EMOTIONAL INTELLIGENCE COMPETENCIES IN PHYSICIAN LEADERS: AN

EXPLORATORY STUDY

Michael J. Deegan, M.D. Executive Doctor of Management Program Weatherhead School of Management ( WSOM ) Case Western Reserve University ( CWRU ) Cleveland, Ohio Telephone: 248-540-3130 ( Birmingham, MI, USA; EDT ) E mail: deeganme@aol.com PROJECT ADVISORS: John D. Aram, Ph.D., Professor, WSOM, CWRU Richard E. Boyatzis, Ph.D., Professor, WSOM, CWRU Christopher P Mulrooney, Ph.D. Consultant, Competency Development & Application, Worldwide; Hay Group, McClelland Center for Research & Innovation Boston, MA

2

ABSTRACT A theory of behavioral competencies associated with emotional intelligence that influence job performance, was tested in physicians who hold leadership positions in their organization. An emotional intelligence competency ( EIC ) survey ( the ECI-2 ) that measured eighteen competencies distributed among four cluster groups ( self awareness, self management, social awareness, relationship management ) was administered to physician leaders and their supervisors, peers and direct reports. The resulting 360 degree assessment permitted a comparison of the physician leaders self assessment scores with one another and those working closely with them. In addition, independent objective and subjective measures of physician leadership effectiveness were available for comparative evaluation. Eleven of thirty-four physician leaders who volunteered to participate had complete profiles and five others had several surveys submitted. One hundred and two self and other surveys form the basis of this exploratory study. Descriptive statistical analysis revealed results restricted to the upper half of the five point Likert scale. Participating physician leader scores often exceeded mean scores from peers, direct reports and supervisors. Specific EICs clustered at the upper and lower ends of the distribution. However, the results range was sufficiently restricted only relative differences were observed. EIC profile results of outstanding physician leaders ( based on independent performance measures: patient satisfaction, clinical quality and productivity, financial performance, innovation, and conflict management ) did not differ appreciably from typical colleagues when compared using a paired t test. The results of the analysis are discussed and refinements applicable to future studies considered. The association of particular EICs and leadership styles is also addressed in the context of the participating physician leader results. KEYWORDS: physician leaders, behavioral competence, emotional intelligence, 360 degree assessment, outstanding & typical performance, leadership styles

3

INTRODUCTION Emotional intelligence, i.e., the ability to monitor one’s own and other’s emotions, to discriminate among them, and to use this information to guide one’s thinking and actions ( Salovey & Mayer, 1990 ) has emerged as an important leadership attribute ( Boyatzis, 1982; Goleman, 2000; Goleman, 2001b; Goleman, 2002 ). Emotional intelligence is made manifest through the use of emotional intelligence competencies, which are underlying characteristics of an individual causally related to effective or superior performance in a job ( Boyatzis, 1982; Spencer & Spencer, 1993 ). Emotional intelligence is essential for leaders attempting to cope with an increasingly complex and fast-paced world. Some of the greatest challenges facing organizations and the people who work in them are learning to cope with discontinuous change, managing massive amounts of data, and working together more effectively ( Cherniss, 2001 ). Emotional intelligence and its manifestation as specific behavioral competencies influences and enhances individual and organizational effectiveness in dealing with these issues. There is a body of evidence suggesting leaders and managers with a well developed complement of emotional intelligence competencies are more effective than peers ( with similar education and experience ) who lack a highly refined set of these qualities ( McClelland, 1998; Boyatzis, 2000; Goleman, 2001b ). Goleman ( 2001b ) cites studies by a consulting firm ( the Hay Group ) that measured financial performance among comparably sized insurance companies. Firms led by chief executive officers who exhibited outstanding emotional intelligence competency levels had better financial performance as measured by both profitability and growth. In another study ( Goleman 2001b ) senior health care executives were more adept at integrating key emotional

4

intelligence competencies such as organizational awareness, and relationship skills such as influence, into their professional activities. McClelland ( 1998 ) demonstrated a link between superior behavioral competency levels and success in earning performance bonuses among senior executives in an international consumer products company and Boyatzis et al. ( 2000 ), while studying partners in a large consulting firm, observed that senior partners with emotional intelligence competency profiles reflecting high levels of self regulatory, self management and social skills consistently contributed higher levels of profitability when compared with their “typical” colleagues. The contribution of emotional intelligence to effective job performance increases as an individual assumes greater responsibility ( Fernandez-Araoz, 2001 ). In circumstances where all incumbents have the requisite cognitive and technical knowledge and relevant experience, the outstanding leaders complement these abilities with maturity, self awareness, empathy, active listening skills and other characteristics that make a meaningful difference in their performance and contribution to the success of the organizations where they work. Much of the literature on emotional intelligence and its accompanying behavioral manifestations has been drawn from studies in business and related industries ( Boyatzis, 1982; McClelland, 1998; Goleman, 1998; Cherniss, 2001; Goleman, 2001b ). There is a paucity of information regarding emotional intelligence competencies and how they influence practitioners and leaders in medicine and related health care professions ( Carrothers, 2000; Bellack, 1999 ). The available literature focuses on the physician – patient relationship rather than the provider’s colleagues and co-workers or the organization where care is delivered ( Ramsey, 1993; Novack, 1997; Violato, 1997; Duberman, 1999; Epstein, 1999 ). The contribution of emotional intelligence to more effective physician leadership remains unexplored. While one might hypothesize physician leaders will demonstrate a repertoire of emotional intelligence competencies similar to managers and professionals

