Pololi, Linda H. MD; Krupat, Edward PhD; Civian, Janet T. EdD; Ash, Arlene S. PhD; Brennan, Robert T. EdD
Faculty physicians and scientists in U.S. medical schools hold leading roles in American health care by educating physicians, driving and applying research in the biomedical and behavioral sciences, and providing sophisticated clinical care for patients, often to underserved groups. As a critical resource and a national investment, this influential sector of the health care workforce must function effectively. With health care costs and reform being nationally debated, we concluded that an appraisal of the existing culture of academic medicine and its effects on the approximately 130,000 faculty in U.S. medical schools was necessary. Below, we report our use of a national faculty survey to examine (1) the extent to which faculty are considering leaving their institution or academic medicine entirely, and their reasons for doing so, and (2) whether faculty perceptions of organizational culture and institutional demographics are associated with faculty intentions to stay at or leave their schools or academic medicine.
Several studies of faculty in academic medicine report high levels of dissatisfaction,1 attrition,2,3 burnout rates,4 unethical behavior in research,5 and an aging faculty.6 Even so, there is little published research on what predisposes faculty to leave academic medicine or the reasons for their dissatisfaction. In two single-institution studies,7,8 Cropsey7 found that attrition of women and minority faculty was due to professional advancement issues, low salary, and chairman/leadership discontent (51% response rate), and Lowenstein8 found that lack of recognition for teaching and clinical excellence, and difficulty in balancing family–career responsibilities, were linked to the intent to leave (38% response rate). Schindler et al,1 in an important four-school study of faculty well-being, documented high levels of depression and anxiety, especially in younger faculty, and, in comparison with findings of a 1985 study,9 found decreased work satisfaction and higher levels of clinically significant depressive symptoms in faculty.1 A recent Association of American Medical Colleges (AAMC) analysis2 of clinical-only physician faculty in 23 self-selected medical schools reported retention rates and briefly assessed satisfaction of this group of faculty (47.5% response rate). Correlates of intent to leave focused on the activity and quality of patient care and whether the workplace cultivated collegiality. Additionally, the cost of replacing one clinical faculty member has been calculated as between $155,000 and $559,000, depending on discipline,10 plus the many less tangible effects on the academic health center (AHC). Faculty turnover costs account for 5% of the annual budget in one AHC.11 Outside the context of medicine, organizational studies suggest that employee perceptions of organizational justice and an ethical climate are related to increased job satisfaction and reduced turnover.12,13
In previous research14–21 conducted by the National Initiative on Gender, Culture and Leadership in Medicine (C-Change),22 we and colleagues provided qualitative data from academic faculty interviews on their professional experiences in AHCs. We learned that faculty continued to find their work meaningful and intellectually stimulating, even though their workplace culture was competitive rather than collaborative and faculty often felt isolated and disconnected from their colleagues.14 Many perceived a lack of alignment between their own values and those exhibited by their schools.16,18 Both women and faculty of color described unconscious and conscious bias.19–21 Building on this prior hypothesis-generating work, for the present study, we developed the C-Change Faculty Survey to measure faculty perceptions of their organizational culture and faculty intentions to stay or leave either their schools or academic medicine in a nationally representative sample of 26 medical schools.
The domains and items of our survey questions were derived in large part from themes identified in the C-Change qualitative studies14–18,20,21 in conjunction with an extensive search of the literature and reviews of relevant instruments.1,23–27 We created 74 items related to advancement, engagement, relationships, feelings about workplace, diversity and equity, leadership, institutional values and practices, and work–life integration. Items used five-point Likert scales (range: 1 = strongly disagree to 5 = strongly agree). We piloted and refined the survey, working with medical colleagues at nonparticipating schools.
Human subject institutional approvals were obtained from Brandeis University, Boston University, and the AAMC.
