Abstract
Faculty attrition at academic health centers (AHCs) is significant at about 11% nationally, with one in five physicians intending to leave, and replacement costs averaging $500,000 per physician. Attrition among AHC faculty leaders is inadequately studied. This study compares reasons to leave between exiting faculty leaders and faculty non-leaders at the University of New Mexico School of Medicine (UNM SOM). The SOM deans interview all exiting faculty using a structured exit survey. 329 faculty non-leaders and 58 faculty leaders left UNM SOM between July 2017 and June 2022. Distributions of each variable were analyzed for statistically significant differences between the two groups using Fisher’s 2-sided exact test. Text comments by leaders were analyzed qualitatively for content using a team-based, iterative process. As compared to non-leaders, exiting faculty leaders were more likely to be professors (51.7% vs 16.7%, p<0.001), and hold tenure (32.8% vs. 12.2%, p=0.001). Faculty leaders were more likely than non-leaders to cite high-level leadership as a reason to leave (41.4% vs. 24.3% p=0.01) and better leadership as a critical issue in development and retention (51.7% vs. 36.8% p=0.04). Qualitative analyses of textual leader comments showed a similar distribution of themes as the quantitative variables when examining open text related to the survey questions related to reasons to leave and the most critical issues. In addition, when asked what would need to change for them to return, qualitative data showed open-ended responses by exiting faculty leaders were twice as frequent to include leadership comments than those by non-leaders (34.2% vs. 16.2%). Exiting faculty leaders disproportionately cite high-level leadership as a reason to leave. The mediatory factors for this association are not known. Investigations to determine the causes for the study findings, and data-driven intervention strategies to retain faculty leaders at SOMs are needed.
Introduction & Literature Review
Faculty attrition at academic health centers (AHCs) is significant at about 11% nationally, with one in five physicians intending to leave (Abbasi, 2022), and replacement costs averaging $500,000 per physician (Shanafelt et al., 2017). Half of new assistant professors leave Schools of Medicine (SOMs) within eight years of hire (Liu & Morrison, 2014) and 57% leave within ten years (Alexander & Lang, 2008). Attrition among AHC faculty leaders is inadequately studied and the reasons why faculty leaders leave an AHC are not well-known. This deficiency constitutes a critical gap in the literature since the replacement of academic leaders costs the AHC even more than academic clinicians, with the costs going beyond financial cost to also include social, societal, and reputational costs. There is, therefore, a great need to sustain this human capital. This study compares reasons to leave between exiting faculty leaders and non-leaders at the University of New Mexico School of Medicine (UNM SOM). The long-term goal of the study is to help develop data-driven strategies to mitigate leader attrition.
Content
Study Design and Sample
In this cross-sectional mixed methods study, investigators collected and analyzed data from exit surveys and interviews from 329 faculty non-leaders and 58 faculty leaders exiting the University of New Mexico School of Medicine (UNM SOM), a medium-sized research-intensive AHC, over five years from July 2017-June 2022.
Procedures
The UNM SOM administers a mandatory exit survey to all faculty using the secure web-based REDCap platform (Harris et al., 2009). In the two-step process described in previous publications, the exit survey is followed by an exit interview with a SOM Dean, either face-to-face or virtually (Greenberg et al., 2021; Sood et al., 2018). Nearly all exiting faculty complete this process since their non-participation results in withheld annual leave payout. The exit survey was matched with a faculty database that monitors faculty during their entire time at UNM SOM, providing reliable estimates of termination dates and demographic characteristics.
Study variables
We studied the association between academic leadership status, faculty status and variables related to leaving. Leaders were defined as Presidents of all types (such as Vice-Presidents), Deans of all types (such as Associate Deans), Chairs of all types (such as Vice Chairs), Division Chiefs, Section Chiefs, and Directors (including fellowship and residency directors) but excluding medical directors. Variables related to leaving included new position type, reasons to leave, and perceived critical issues in faculty development and retention.