5

in other fields there are good reasons to explore and validate this hypothesis ( LeTourneau & Curry, 1997; Guthrie, 1999 ). Guthrie ( 1999 ) summarizes several of the differences frequently observed when physicians and managerially trained health care executives are compared. These differences are summarized in the following chart. Differences Between Physicians & Executives Physician Executive • Expert mindset • Planner • “Do-er” • Designer • Works one-on-one, in sequence • Organizer • Values autonomy • Values collaboration • Conservative / reactive decision maker • Team player • Success = intrinsic results such as • Success = surrogate measures patient-physician relationships, problem such as financial benchmarks solving, income generation • Little or no business training or • Formal management education experience • Results demonstrated in patient • Results demonstrated in outcomes and quality care quantitative measures ( Guthrie, 1999 ) Guthrie ( 1999 ) describes a number of implications that may emerge from these distinctive profiles including: (1) barriers to effective communication between physicians and executives; (2) differing views of the importance of organizational vs. individual objectives; (3) different views regarding the relative importance of patients and the care they receive; (4) a delayed, often conservative response to the need for organizational change among physicians; and (5) direct economic conflict between reasonable objectives of a hospital or health system and the individual success of physicians. Physicians and health care executives often view the same issue through distinctly different lenses and, as a result, arrive at different causal explanations and select alternative corrective actions.

6

The intent of this project was to conduct a preliminary, exploratory study examining the influence of emotional intelligence competencies on the effectiveness of physicians in formal leadership roles. A key objective was to gain an understanding of how important these attributes were to physicians in leadership roles and whether these competencies discriminate outstanding from typical leadership performance. This knowledge could be instrumental in the development of contemporary physician leaders and the selection, education and mentoring of future physicians and their leaders. Effective planning, implementation, and delivery of health care services in the United States is dependent upon current and future physician leaders’ ability to work effectively with others and see societal as well as an individual perspective regarding their profession and its import. CONTEXT At present health care delivery in the U.S. is in turmoil. During the past twenty-five years dramatic changes in the financing and delivery of health care have combined with remarkable scientific progress to create a paradoxical situation. On the one hand a majority of Americans enjoy access to the most advanced, technologically sophisticated health care in the world, while at the same time more than forty million Americans lack basic, systematic care because they do not have health care insurance and, therefore, cannot gain access to the delivery system except in emergent or life-threatening situations. Throughout this period American physicians have been under duress. Changes in the methods of reimbursement for services rendered, challenges to their clinical autonomy from managed care insurers ( Herzlinger, 1997 ), the legitimization of alternative medicine ( Saks, 1995 ), and the emergence of better informed, more inquisitive patients seeking more influence on care decisions ( Schneider, 1998 ), have each contributed to a sense of alienation and loss among American physicians. Physician

7

morale in the U.S. is at its lowest level in decades. Increasing numbers of middle-aged and older physicians are retiring prematurely. Applications to U.S. medical schools have declined for the fourth consecutive year ( Barzansky, 2000 ). The hegemony physicians held over the American health care delivery system is broken. The model that dominated the delivery of medical and surgical services for more than fifty years has been shattered without an attractive alternative to replace it. Despite this dilemma, physicians continue to provide medical and surgical services in increasingly complex care settings and delivery models. Many have accepted salaried positions and work as highly compensated professional employees in large health care delivery systems. Contemporary care is often planned, organized and delivered by a team rather than an individual physician. These changes require new competencies in collaboration and teamwork. Increasingly diverse patient populations and co-workers require enhanced sensitivity to cultural differences. Decisions regarding the most effective use of limited resources benefit from the ability to listen to alternatives with an open mind and to constructively negotiate creative solutions that optimize resources and outcomes. All of these practices require the exercise of emotional intelligence and the demonstration of related competencies. To be effective in this setting, physicians and their leaders must work to improve their understanding and use of emotional intelligence competencies in a manner similar to the way they embraced the explosion of scientific knowledge during the twentieth century. Unfortunately, too little attention has been devoted to this subject by the medical profession when selecting leaders. The latter are virtually always chosen from within the discipline; whether it is family medicine, cardiology, neurosurgery or any of the other twenty to thirty recognized medical and surgical specialties. Depending on the nature of the organization, these nascent leaders are usually