Selecting schools. Schools were selected in a multistage process. First, we included the 5 schools that were part of the consortium associated with the project from which this study derived. The 5 C-Change schools had been selected to vary among themselves in important ways and encompass a range of key attributes of medical schools (public/private, region of the United States, National Institutes of Health [NIH] research intensive, or community care focused). The mean aggregate faculty demographics of these 5 schools matched those of all the 126 AAMC member schools at the time. Our survey was administered in these 5 schools as a pretest survey and very early in the school leaders’ participation in the C-Change Learning Action Network. Then, with the assistance of the AAMC and using the AAMC roster of all its 126 member medical schools at that time, we constructed a stratified random sample of an additional 21 medical schools to ensure that the resulting 26 schools (1) spanned all school types (including 1 small and 1 historically black school) and (2) achieved a distribution similar to the overall proportion of AAMC member schools across eight strata defined by four geographic regions crossed with public/private status. Each dean was asked permission to approach faculty and to provide e-mail addresses. If consent was not granted, a school from the same stratum was randomly selected as a replacement. The same stratified random selection of faculty was used in all 26 schools. Because the first 5 schools had not been selected at random, the data from these schools were systematically compared with those of the other 21. We found no significant differences in respondents’ perceptions on 11 of 12 scales created to measure dimensions of the culture (see Table 1). The exception was the relatedness/inclusion scale, on which the 5-school faculty sample scored modestly higher; all demographic characteristics of the faculty were alike. We cannot name the participating schools because they were assured anonymity; however, the systematic random sampling of 21 of the 26 schools and the measured similarity of the 5 nonrandomly and 21 randomly selected schools gave us confidence that the sample 26 schools is reasonably representative of all U.S. medical schools. Table 2 compares characteristics of the study schools with all AAMC member schools.
Sampling faculty within schools. The AAMC provided lists of full-time faculty at each school, including faculty demographic characteristics. For sampling, each faculty member was categorized by sex and chronological age: under 39 years old; 39 to 47 years; and 48 years and older. Faculty selection began with equal allocation stratified sampling to ensure adequate numbers of women in the respondent pool for analysis with a feasible total sample size. At each school, 25 faculty were randomly selected from each of 6 sex-by-age categories for a base sample of 150 per school. To ensure adequate numbers of faculty in 2 other groups of interest, we added faculty at each school to the 150 members sampled by sex and age. Faculty who identified as underrepresented minorities in medicine (URMM) were one such group. We followed NIH definitions in coding the following as URMM: American Indian or Alaska Native, black or African American, Hispanic/Latino, Native Hawaiian, or other Pacific Islander. At each school, we added URMM faculty to each age group, up to a maximum of 20 URMM faculty per group. When fewer than 20 URMM faculty were available in a category, faculty from the next-younger category were selected until the target was reached or the pool of URMM faculty was exhausted.
Female surgeons were another group of interest because of the particularly low representation of women in this specialty. To ensure their representation, additional female surgeons were selected until the pool was exhausted. By oversampling URMM faculty and female surgeons, we were able to maximize statistical power for testing comparative questions for these groups and to increase the precision for estimates in what might otherwise be sparse categories when examining characteristics within these groups such as faculty rank. Weights were employed to adjust for oversampling in survey analyses.
The stratified selection process and inclusion of additional URMM faculty and female surgeons resulted in a list of 4,578 sampled faculty. The survey was administered electronically, with reminders at two- to three-week intervals and eventual follow-up with phone contact and a hard copy mailing when needed. The survey was distributed, as schools were recruited, in waves from 2007 through early 2009. Faculty were assured that their responses would be anonymous.
Using the five-point scale described earlier (i.e., strongly disagree, somewhat disagree, neither agree nor disagree, somewhat agree, strongly agree), respondents were asked the extent to which they endorsed the following statement: “In the past 12 months, I have seriously considered leaving my current institution.” A second, analogously phrased item asked about “seriously considering leaving academic medicine.” Those responding with one of the “agree” categories were asked their reason for considering leaving. Responses to these items generated five discrete groups: (1) stayers—those who had not seriously considered leaving, (2) those considering leaving their school but not academic medicine because of dissatisfaction, (3) those considering leaving academic medicine because of dissatisfaction, and those considering leaving their school and/or academic medicine either (4) to retire or (5) for personal/family reasons.