Data Analysis
In this study, investigators used a convergent parallel design, analyzing quantitative and qualitative data separately and then converging to compare and interpret results. The UNM Human Research Protection Office, which served as the Institutional Review Board of record (HRPO 17–347), approved this study.
Quantitative analyses included a tabulation of the proportions of each variable to compare faculty leader and non-leader responses of the selected exit survey variables. Significant differences (p<0.05) in variable proportions were compared using Fisher’s exact two-sided tests of significance. Stata 15 (College Station, Texas) software was used for all tabulations and statistical analyses.
Qualitative analyses of textual exit survey responses were analyzed for content using a team-based, iterative process and NVivo software (version 14, Lumivero). After each closed-ended question, respondents had the opportunity to provide additional information, which was coded using Shanafelt and Noseworthy’s (Magdaleno et al., 2022; Shanafelt & Noseworthy, 2017) dimensions of provider burnout as themes and added emergent codes. In the current project, investigators conducted content analysis by querying distribution of leader qualitative responses against all codes for all open-ended survey questions and then queried distribution specific to the survey questions asked.
Results
Of the 58 exiting faculty leaders included, only five were senior leaders (i.e., chairs and deans). Faculty leaders were more likely than non-leaders to be professors and hold tenure (p ≤ 0.001, Table 1). There were no other significant differences in demographic variables between exiting faculty leaders and non-leaders.
Table 1.
Demographics: Exiting Faculty Leaders Versus Non-Leaders
| Variable | Categories | Non-Leaders %(n=329) | Leaders %(n=58) | p-value |
|---|---|---|---|---|
| Gender | Male | 45.9 | 51.7 | 0.55 |
| Female | 48.9 | 41.4 | ||
| Gender Diverse | 0.6 | 0.0 | ||
| White | Yes | 81.8 | 79.3 | 0.71 |
| Asian | Yes | 17.3 | 15.5 | 0.85 |
| Ethnicity | Hispanic | 10.9 | 5.2 | 0.24 |
| MD is Terminal Degree | Yes | 72.6 | 75.9 | 0.87 |
| FTE | 1 | 78.4 | 87.9 | 0.31 |
| 0.5= < 1.0 | 17.0 | 10.3 | ||
| < 0.5 | 4.6 | 1.7 | ||
| Rank | Assistant Professor | 49.8 | 20.7 | <0.001 |
| Associate Professor | 24.6 | 24.1 | ||
| Professor | 16.7 | 51.7 | ||
| Lecturer | 6.7 | 3.5 | ||
| Status | Tenure | 12.2 | 32.8 | 0.001 |
| Non-Tenured | 73.6 | 53.5 | ||
| Rolling Contract | 14.3 | 13.8 |
Exiting faculty leaders were less likely than non-leaders to transfer to private practice (p=0.04); none of the other new position types were significantly different between the two groups (Table 2).
Table 2.
Exited Faculty by New Position Type
| New Position is: | No Leaders % (n=329) | Leaders % (n=58) | p-value |
|---|---|---|---|
| Academic Institution | 40.7 | 48.3 | 0.31 |
| Other | 28.9 | 29.3 | 1.00 |
| Private Practice | 24.3 | 12.1 | 0.04 |
| Industry | 4.9 | 6.9 | 0.52 |
As compared to non-leaders, exiting faculty leaders were almost twice as likely to express concern with high-level leadership (i.e., departmental or higher-level) as a reason to leave (p=0.01, Table 3). Conversely, there may be a trend for leaders to less frequently cite concerns with lower-level leadership (i.e., divisional leadership, p=0.07) and more frequently cite inadequate ancillary support staff and services (p=0.09) and a more prestigious institution/company (p=0.09) as reasons to leave than non-leaders. None of the other reasons to leave variables were significantly different between the two groups (Table 3).
Table 3.