8

accomplished, recognized practitioners in their professional community. They may be highly regarded by peers for their professional or technical achievements or they may have the largest patient referral base or extraordinary success in obtaining resources from foundations and grant-making agencies. For the last five decades the keys to securing a physician leadership position were evidence of success in the practice field and integration and demonstration of the normative values and beliefs most important to physician colleagues. Emotional intelligence, when considered, was a secondary factor in the selection of most physician leaders. Nowhere is the need to build awareness of emotional intelligence and its related competencies more important than with contemporary physician leaders who need these skills if they are to be effective role models and representatives for the physician community. Unfortunately, there has been little systematic study of physician leaders and the characteristics that discriminate highly effective performers from those who are excellent practitioners or clinical investigators but falter when thrust into a complex leadership role where their clinical skills are not sufficient for the task ( LeTourneau & Curry, 1997; Guthrie, 1999 ). This study was intended to begin to uncover some of the traits that distinguish outstanding from average or typical physician leaders. The study was designed to acquire preliminary information on the emotional intelligence profiles of practicing physicians who also hold formal leadership roles in their organization. There is a dearth of reliable published information on this topic. Furthermore, the selection, education, and socialization of physicians is sufficiently distinct ( Becker, 1961; Moore, 1970; Friedson, 1975 ) it may be imprudent to assume those selected for managerial and leadership roles will be more similar than different from professional managers. For example, many physicians adopt a “pace-setting” leadership style; one that emerges from the individual contributor / practitioner model prevalent in American society

9

and professional practice throughout the twentieth century. This style has been shown to be one of the least effective styles for leading in most contemporary settings ( Goleman, 2000; Goleman, 2002 ). Key emotional intelligence competencies that underlie this style include initiative, conscientiousness and a drive for achievement; all features characteristic of many physicians in the U.S. Absent or under-represented are empathy, relationship building, communication, collaboration and team leadership – each a key competency demonstrated by leaders who adopt the more effective affiliative, democratic or visionary styles ( Goleman, 2002 ). The key question this study addressed is whether emotional intelligence competencies were correlated with outstanding leadership performance by physicians in managerial roles. Variations in the emotional intelligence competencies of participating physician leaders ( measured by the ECI-2 survey instrument ) were the independent variables, while leadership effectiveness, as determined by a set of independent, objective measures regularly utilized by the sponsoring organization, was the dependent variable. Additional research questions included: Are there differences in the patterns and / or levels of competencies demonstrated by outstanding and typical physician leaders? Do particular competencies or clusters of competencies correlate with specific outstanding performance as determined by independent measures of leadership effectiveness?

10

METHODS & DESIGN Study site. The study was conducted in a large, vertically integrated health care delivery system in the north-central U.S. The organization’s principal business units include multiple hospitals, a large health maintenance organization, and an academically oriented medical group employing nine hundred to one thousand full and part time physicians and thirty-five hundred other staff members. The medical group operates and staffs numerous ambulatory care centers throughout its service area. Participants. The medical group operates as a semi-autonomous business unit and is led by a physician chief executive officer. Regional medical directors and discipline specific ( surgery, medicine, pediatrics, etc. ) department chairs report to the CEO. Each ambulatory center ( approximately forty ) has a physician-in-charge who is accountable to the regional medical director. The traditional departments have one or more division heads accountable to the department chairs. There are approximately twenty-five chairs and regional medical directors, forty physicians-in-charge, and twenty to thirty division heads. Study participants were all volunteers ( see the Institutional Review Board approval process described later in this section ). Most of the participants ( > ninety percent ) were either physicians-in-charge or division heads in ambulatory centers. They are the first tier of physicians holding formal leadership positions. All maintained an active clinical practice in addition to their managerial responsibility. In order to assure confidentiality and encourage participation, limited demographics were obtained from each participant including their leadership title and specialty discipline, e.g., pediatrics, family medicine, internal medicine, emergency medicine, etc. No information ( gender, years in leadership position, location of practice, etc) was collected that could be used to link individual results with a particular individual.

11

Design elements. After IRB approval was obtained participants were recruited by the principal investigator. Potential participants were provided information describing the study, its relevance, and the likely benefits for the individual and the organization. When an individual agreed to participate they were asked to complete and return a “self” version of the Emotional Competency Inventory – 2 ( ECI-2 ) survey instrument used throughout the study ( see below ). Participants were also asked to recruit a sufficient number of superiors, peers, direct reports and others ( 6 to 8 total ) who completed a similar “other” version of the ECI-2; thereby providing a three hundred and sixty degree assessment of each participant’s emotional intelligence competency profile. Instrument. The ECI-2 ( Boyatzis, 2000 ) used in the study is an instrument that permits assessment of multiple competencies ( eighteen ) grouped in four clusters: Self Awareness ( emotional self awareness, accurate self assessment, self confidence ); Self Management ( emotional self control, trustworthiness, adaptability, optimism, achievement orientation, initiative ); Social Awareness ( empathy, service orientation, organizational awareness ); and Relationship Management ( teamwork and collaboration, developing others, influence, change catalyst, conflict management, leadership ). As utilized, the instrument had seventy-three questions distributed among the four clusters and eighteen competencies. Respondents used a five point Likert scale to describe themselves ( participants ) or another person ( others ) with regard to each competency. Respondents also had the option to reply “I don’t know” or “ I have not had the opportunity to observe” to each item. The ECI-2 has its origins in the Self Assessment Questionnaire ( SAQ ) developed by Boyatzis and others ( Boyatzis, 1994; Boyatzis, 1995 ). The latter instrument has its basis in competencies developed through the use of behavioral event interviews in multiple company studies provided as a consultation service to business, industry and