Several personal and professional demographic items were used as predictors: sex; URMM status; age; having a medical versus another terminal degree; percent time devoted to research; rank; primary role of clinician, researcher, administrator, or educator; and holding a leadership position. These characteristics were gathered from survey response data. Additional professional attributes included receipt of mentoring (or not) and whether their medical school (as opposed to their hospital) predominantly affected their job satisfaction. School-level variables included NIH award ranking and (from the AAMC) sector (public versus private), region, school size, and percentage of faculty who are women.
We constructed weights based on sex, age, and URMM status of all faculty at the 126 AAMC member schools in 2008 to be able to generalize our findings to the national population of academic faculty. To address missing values in demographic data and scales representing dimensions of the culture, 10 multiply imputed data sets were estimated using IVEware 2002 (Survey Research Center, Institute for Social Research, University of Michigan).28 Under certain assumptions, multiple imputation yields unbiased point estimates and confidence intervals.29 IVEware uses chained equations in combination with a Markov chain Monte Carlo method.
To determine the conceptual structure underlying faculty responses, we subjected 46 items related to institutional culture to a factor analysis using SAS/STAT Version 8.2 for Windows, 2004 (SAS Institute, Cary, North Carolina). First, we examined unrotated principal component loadings showing the linear consistency among all items, retaining items with unrotated factor loadings ≥ 0.40.30 Then, we used an equamax rotation to identify distinct factors, or subdimensions, of institutional culture. We used these in conjunction with semantic review of the items to guide final scale development; an additional five scales were content derived. Negatively stated individual questions were reverse coded, responses summed, and scores divided by the number of items in each scale. Cronbach α reliability coefficients were estimated to assess the internal consistency of each scale (see Table 1 for a complete list of scales and abbreviated items).
We used a two-level (individual and school) multinomial logit model31 to estimate predictors of membership in each leaver group using HLM 7 for Windows, 2011 (Scientific Software International, Inc., Lincolnwood, Illinois). Because these models are complex, in particular with regard to parallel equations for each of the four types of leaving intentions, and because some of the leaver groups were sparsely populated, the models were constructed by testing progressive blocks of variables as indicated in Table 3, starting with personal demographics, and by conducting joint significance tests to retain or drop each group of variables using the multiple general linear hypothesis testing available in HLM 7. Only significant variable blocks were retained. However, female and URMM were retained to test two-way interaction effects in a later model. Finally, a reduced model was created consisting only of variables that were significant predictors for at least one of the four outcome categories. The reduction alleviates multicollinearity concerns and improves the efficiency of the estimation.
In the model presented, odds ratios for the scale variables were interpreted for a one-unit change on the original Likert scale (e.g., from “3 = neither agree nor disagree” to “4 = somewhat agree”), and the odds ratio for age was interpreted for a one-decade change in age. All continuous variables were centered on their grand means. All binary variables were coded 0/1, with the variable name representing the 1 value; for instance, the odds ratio associated with female represents the odds associated with being female versus being male.
Of the 4,578 faculty invited to participate, 2,381 responded for a response rate of 52%. The median school response rate was 54%; one school had a rate of 28%, whereas the others ranged from 40% to 62%. These calculated response rates are conservative because we assumed that all nonrespondents were survey eligible. Because some respondents either skipped or answered “Don’t know” to one or both questions related to intentions to leave, outcome data were missing on 387 respondents, yielding an analysis sample of 1,994. Of these, 1,062 (53%) were female, and 475 (24%) were URMM faculty.