Reasons to Leave: Exiting Faculty Leaders Versus Non-Leaders
| Multiple Check List Variables: What was the primary or other factors that influenced your decision to leave? | Non-Leaders % (n=329) | Leaders % (n=58) | p-value |
|---|---|---|---|
| Concern with Departmental or higher-level leadership | 24.3 | 41.4 | 0.01 |
| Personal/family matters | 35.9 | 37.9 | 0.77 |
| Work environment | 29.8 | 36.2 | 0.36 |
| Work life balance | 28.6 | 32.8 | 0.53 |
| Greater career opportunity | 36.2 | 27.6 | 0.23 |
| Salary | 31.9 | 25.9 | 0.44 |
| Inadequate ancillary support staff and services | 15.8 | 25.9 | 0.09 |
| Lack of recognition | 20.4 | 24.1 | 0.49 |
| Inadequate mentorship | 14.0 | 17.2 | 0.54 |
| More prestigious institution/company | 8.2 | 15.5 | 0.09 |
| Better geographic location | 14.3 | 12.1 | 0.84 |
| Retirement* | 13.1 | 13.8 | 0.84 |
| Concern with Divisional leadership | 16.4 | 6.9 | 0.07 |
| Inadequate adherence to the FTE split as per hiring contract | 6.1 | 6.9 | 0.77 |
| Lack of promotion | 9.1 | 3.5 | 0.20 |
| Concern with Departmental finances | 7.3 | 3.5 | 0.40 |
| Concern with Diversity/Equity/Inclusion* ** | 1.5 | 3.5 | 0.28 |
| Tenure status | 4.0 | 1.7 | 0.70 |
| *Primary only, **added June 2021 |
Similarly, exiting faculty leaders were more likely to cite better leadership as a perceived critical issue in development and retention compared to non-leaders (p=0.04, Table 4). None of the other differences in critical issues in development and retention were significant between the two groups, (Table 4).
Table 4.
Exited Faculty by Critical Issue in Development and Retention
| What do you perceive to be the most critical issues in faculty development and retention? (Multiple Check List Variables) | Non-Leaders % (n=329) | Leaders % (n=58) | p-value |
|---|---|---|---|
| Better leadership | 36.8 | 51.7 | 0.04 |
| Better Compensation | 53.5 | 53.5 | 1.00 |
| More support for academic pursuits | 33.1 | 43.1 | 0.18 |
| More effective mentoring of Junior Faculty | 41.3 | 41.4 | 1.00 |
| Better organized faculty development efforts | 29.2 | 29.3 | 1.00 |
| More recognition for academic pursuits | 21.9 | 27.6 | 0.40 |
| Better work-life balance | 26.4 | 27.6 | 0.87 |
| Better opportunities for research and/or teaching | 26.1 | 24.1 | 0.87 |
| More time for academic pursuits | 27.7 | 22.4 | 0.52 |
| Greater appreciation of diversity | 12.8 | 19.0 | 0.22 |
| Better adherence to FTE split as per hiring contract | 11.2 | 17.2 | 0.20 |
| Loan repayment | 15.5 | 8.6 | 0.22 |
| Better spousal and family support | 7.9 | 6.9 | 1.00 |
Qualitative data showed similar distribution of themes as the quantitative variables, when examining open text related to the survey questions on reasons to leave and the most critical issues for development and retention. For example, one leader explained, “Chair searches are a popularity contest…This prevents change. Chair term should be limited.” Additionally, when asked what would need to change for them to return, qualitative data showed leaders doubly included leadership comments than those by non-leaders (34.2% vs. 16.2% of open-ended responses, Table 5). Other factors that were two or more times more common in leaders were related to career development, autonomy, opportunities for collaboration, and retention/staffing limitations. Non-leaders’ comments on personal/family matters were two times more frequent than leaders See Table 5 for example comments.
Table 5.