12

government ( Spencer & Spencer, 1993 ). The SAQ was extensively validated by Boyatzis ( 1997 ). The ECI 360 was pilot tested in 1998 and has been revised and tested with a large number of respondents. The current version, the ECI-2, has been tested on four thousand managers and an accompanying “other” cohort exceeding ten thousand respondents ( Boyatzis, 2001 ). The instrument was tested for reliability and validity according to the standards for Educational and Psychological Testing ( AERA, 1999 ). Cronbach alphas ( a measure of internal consistency and, therefore, reliability ) for the self assessment phase of the instrument range from 0.62 for adaptability to 0.87 for change catalyst, while composite other components had values of 0.80 for emotional self awareness to 0.95 for empathy. The ECI-2 is also supported by construct validity evidence, content validity evidence and generalization validity evidence based on its predecessor, the SAQ ( Spencer & Spencer, 1993; Boyatzis, 1994; Boyatzis, 1995; Boyatzis, 1997 ). Results from the ECI-2 360 were organized, collated and scored manually. Self respondent scores were utilized as reported. Other respondent scores for each participant were averaged within categories ( peers, superiors, direct reports, others ) and these category averages were used in calculating an “all others” average score for each item ( seventy three ) and each competence ( eighteen ). Independent measures of leadership effectiveness. Senior physician leaders from the departments or divisions where participants were drawn defined a list of independent measures used by the organization to assess physician leadership effectiveness ( as distinguished from clinical effectiveness when functioning as an individual provider ). Four objective: patient satisfaction for the individual participant’s area of responsibility as measured by a national patient satisfaction survey administered twice a year, unit gross financial contribution for 2001, clinical quality measures related to

13

ongoing care of diabetic patients, and physician productivity in area(s) of responsibility; and two subjective: innovation effectiveness and conflict / challenge management; measures were specified, weighted as to relative importance ( see Table 3 ), and a composite score determined for each participating physician leader ( Table 3 ). Outstanding physician leaders were distinguished from their typical colleagues by a score of 3.3 or greater on their composite scores. This threshold was selected by the non-participant senior physician leaders who chose the performance measures. It was based on data routinely collected by the organization to assess performance effectiveness. The 3.3 cutoff is not an established internal or external benchmark. It was chosen expressly for this study with the intent of distinguishing outstanding performers. Statistical analysis.. Distributions of the results were examined using histograms and box-plots ( Norusis, 2000 ). A paired sample t test method ( Norusis, 2000 ) was used to test whether differences between outstanding and typical performers ( as determined by independent, unrelated measures; see above ) were significant. Institutional Review Board (IRB) approval. The proposed project was reviewed and approved by the IRB at the participating institution and the IRB at Case Western Reserve University. Informed consent was obtained from all participants, both the physician leaders and those who agreed to complete an “other” survey on their behalf. Confidentiality was maintained by using coded survey result forms and the services of a third party data research coordinator who assured none of the survey results could be linked to specific participants.

14

RESULTS Observations related to descriptive statistical analyses. Thirty-four (34) physician leaders volunteered to participate in the study. Completed surveys ( self and / or others ) were received from twenty (20). Eleven (11) participating physician leaders had complete profiles and five (5) had several surveys submitted by others ( peers, supervisors, direct reports, etc. ) but their self surveys were not received. Four (4) other physicians either did not submit a self survey (2) or no other surveys were received (2) to accompany their self surveys. Data from the latter four (4) were excluded from subsequent analysis. A total of one hundred and two (102) self and other surveys ( physician leaders-11, supervisors-17, peers-38, direct reports-24, others-12 ) form the basis of the subsequent analysis and related remarks. All but two (2) of the physicians who completed self surveys and are included in the study were primary care practitioners with a majority of their practice in an ambulatory setting. Descriptive statistics for the one hundred and two (102) surveys are presented in Table 1 and Figures 1a and 1b. The figures were obtained by calculating a mean for the individual survey results from each subgroup ( self, peers, supervisors, direct reports, others ) associated with a participating physician then, averaging all of the individual results within a category. The ranges observed were restricted to the upper half of the five point Likert scale. Mean low values varied from 3.42 ( emotional self awareness – direct reports ) to 3.84 ( conflict management – all others ), while mean high values ranged between 4.36 ( empathy – self ) to 4.67 ( self confidence – direct reports ). Values below three (3.0) were recorded but were never sufficient to lower a mean below 3.0. Distributions were examined using box-plots prepared for each subgroup ( Figures 1a and 1b ).

15

In contrast to the reports of others who measured physicians’ humanistic ( respect, integrity, compassion ) ( Ramsey, 1993 ) or psychosocial ( Violato, 1997 ) characteristics and found physician self ratings lower than the values assigned by other raters, a similar pattern did not prevail in this study. Participating physician leaders who completed a self survey frequently rated themselves high, for example, on emotional self awareness ( 4.02 ), accurate self assessment ( 4.14 ), empathy ( 4.36 ) and conflict management ( 3.99 ). Their scores often exceeding the mean scores of the all others subgroup as well as the individual subgroups ( Tables 1 & 2 ). Two (2) representative physician profiles are illustrated in Figures 2a & 2b. Other physician profiles were similar. Each had unique aspects. No “typical” or “characteristic” pattern was apparent. When the individual emotional intelligence competency results were arrayed from lowest to highest using mean values within participating categories ( Table 2 ) interesting patterns became apparent. There was a disproportionate gathering of competencies from the self awareness ( emotional self awareness, accurate self assessment ) and self management ( adaptability, initiative, optimism ) clusters, as well as conflict management ( relationship management cluster ) at the lower end of the array. This was accompanied by a second cluster including the social awareness competencies, empathy and service orientation, teamwork and collaboration ( relationship management cluster ), self confidence ( self awareness cluster ) and trustworthiness ( self management cluster ) competencies at the opposite ( high ) end of the distribution. Most of the competencies comprising the relationship management cluster were distributed between the extremes. The potential relevance of these observations will be considered in the discussion. Several other observations among and between the self and other survey results deserve mention. Some of the most notable disparities were observed in comparing