Among respondents with complete data, 1,142 (57%) were stayers. The remaining 852 (43%)—all “leavers”—were divided among those considering leaving their school (but not academic medicine) because of dissatisfaction (273 [14%]), those considering leaving academic medicine altogether because of dissatisfaction (421 [21%]), those considering leaving their school and/or academic medicine for personal/family reasons (109 [5%]), and potential retirees (49 [2%]). Characteristics of the sample and descriptive statistics for faculty-level predictors are presented in Table 3.
Schools were almost evenly split between public (14 [54%]) and private (12 [46%]) and were similarly distributed by region in the universe of AAMC member schools: Northeast, 8 (31%); South, 8 (31%); Central, 6 (23%); and West, 4 (15%). School faculty size ranged from about 200 to 1,600, with an average of 935; NIH award ranks spanned nearly the entire range, with an average rank of 60.4. Faculty at the 26 schools were 33% women (versus 32% across all schools nationally).
Scales associated with faculty staying versus considering leaving
The rotated solutions identified 8 factors, which were used to create 7 scales as identified in Table 1. The 5 additional content-derived scales and associated statistics for all 12 scales are also displayed in Table 1. Results for the multilevel multinomial logit model for all four categories of possible leavers are displayed in Table 4.
For none of the leaver groups was there a significant association between intention to leave and URMM status, faculty rank, percentage of time devoted to research, holding a school leadership position, whether their medical school (versus their hospital) predominantly affected their job satisfaction, school size, sector (public versus private), geographic region, NIH award ranking, or the percentage of faculty who are women. Two scales failed to predict leaving: perceptions of gender equity and institutional change efforts for diversity.
Faculty considering leaving because of dissatisfaction
Figure 1 summarizes the findings for intention to leave because of dissatisfaction. Among the scales associated with desire to leave one’s institution and academic medicine were lower senses of relatedness/inclusion, engagement, self-efficacy, values alignment, and institutional commitment to improve support for faculty. Low perceived institutional support was related to consideration of leaving one’s institution but not leaving academic medicine, and high ethical/moral distress was associated only with leaving academic medicine altogether. Younger faculty were more likely to consider leaving academic medicine for dissatisfaction than were older faculty (see Table 4). Faculty with higher leadership aspirations were more likely to consider leaving their institution and leaving academic medicine (see Table 4).
Faculty considering leaving because of retirement or personal/family reasons
Not surprisingly, retirement was linked to age, although higher engagement and work–life integration scores were associated with less consideration of retirement. For the 5% considering leaving either their school or academic medicine (combined) for personal/family reasons, younger faculty were more likely to be at risk of leaving, as were those with medical degrees compared with those with other degrees. Faculty who primarily were researchers were more likely to consider leaving than those who primarily were clinicians. Faculty with higher engagement and self-efficacy scores were less likely to consider leaving for personal/family reasons. For women only, higher scores on leadership aspirations were associated with lower likelihood of leaving for personal/family reasons.
Discussion and Conclusions
Our survey’s findings about the extent to which faculty intend to leave are consistent with findings from the AAMC analysis of actual faculty retention and turnover2 and with Schindler and colleagues’1 findings in four schools. Our study delves into why faculty feel so dissatisfied that they intend to leave either their own institutions or their careers in academic medicine. The central and concerning finding is that faculty dissatisfaction was saliently associated with faculty members’ negative perceptions and distress about the nonrelational and ethical culture of the workplace. Significant predictors of intention to leave included feeling vulnerable and unconnected to colleagues, moral distress, perceptions of the culture being at times unethical, and feelings of being adversely changed by the culture. Low self-efficacy and sense of engagement and a lack of alignment of faculty members’ personal values with perceived institutional values also predicted intention to leave. In marked contrast, objective characteristics such as school size, region, or sector (public versus private) were not associated with dissatisfaction and considering leaving. None of the following either independently predicted, or moderated, dissatisfaction and considering leaving: the faculty member’s sex, URMM status, or perceptions associated with work–life integration (see List 1). Similarly, Lowenstein et al8 found no sex difference in intention to leave academic medicine.
