident interventions, although few in number, have reportedly been very positively received. One-year postintervention follow-up showed no, or many fewer, complaints about residents still at the institution (the others had been in their last year of residency at the time of the interventions) (unpublished data).
The PARS process currently addresses patient/family complaints. The same intervention processes, however, might be employed when concerns about behavior are reported by staff, referring physicians, vendors, residents, students, or others who have business with the AMC. Despite barriers to reporting and taking action, when institutional leadership is committed to reducing and eventually eliminating unprofessional behavior, policies and surveillance systems are in place, and training has been effective, these barriers can be overcome.
Institutional Resources for Dealing with Unprofessional Behavior
Despite the foregoing programs of education and intervention, a small proportion of physicians persist with unprofessional behavior that disrupts patient and peer working relationships. Because these physicians are unable to make behavioral changes on their own, their deans and their department chairs must therefore require change, suggest sources of assistance for making those changes, and hold the physician accountable for changing. If the unprofessional behavior nevertheless persists, disciplinary action leading to termination may result. Fortunately, our leaders have access to substantial support services, including
▪ A Joint Commission on Accreditation of Health Care Organizations-mandated Physician Wellness Program led by a skilled internist (http://www.mc.vanderbilt.edu/root/vumc.php?site=cph&doc=683);
▪ A Comprehensive Assessment Program for Professionals (http://www.mc.vanderbilt.edu/root/vumc.php?site=vcap) that involves individualized medical and psychological evaluation and treatment planning; and
▪ A Center for Professional Health (http://www.mc.vanderbilt.edu/root/vumc.php?site=cph) offering group classes for physicians whose unprofessional behavior includes misprescribing controlled substances, crossing sexual boundaries, or classic disruptive behavior.
Finally, just as the persons who have been disruptive may require attention, so may those whose work has been disrupted. Service recovery for staff, patients, and other physicians should be considered and implemented.
The Benefits and Responsibilities of Addressing Unprofessional Behavior
Why bother addressing unprofessional and disruptive behavior? After all, AMC leaders face a host of competing priorities for their time and, of course, money. All elements of the infrastructure described above must be supported for the overall program to work. The principal cost of successfully implementing this program is leadership and faculty time that might otherwise be spent generating grants or patient care income, providing service, and teaching other elements of the crowded curriculum. Additional expenses include surveillance systems and personnel to implement them, professional development training programs, and costs associated with service recovery (making right what was wrong). These costs will vary by institution. For example, many hospitals with active patient/guest-relations programs have one FTE patient advocate/ombudsman per 100 to 150 hospital beds, and faculty time for planning and teaching (formally and informally) the programs like those listed in Table 1 can be variously estimated.
Although the costs of such programs are important considerations, an equally important calculus involves the costs of failing to teach about and deal with unprofessional behavior. Therefore, besides being the right thing to do, and despite the significant challenges, dealing with disruptive behavior has the potential to yield several important cost-saving benefits, including
1. Improved staff satisfaction and retention;
2. Enhanced reputation for the AMC and its leaders;
3. Creation of a culture of professionals who are important role models for students, residents, staff, and one another;
4. Improved patient safety attributable to greater staff willingness to speak up when they observe problems in patient care;
5. Reduced liability exposure and risk-management activity; and
6. Overall more productive, civil, and desirable work environments.
In summary, the challenges for leaders in academic medicine are to think not only about how best to promote professionalism, but simultaneously to renew our commitment to addressing unprofessional or disruptive behaviors whenever they occur, to adopt or adapt a framework for understanding approaches to and processes for dealing with unprofessional behavior, to develop or revisit behavior-related institutional policies, to employ surveillance tools, to train our leadership teams on monitoring and intervention, and to become truly accountable to, and for, one another.
The authors appreciate the editorial assistance of Rick Moore and Anna Caruso Hayden. This work was supported, in part, by the Vanderbilt Center for Patient and Professional Advocacy, the Vanderbilt Office of Risk and Insurance Management, and the Office of the Dean, Vanderbilt University School of Medicine.