16

physician self survey results with those of their supervisors, peers and direct reports ( Table 2 ). Physician self ratings for emotional self awareness, accurate self assessment and influence were all higher than the relative assignment of these competencies by direct reports, peers and supervisors. A note of caution is warranted in this regard. The differences cited are only apparent in the relative rankings within and between subgroups ( Table 2 ) and would not be appreciated if only absolute scores were considered. Observations related to independent measures of physician leadership effectiveness. Independent measures of physician leadership effectiveness were provided by the participating organization. A description of each measure and its relevance was included in the methods section. The parameters ( Table 3 ) address several important aspects of the physician leaders’ non-clinical responsibility. The relative weights were assigned by non-participant physician leaders and reflect key aspects of physician leadership accountability and effectiveness for the organization. The results, including a composite score, for ten (10) of the physician leaders participating in the study are summarized in Table 3. Outstanding performance ( a composite score of 3.3 or more ) was achieved by four (4) participants. Their emotional intelligence competency profile scores were compared with the scores of six (6) participants with typical composite scores on the independent effectiveness measures ( Tables 3 & 4 ). A paired t test method was utilized to compare the outstanding and typical mean emotional intelligence competency scores. The means for each emotional intelligence competency and the level of significance achieved when tested as described are shown in Table 4. No statistically significant differences ( p < 0.05 ) were found for any of the emotional intelligence competencies when outstanding and typical physician leader self, all other, or supervisor scores were tested. Outstanding and

17

typical physician self survey comparisons revealed marginally significant differences in achievement orientation ( p = 0.063) with the typical ( not the outstanding ) physician group demonstrating higher scores. Contrariwise, when differences between the “all other” means for each competency were determined, the outstanding group had better scores for trustworthiness ( p = 0.058 ) and achievement orientation ( p = 0.094 ). Comparison of the supervisory survey scores for the two groups failed to reveal significant differences. DISCUSSION This exploratory study revealed high levels of emotional intelligence competencies among the participating physician leaders. Strong performance was evident in all of the emotional intelligence competency clusters. Social awareness skills, notably empathy and service orientation, were prominently represented and complemented by high levels of trustworthiness, teamwork and self confidence. The midrange of results were populated by competencies associated with the relationship management and, to a lesser extent, the self management clusters. Among the former, developing others, was consistently displayed at high levels. Achievement orientation was also prevalent in self and other survey results. Relatively lower but sound scores were associated with a number of the self awareness ( emotional self awareness, accurate self assessment ) and self management ( adaptability, optimism, initiative ) competencies. It was this group where differences in the relative rankings of specific competencies by the physician leaders and their peers, supervisors and direct reports were apparent. The other subgroups consistently ranked emotional self awareness, accurate self assessment and conflict management lower than physician leaders scored themselves. In general, there was more congruence among the self, peer and supervisor results than with direct report ratings. The largest number of

18

relative disparities were observed when the results from physician leader self surveys were compared with direct report scores and rankings. Unfortunately, the research questions posed were not answered in a definitive manner. The ECI-2 survey results did not reveal meaningful differences between outstanding and typical physician leader emotional intelligence competency profiles when these groups were separated by the introduction of independent, objective measures of leadership effectiveness. Nor were there notable differences in the patterns of emotional intelligence competencies displayed by the typical and outstanding leaders who participated in the study. There are several factors that may have contributed to this outcome. The number of physician leaders who participated was modest; making discrimination difficult. The range of results displayed is relatively narrow. A feature that may necessitate a larger sample in order to detect modest differences. The physicians who chose to voluntarily participate in the study may represent a unique or select population relative to all potential participants. Finally, there is a possibility others ( peers, direct reports, supervisors ) selected by the participating leaders to contribute to the three hundred and sixty degree assessment might represent a group of respondents especially supportive of their physician leaders. These reservations must be placed in the context of broader generic threats to the validity of studies on emotional intelligence. Cherniss ( 2001 ), in a chapter entitled “Emotional Intelligence and Organizational Effectiveness”, included a section on “Unresolved Issues & Dilemmas” regarding the subject of emotional intelligence and related competencies. Three of the topics he considered are germane to this study. The first relates to the definition of the concept as well as the distinction between emotional intelligence and emotional intelligence competencies. Cherniss’ discussion is paralleled and complemented by Goleman ( 2001a ), writing in the same text.