Relationships and inclusion
Unrelatedness—expressed as faculty feeling isolated and invisible—correlates with intent to leave. This finding accords with the tenet of psychological theory that social environments characterized by a lack of connectedness and low self-efficacy hinder intrinsic motivation and thwart individual innate psychological needs.32 Although relationships are increasingly recognized as critical in shaping medical training33,34 and organizational health care outcomes,35 few studies exist on relationships among colleagues.35,36 In emergency medicine faculty, Lowenstein et al8 also showed that intention to leave was related to “departmental lack of a sense of academic community.” Despite ongoing efforts to enhance professionalism and humanistic approaches in academic medicine, there seems to be little recognition of the existing significant relational barriers to these goals. Our survey findings highlight the importance of relationship formation in medical schools and teaching hospitals.
Schindler and colleagues1 found clinically significant symptoms of depression among approximately 20% of both men and women faculty in four medical schools, as well as high anxiety ratings, with younger faculty being most affected. Such problems could be contributing to our respondents’ feelings of vulnerability, being unconnected to colleagues, and lower self-efficacy and sense of engagement. The causal relationship of depression and intention to leave cannot be determined in this study, nor can whether lack of values alignment, working in a nonrelational culture, and ethical/moral distress contribute to depression. On the other hand, we note that several of the survey questions elicited very positive responses; for example, almost all faculty found their work personally satisfying, and about three-quarters felt energized by their work. How the mental status of faculty affects their sense of satisfaction and intention to leave—and the well-being of faculty generally—warrants further study.
Faculty ethical/moral distress
Our findings are congruent with meta-analyses of 25 years of organizational justice research outside medicine. These studies suggest that employee perceptions of organizational justice and an ethical climate are related to increased job satisfaction, trust in leadership, enhanced performance, commitment to one’s employer, and reduced turnover.12,13
The scale of ethical/moral distress (see Table 1) reflects reactions to the prevailing norms and possible erosion of professionalism and increased organizational self-interest. There is a growing belief that organizations influence and are responsible for the ethical or unethical behaviors of their employees.37,38 To our knowledge, faculty perceptions of “moral atmosphere” and “just community” embedded in our survey have not been previously investigated in academic medicine, even though the ethical concepts of professionalism and justice can be used to guide the pursuit of excellence in the missions of medical schools. Several scholars have called for academic medicine to attend to its social justice and moral mission.39 Faculty perceptions of organizational justice are pivotal to the critical issue of professionalism in medicine. The ethical/moral distress scale in the survey reported here included items such as “the culture of my institution discourages altruism” and “I find working here to be dehumanizing.” (See Table 1 for other items in this scale.) In that ethical/moral distress was more strongly related to intent to leave academic medicine entirely than intent to leave one’s own institution, these negative feelings among faculty must be particularly disheartening to them and may color major career decisions. To understand why faculty with high ethical/moral distress are more likely to leave academic medicine requires more targeted data and analysis.
In addition to effects on intention to leave, the detrimental culture for faculty members constitutes part of the hidden curriculum for medical students, who often become less altruistic and more cynical through the four years of medical school.40,41 We concur with Inui42 that if faculty project that the moral, ethical, professional, and humane values articulated in the formal curriculum are not reinforced in their own experience as faculty (through the medium of the hidden or informal curriculum), the goals of educating and graduating competent, professional, and humanistic physicians may be undermined.
Age of faculty members
Similar to AAMC analyses,2 we found that younger faculty were more likely to consider leaving academic medicine for dissatisfaction than their older colleagues. Differential loss of young faculty is especially problematic because faculty development is a long-term investment for individual schools and for the nation’s health care. One of the few benefits of turnover is the opportunity to recruit fresh ideas and energy. Furthermore, the possibility of a dwindling pool of trainees seeking academic medical careers exacerbates concerns over an “aging” medical faculty.6 We found that faculty aspiring to leadership were more likely to consider leaving not only their institution but also academic medicine. Therefore, succession planning and developing faculty for local leadership may benefit schools wishing to reduce faculty turnover.