19

He suggests that Bar-On, who introduced the term Emotional Quotient ( 1988 ), places emotional intelligence within a theory of personal well-being; while Salovey & Mayer’s concepts ( 1990 ) are best located within the sphere of intelligence theory, and his own ( Goleman’s ) model fits well with a theory of performance ( Goleman, 1998; Boyatzis, 1982; Spencer & Spencer 1993 ). Goleman ( 2001a ) also acknowledges the similarities between emotional intelligence and the concepts of intra- and inter-personal intelligence introduced by Gardner in his theory of multiple intelligences ( 1983 ). In essence Cherniss acknowledges a need for further clarity and greater consensus with regard to the definition of emotional intelligence. A second area where further development is warranted concerns the tests used to measure emotional intelligence and related competencies. There are five well described instruments in the literature ( Gowing, 2001 ). Only one, the Multifactor Emotional Intelligence Scale or MEIS ( Salovey & Sluyter, 1997 ) concerns itself with the measurement of emotional intelligence, while the others ( the EQI of Bar-On ( 1997 ), the the ECI 360, and the EQ map of Cooper & Arioli ( 1997 ) measure variations of behavioral evidences related to emotional intelligence. Gowing ( 2001 ) provides a thoughtful description of each test, its strengths and limitations while Davies (1998 ) offers a more critical assessment of the tests and their validity. A third challenge is the need to reconcile the predictive power of emotional intelligence relative to IQ. This is a complex, long-standing set of concerns where proponents of different persuasions have devoted significant efforts to debating the merits of each ( McClelland, 1973; Barrett & Depinet, 1991; Mayer & Salovey, 1993 ). This is interesting in the context of the current study where all the participants are intelligent and well educated. It is unlikely differences in the respective IQ scores could explain variations in physician leadership effectiveness. If this assumption is correct then

20

other attributes ( emotional intelligence competencies? ) may account for any differential and more extensive testing may provide a basis for validating this assumption. A search for related studies in the literature met with limited success. No other studies addressing the relationship between behavioral characteristics and independent measures of physician leadership effectiveness were identified. There is one report ( Duberman, 1999 ) comparing standard measures of primary care physician effectiveness ( in a managed care setting ) with behavioral competencies identified using behavioral event interview methods. It demonstrated modest correlation between clinical effectiveness and certain behavioral competencies, e.g., empathic care-giving. However, the subject population was practitioners not physicians in leadership roles and the study sample was small. There is complementary literature that affirms the importance of emotional intelligence in clinical practice. These studies emerge from the relationship-centered care movement ( Tresolini,1994 ), the selection and education of new physicians ( Carrothers, 2000 ), a growing interest in the retention of competence throughout a physician’s entire career ( Ramsey, 1993; Novack,1997; Violato,1997; Irvine, 1999 ), and a decline in medical professionalism ( Swick,2000 ). Unfortunately, this literature also suffers from a lack of consensus regarding terms, definitions, methods, and instruments of analysis, making it difficult to directly compare the results of different studies ( including the present one ). Despite these limitations there is a recognition and re-discovery that humanistic and socio-behavioral attributes are important aspects of physician effectiveness ( Ramsey, 1993; Violato, 1997 ). Furthermore, these qualities can be selected ( Carrothers, 2000 ), measured ( Carrothers, 2000; Ramsey, 1993; Violato, 1997 ), and strengthened through awareness and practice ( Novack, 1997; Epstein, 1999 ). Much of this work was recently summarized by Epstein & Hundert ( 2002 ). In their review they identify seven

21

dimensions of professional competence shared by physicians including cognitive, technical, integrative, contextual, relationship, affective / moral, and habits of mind. Sub- elements included under several of these headings are dimensions of the emotional intelligence framework utilized in this study. Another initiative affirming the importance of emotional intelligence competencies in clinical practice is the Outcomes Project sponsored by the Accreditation Council for Graduate Medical Education ( ACGME ). The latter group is responsible for general policy and oversight of postgraduate medical education in the U.S. Two years ago it initiated a major transformation of its residency review and accreditation process ( ACGME ), re-directing emphasis toward accountability for demonstrable competence development appropriate to each primary care, medical or surgical specialty. A set of six general competencies applicable to all programs ( patient care, medical knowledge, practice-based learning and improvement, interpersonal and communication skills, professionalism and systems-based practice ) assign significant importance to the development and measurement of emotional intelligence competencies ( ACGME ). A challenge facing the ACGME and residency program directors across the country is the creation of reliable, robust, comparable methods for developing, teaching, and measuring the cognitive and non-cognitive competencies required by this initiative. An important aspect of this preliminary study is the relationship between the competencies identified by the survey and alternative leadership styles. Recent publications by Goleman ( 2000 ) and Goleman, Boyatzis & McKee ( 2002 ) provide a framework linking key emotional intelligence competencies to six leadership styles: visionary, affiliative, democratic, coaching, pacesetting and commanding. These styles are complementary not mutually exclusive. The authors encourage flexibility and the use of different styles for different purposes. Proficient leaders will recognize the benefits

22

associated with different approaches and will master the key emotional intelligence competencies so they can utilize each style in an authentic and effective manner. Leading professionals, especially physicians, is a challenging task exacerbated by the changes confronting health care today. Core attributes of most physicians include expertise, autonomous behavior, collegiality, and service to others. As a consequence of the way American physicians have been selected, educated, and socialized during their training many are highly competitive, relatively independent practitioners. They often eschew teamwork and collaboration and other affiliative behaviors. Their education and socialization fosters pacesetting or commanding leadership styles that may be appropriate in certain clinical circumstances, e.g., a busy emergency department or a critical care unit, but could be counter-productive when used in other care settings. Democratic, affiliative or coaching styles are likely to be effective when working with physicians. Like most professionals, physicians are very democratic and resist hierarchical, command and control leaders ( unless they are leading! ). The physician leaders surveyed in this study may not be typical. They demonstrated high levels of empathy, as well as teamwork and collaboration competencies. These fit well with democratic, affiliative and coaching leadership styles. Study physicians also scored well in developing others, an important attribute for effective coaching. Trustworthiness or transparency and service orientation are two other competencies among the top five manifest by this group. Both are important when building trust and reliability with followers and encouraging a customer focus when serving patients. Relative weaknesses revealed by the survey include conflict management and influence skills; both valuable when adopting a democratic leadership approach and a potential developmental opportunity. As suggested above this “profile” may not be characteristic of physician leaders in general. Factors that may be influencing it include the employed group practice status of the participants, their primary care orientation, and the medical group’s investment and

23

commitment to managed care principles and practices. While a combination of democratic, coaching and affiliative leadership styles may serve physicians well in relating to colleagues and other members of the health care team, incorporating visionary leadership competencies is essential as a physician executive assumes broader leadership responsibilities. If a pacesetting and / or commanding style has been an important part of a physician leader’s portfolio it would be prudent to reduce reliance on these options as confidence in other styles is practiced and refined. Beginning with the early design phase of this project there was a concern that endogeneity , i.e., the potential for the dependent variable ( physician leadership performance relative to independent criteria ) to unduly influence the independent variables ( emotional intelligence competencies ), could be operative. It was determined the most effective way to reduce or eliminate this possibility was to select performance indicators that were unlikely to be influenced in this way. The performance measures chosen to distinguish outstanding and typical performers are not likely to be influenced in this manner and it is unlikely defensible claims could be advanced that outstanding performers received special considerations and were able to reciprocate by demonstrating higher levels of the behaviors relevant to the measured emotional intelligence competencies. However, another “confounding variable” has emerged that is not so easily dismissed. It is the inherent difficulty of separating the relative contribution of the physician in his or her leadership role from their role as clinician and caregiver. Both have the potential to elicit similar behavioral responses. Explicit instructions advising participants to distinguish these roles were not provided. Even if such instructions had been given it would be difficult to distinguish and weight the relative contributions in in a leader / caregiver’s daily behaviors. Given the exploratory nature of this initiative

24

future studies should address this concern through the use of alternative instruments and / or the development of specific questions designed to facilitate this distinction. Other features that should be incorporated into subsequent studies include a larger sample of physicians in leadership positions; a balanced mix of primary care, medical and surgical subspecialty leaders; gender and cultural diversity; the impact of longevity in a leadership position; and the influence of formal management education ( MHA, MBA, other ) on physician leadership effectiveness.

25

REFERENCES ACGME Outcome Project. 2000. Accreditation Council for Graduate Medical Education Web site. Available at http:// www.acgme.org. Bar-On, R. 1988. The development of an operational concept of psychological well- being. Unpublished doctoral dissertation, Rhodes University, South Africa. Bar-On, R. 1997. The Emotional Quotient Inventory ( EQ-I ): Technical Manual. Toronto: MultiHealth Systems. Barrett, G.V., & Depinet, R.L. 1991. A reconsideration of testing for competence rather than intelligence. American Psychologist, 46:1012-1024. Barzansky, B., Jonas, H.S., & Etzel, S.I. 2000. Educational programs in U.S. medical schools. J of the American Medical Association, 284: 1114-1120. Becker, H., Geer, B., Hughes, E.C., & Strauss, A. 1961.Boys in White: Student Culture in Medical School. Chicago: U of Chicago Press. Bellack, J.P. 1999. Emotional intelligence: a missing ingredient? J of Nursing Education, 38:3-4. Boyatzis, R.E. 1982. The Competent Manager: A Model for Effective Performance. New York: Wiley. Boyatzis, R.E. 1994. Stimulating self-directed change: a required MBA course called Management Assessment and Development. J of Management Education 18: 304-323. Boyatzis, R.E., Cowen, S.S., & Kolb, D.A. 1995. Innovation in Professional Education: Steps on a Journey to Learning. San Francisco: Jossey-Bass. Boyatzis, R.E., Wurzler, J., & Wright, R. 1997. Competency development in graduate education: a longitudinal perspective. Paper presented at the first world conference on self-directed learning, Group for Interdisciplinary Research in Autonomy & Training. Montreal: Quebec. Boyatzis, R.E., Goleman, D., & Rhee, K.S. 2000. Clustering Competence in Emotional Intelligence: Insights from the Emotional Competence Inventory. In R. Bar-On & J.D.A. Parker ( Eds. ) The Handbook of Emotional Intelligence: 343-362. San Francisco: Jossey-Bass. Boyatzis, R.E. 2001. Personal communication, Oct 25. Carrothers, R.M., Gregory, S.W., & Gallagher, T.J. 2000. Measuring emotional intelligence of medical school applicants. Academic Medicine, 75: 456-463. Cherniss, C. 2001. Emotional Intelligence & Organizational Effectiveness. In C. Cherniss & D. Goleman ( Eds. ) The Emotionally Intelligent Workplace: 27-44. San Francisco: Jossey-Bass.