Of note, our findings for both women and men did not associate work–personal life problems with dissatisfaction and leaving academic medicine.
Study limitations and strengths
This study used faculty data from a not fully random sample of AAMC member schools. However, most of the schools (n = 21) were randomly selected, and all 26 together were highly representative of all AAMC member schools at the time (n = 126), all of which were MD-granting medical schools in the United States accredited by the Liaison Committee on Medical Education, as indicated in Table 2. Within schools, faculty were randomly selected, although the AAMC roster used for sampling was not fully up-to-date regarding recent faculty attrition, nor were the e-mail lists of their faculty provided by the schools we studied. The response rate could have been higher had we been able to exhaustively determine all sampled faculty who were ineligible for inclusion. Especially given this, a 52% response rate compares well with those of other reported surveys of medical faculty1,5,7–9,43 and seems likely to underreport dissatisfaction and leaving.
We also report intent to leave rather than actual faculty attrition; in studies outside medicine, intent to leave one’s faculty position was the strongest predictor of actual voluntary turnover.44 The reasons for actual attrition are much harder to observe because once faculty members leave, they are hard to locate, the number who respond to surveys is very low, and their answers are subject to recall bias.
This study provides substantial detail to our previous understanding of faculty dissatisfaction and attrition.1,3,7,9,45,46 Because survey items were based on a hypothesis-generating qualitative analysis of prior faculty interviews, our survey included numerous noncustomary questions and domains relating to relationships, values, ethical and moral climate, being changed by the culture, diversity, equity, and support/advancement. Our faculty sample included members from both clinical practice and basic sciences and also identified major roles of the faculty respondents and detailed demographic data.
Our work adds to the literature in that our findings provide new insights into understanding the dimensions that contribute to the culture in academic medicine and its effects on faculty intent to leave. Because these same factors predict performance in nonmedical settings,12,13 this suggests that the factors may reduce the productivity of medical faculty.
We hope that medical schools can use these findings to develop organizational structures that not only support intellectual endeavors but also support relationship formation among faculty, including those in leadership.47,48 The Relationship Centered Care Initiative has focused on this goal; for instance, Cottingham and colleagues42 gathered stories of positive relational patterns among faculty and used these to foster mindfulness and to enhance relational practices. Incorporating activities for faculty and leaders that encourage reflection on the meaningfulness of their work and core convictions could help address moral and ethical issues and help prevent faculty from feeling that they have been adversely changed in the culture.48 This core change in medical school culture could encourage trustworthy relationships and support the humanistic needs of health professionals. A supportive culture could also positively affect students and physicians in training and facilitate the inclusiveness and collaboration essential for creative research productivity and optimal patient care. Although the beliefs and attitudes of faculty are fortunately aligned with the purported missions of academic medicine, our survey findings suggest that the culture of medical schools is a barrier to fulfilling these professed goals. We have some optimism that increased awareness of the issues discussed in this report will support ongoing efforts to positively change the culture for academic health professionals. It bodes well that the faculty we studied retained an awareness of a moral imperative and their own deeply held values, such that the negative aspects of their work environment did not obliterate their desire to act authentically and professionally in meeting the demands of their roles and responsibilities.
Acknowledgments: The authors wish to thank all the members of the C-Change research team who participated in developing items for the survey and supporting the process of C-Change work: Phyllis Carr, Lisa Cooper, Peter Conrad, Michael Goldstein, David Kern, Sharon Knight, and Eugene Schnell. The authors thank the AAMC for assisting in the initial phase of the project, and Lisa Rosenbaum and Jeffrey Prottas for generously offering their helpful suggestions on an earlier draft of this manuscript. The authors are indebted to the medical faculty who generously shared their perspectives in the survey.
Funding/Support: The authors gratefully acknowledge the critical funding support of the Josiah Macy Jr. Foundation and Brandeis University Women’s Studies Research Center. Funding supported the design and conduct of the study and the collection, management, analysis, and interpretation of the data. Supplemental funds to support data analysis were provided by the U.S. Health and Human Services Office of Public Health and Science, Office on Women’s Health, and Office of Minority Health; National Institutes of Health, Office of Research on Women’s Health; the Agency for Healthcare Research and Quality; the Centers for Disease Control and Prevention; and the Health Resources and Services Administration.
Other disclosures: The C-Change Faculty Survey and its items, described in this report, are copyrighted by C-Change, Brandeis University. Please contact Dr. Pololi to use this survey.
Ethical approval: Human subject institutional approvals were obtained from Brandeis and Boston Universities and the AAMC.
1. Schindler BA, Novack DH, Cohen DG, et al. The impact of the changing health care environment on the health and well-being of faculty at four medical schools. Acad Med. 2006;81:27–33
2. Corrice AM, Fox S, Bunton SA. Retention of full-time clinical M.D. faculty at US medical schools. AAMC Analysis in Brief. February 2011;11:1–2
3. Alexander H, Lang J. The long-term retention and attrition of U.S. medical school faculty. AAMC Analysis in Brief. June 2008;8:1–2
4. Shanafelt TD, Sloan JA, Haberman TM. The well-being of physicians. Am J Med. 2003;114:513–519
5. Martinsen BC, Anderson MS, Vries R. Scientists behaving badly. Nature. 2005;435:737–738
6. Alexander H, Liu CQ. The aging of full-time U.S. medical school faculty: 1967–2007 AAMC Analysis in Brief. October 2009;9:1–2
7. Cropsey KL, Masho SW, Shiang R, Sikka V, Kornstein SG, Hampton CL. Why do faculty leave? Reasons for attrition of women and minority faculty from a medical school: Four-year results. J Womens Health (Larchmt). 2008;17:1111–1118
8. Lowenstein SR, Fernandez G, Crane LA. Medical school faculty discontent: Prevalence and predictors of intent to leave academic careers. BMC Med Educ. 2007;7:1–8
9. Linn LS, Yager J, Cope D, Leake B. Health status, job satisfaction, job stress, and life satisfaction among academic and clinical faculty. JAMA. 1985;254:2775–2782
10. Schloss EP, Flanagan DM, Culler CL. Some hidden costs of faculty turnover in clinical departments in one academic medical center. Acad Med. 2009;84:32–36
11. Waldman JC, Kelley F, Smith AS. The shocking costs of turnover in healthcare. Health Care Manage Rev. 2004;29:2–7
12. Colquitt JA, Conlon DE, Wesson MJ, Porter CO, Ng KY. Justice at the millennium: A meta-analytic review of 25 years of organizational justice research. J Appl Psychol. 2001;86:425–445
13. Martin KD, Cullen JB. Continuities and extensions of ethical climate theory: A meta-analytic review. J Bus Ethics. 2006;69:175–194
14. Pololi L, Conrad P, Knight S, Carr P. A study of the relational aspects of the culture of academic medicine. Acad Med. 2009;84:106–114
15. Carr P, Pololi L, Knight S, Conrad P. Collaboration in academic medicine: Reflections on gender and advancement. Acad Med. 2009;84:1447–1453
16. Pololi L, Kern DE, Carr P, Conrad P, Knight S. The culture of academic medicine: Faculty perceptions of the lack of alignment between individual and institutional values. J Gen Intern Med. 2009;24:1289–1295
17. Conrad P, Carr P, Knight S, Renfrew MR, Dunn M, Pololi L. Hierarchy as a barrier to advancement for women in academic medicine. J Womens Health (Larchmt). 2010;19:799–805
18. Pololi L, Kern DE, Carr P, Conrad P. Authors’ reply: Faculty values. J Gen Intern Med. 2010;25:647