26

Cooper, R.K. 1997. EQ map overview. San Francisco: Q-Metrics. Davies, M., Stankov, L., & Roberts, R.D. 1998. Emotional Intelligence: In search of an elusive construct. J of Personality & Social Psychology, 75: 989-1015. Duberman, T.L. 1999. Assessing the critical behavioral competencies of outstanding managed care primary care physicians. Managed Care Interface, 76-80. Epstein, R.M. 1999. Mindful practice. J of the American Medical Association, 282: 833-839. Epstein, R.M. & Hundert, E.M. 2002. Defining and assessing professional competence. J of the American Medical Association, 287: 226-235. Fernandez-Araoz, C. 2001. The Challenge of Hiring Senior Executives. In C. Cherniss & D. Goleman ( Eds. ), The Emotionally Intelligent Workplace: 182-206. San Francisco: Jossey-Bass. Friedson, E. 1975. Doctoring Together: A Study of Professional Social Control. New York: Elsevier. Gardner, H. 1983. Frames of Mind: The Theory of Multiple Intelligences. New York: Basic Books. Goleman, D. 1995. Emotional Intelligence. New York: Bantam. Goleman, D. 1998. Working With Emotional Intelligence. New York. Bantam. Goleman, D. 2000. Leadership that gets results. Harvard Business Review. Mar-Apr: 78-92. Goleman, D. 2001a.Emotional Intelligence: Issues in Paradigm Building. In C. Cherniss & D. Goleman ( Eds. ), The Emotionally Intelligent Workplace: 13-26.San Francisco: Jossey-Bass. Goleman, D.2001b. An EI-based Theory of Performance. In C.Cherniss & D. Goleman ( Eds. ) The Emotionally Intelligent Workplace: 27-44. San Francisco: Jossey-Bass. Goleman, D., Boyatzis, R.E., & McKee, A. 2002. Primal Leadership: Realizing the Power of Emotional Intelligence. Boston: Harvard Business School Press. Gowing, M.K. 2001. Measurement of Individual Emotional Competence. In C. Cherniss & D. Goleman ( Eds. ), The Emotionally Intelligent Workplace: 83-131. San Francisco: Jossey-Bass. Guthrie, M.B. 1999. Challenges in developing physician leadership. Frontiers in Health Services Management, 15: 3-28. Herzlinger, R.E. 1997. Market-driven Health Care: Who Wins Who Loses in the Transformation of America’s Largest Service Industry. Reading, MA: Addison- Wesley.

27

Irvine, D. 1999. The performance of doctors: the new professionalism. Lancet, 353: 1174- 1177. LeTourneau, B. & Curry, W. 1997. Physicians as executives: boon or boondoggle? Frontiers in Health Services Management, 13: 3-25. Mayer, J.D. & Salovey, P. 1993. The intelligence of emotional intelligence. Intelligence, 17: 433-442. McClelland, D.C. 1973. Testing for competence rather than intelligence. American Psychologist, 28: 1-14. McClelland,D.C. 1998. Identifying competencies with behavioral event interviews. Psychological Science, 9: 331-340. Moore, W.E. 1970. The Professions: Roles & Rules. New York: Sage. Norusis, M.J. 2000. SPSS 10.0: Guide to Data Analysis. Upper Saddle River, NJ: Prentice-Hall. Novack, D.H., Suchman, A.L., Clark, W., Epstein, R.M., Najberg, J., Kaplan, C. 1997. Calibrating the physician: personal awareness and effective patient care. J of the American Medical Association, 278: 502-509. Ramsey, P.G., Wenrich, M.D., Carline, J.D., Inui, T.S., Larson, E.B., LoGerfo, J.P. 1993. Use of peer ratings to evaluate physician performance. J of the American Medical Association, 269: 1655-1660. Saks, M. 1995. Professions & the Public Interest: Medical Power, Altruism, & Alternative Medicine. London: Routledge. Salovey, P. & Mayer, J.D. 1990. Emotional Intelligence. Imagination, Cognition & Personality, 9: 185-211. Salovey, P. & Sluyter, D.J. ( Eds. ). 1997. Emotional Development & Emotional Intelligence: Implications for Educators. New York: Basic. Schneider, C.E. 1998. The Practice of Autonomy: Patients, Doctors & Medical Decisions. New York: Oxford. Spencer, L.M. & Spencer, S.M. 1993. Competence at Work: Models for Superior Performance. New York: Wiley. Standards for Educational & Psychological Testing. 1999. Washington, D.C.: American Educational Research Association. Swick, H.M. 2000. Toward a normative definition of medical professionalism. Academic Medicine, 75: 612-616. Tresolini, C. & the Pew-Fetzer Task Force. 1994.Health Professions Education and

28

Relationship-Centered Care: Report of the Pew-Fetzer Task Force on Advancing Psychosocial Health Education. San Francisco: Pew Health professions Commission. Violato, C., Marini, A., Toews, J., Lockyer, J., Fidler, H. 1997. Assessing physicians in practice. Academic Medicine, 72: S82-S84.