19. Pololi LH Changing the Culture of Academic Medicine: Perspectives of Women Faculty. 2010 Hanover, NH Dartmouth College Press
20. Pololi L, Cooper LA, Carr P. Race, disadvantage and faculty experiences in U.S. academic medicine. J Gen Intern Med. 2010;25:1363–1369
21. Pololi LH, Jones SJ. Women faculty: An analysis of their experiences in academic medicine and their coping strategies. Gend Med. 2010;7:438–450
23. Butler JK. Toward understanding and measuring conditions of trust: Evolution of a conditions of trust inventory. J Manage. 1991;17:643–663
24. Pololi L, Price J. Validation and use of an instrument to measure the learning environment as perceived by medical students. Teach Learn Med. 2000;12:201–207
25. Pololi LH, Dennis K, Winn GM, Mitchell J. A needs assessment of medical school faculty: Caring for the caretakers. J Contin Educ Health Prof. 2003;23:21–29
28. Raghunathan TE, Lepkowski JM, Van Hoewyk J, Solenberger PW. A multivariate technique for multiply imputing missing values using a sequence of regression models. Surv Methodol. 2001;27:85–95
29. Rubin DB Multiple Imputation for Nonresponse in Surveys. 1987 New York, NY Wiley
30. Hatcher L A step-by-step approach to using the SAS system for factor analysis and structural equation modeling. 1994 Cary, NC SAS Institute
31. Raudenbush SW, Bryk AS Hierarchical Linear Models: Application and Data Analysis Methods. 20022nd ed Newbury Park, CA Sage
32. Ryan RM, Deci EL. Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. Am Psychol. 2000;55:68–78
34. Christakis DA, Feudtner CTemporary matters.. The ethical consequences of transient social relationships on medical training. JAMA. 1997;278:739–743
35. Safran DG, Miller W, Beckman H. Organizational dimensions of relationship-centered care. J Gen Intern Med. 2006;21(1 Suppl):S9–S15
36. Institute of Medicine. To Err Is Human: Building a Safer Health System. 2000 Washington, DC National Academy Press
37. Victor B, Cullen JB. The organizational bases of ethical work climates. Adm Sci Q. 1988;33:101–125
38. Higgins A, Power C, Kolhlberg LKurtines W, Gewirtz J. The relationship of moral atmosphere to judgments of responsibility. Morality, Moral Behavior and Moral Development. 1984 New York, NY Wiley:74–106
39. Chervenak FA, McCullough LB. Professionalism and justice: Ethical management guidelines for leaders of academic medical centers. Acad Med. 2002;77:45–47
40. Feudtner C, Christakis CA. Do clinical clerks suffer ethical erosion? Students’ perceptions of their clinical environments and personal development. Acad Med. 1994;69:670–679
41. Crandall SJ, Volk RJ, Loemker V. Medical students’ attitudes toward providing care for the socially underserved: Are we training socially responsible physicians? JAMA. 1993;269:2519–2523
42. Cottingham AH, Suchman AL, Litzelman DK, et al. Enhancing the informal curriculum of a medical school: A case study in organizational culture change. J Gen Intern Med. 2008;23:715–722
43. Sachdeva AK, Cohen R, Dayton MT, et al. A new model for recognizing and rewarding the educational accomplishments of surgery faculty. Acad Med. 1999;74:1278–1287
44. Bluedorn A. The theories of turnover: Causes, effects and meaning. Res Sociol Organ. 1982;1:75–128
45. Buchbinder S, Wilson M, Melick C, Powe N. Primary care physician job satisfaction and turnover. Am J Manag Care. 1995;5:1431–1438
46. Beasley B, Kern D, Kolodner K. Job turnover and its correlates among residency program directors in internal medicine. Acad Med. 2001;76:1127–1135
47. Pololi LH, Evans A. Culture change through faculty mentoring programs at Weill Cornell Medical College. [unpublished]
48. Pololi LH, Kern DE, Krupat E. C-Change Learning Action Network: Learning from evidence and experience as a culture change process in academic medicine. [unpublished]