Qualitative codebook with distribution of leader and non-leader comments in response to, “What needs to change at your UNM college/school for you to return?”
| Name | Description & Example Comments | Non-Leaders % (n=287)* | Leaders % (n=53)* |
|---|---|---|---|
| WORKLOAD & JOB DEMANDS, CONTROL & FLEXIBILITY | |||
| a. Compensation | Compensation package (including paid time off and other benefits), getting paid for more job duties. “…Get everybody up to the 25th percentile. Loan repayment would help faculty recruitment…” | 13.6 | 9.8 |
| b. Full-time equivalent (FTE)/ appointment | Time spent among types of duties (research/teaching/clinical). “Faculty don’t have time and resources needed to accomplish academic goals…clinical practice needs overwhelm academic goals.” | 6.1 | 2.4 |
| c. Workload | General comments about workload, vague mentions of burnout. “I should have not accepted the interim chief status of another division since this increased my workload.” | 2.2 | 0.0 |
| EFFICIENCY & RESOURCES | |||
| a. Efficiency of practice | Organization, patient processes, EHRs. Things others can do (e.g., support staff). “MDs waste time doing clerical work in clinics. Teambased model did not work as well because of lack of MAs.” | 4.4 | 7.3 |
| b. Resources | Funding/resources for facilities, job positions. “The IT system belongs to the stone ages…Everyday niggles such as computer not turned on in the clinic decrease your efficiency.” | 6.6 | 2.4 |
| c. Retention/staffing | Turnover, being short-staffed. “Limited resources make innovation and recruitment difficult. The last three faculty recruits were lost to Presbyterian due to poor and inequitable salary.” | 2.2 | 7.3 |
| MEANING IN WORK | |||
| a. Career development | Career advancement, leadership opportunities, faculty/professional development, mentorship, periodic reviews, and promotion. “I did not have a designated mentor…every faculty [needs] a mentor.” | 4.0 | 12.2 |
| b. Feeling valued/supported | Feeling like they have a voice or autonomy, valued. Being treated fairly. “[I] lost the belief that senior leadership is interested in listening to me. Lack of responsiveness to my concerns…” | 7.9 | 2.4 |
| c. Fit, career change | “Fit” between personal interests/skills and UNM/department/school. “Work in a different department to advise students on careers.” |
4.0 | 0.0 |
| CULTURE & VALUES; SOCIAL SUPPORT & COMMUNITY AT WORK | |||
| a. Culture & people | Behaviours of people, workplace culture, transparency. “Less ambiguous department finances…the department confiscated rather than paying out substantial funds in the spending accounts. [This] has had a significant detriment on …” | 4.4 | 2.4 |
| □ Leadership | Comments about leadership, specific characteristics/traits, or actions/inactions. “I hope that new institutional leadership will bring lasting, effective, and meaningful change…morale is incredibly low.” | 16.2 | 34.2 |
| b. Autonomy (dept) | Need for department/area to have autonomy to do more independently. (Example quotes would identify specific programs and/or individuals.) | 1.3 | 4.9 |
| c. Clinical mission | Patient care and environment, clinical mission/service/structure. “A stronger commitment to value-based care. UNM is really a hospital system and not truly a health system.” | 6.1 | 9.8 |
| d. Education mission | Teaching environment, structure. (Example quotes would identify specific programs and/or individuals.) | 3.5 | 0.0 |
| e. Research mission | Research environment, structure. “…the major issue is the time it takes for each faculty member to deal with the logistics of research, taking away from real research time…inefficient systems, under-staffing and lack of staff support, lack of communication between university offices/systems, IT issues, etc.” | 3.5 | 2.4 |
| f. Values | Organizational values alignment. Diversity of faculty, treatment/ consideration of cultural/ethnic minorities or women. “Decrease in gender bias with respect to advancement, leadership and salary.” | 4.0 | 7.3 |
| g. Collaboration | Opportunities or willingness to collaborate among people or departments. “The separate silos of UH, HSC, Med… This negatively affects patient care, program development, hiring and retention, cost, research, scholarship, teaching, time and resource allocation, morale, collaboration.” | 1.8 | 4.9 |
| WORK-LIFE INTEGRATION | |||
| a. Hours/schedule | Vague “work-life balance” comments; call schedule, night/weekend work, long hours, ability to take time off, part-time work, ability to spend time with family/on non-work activities. “There is no shift work on weekend calls that last 72 hours. Being on call is burdensome. I need to spend more time with a grandchild.” | 4.4 | 0.0 |
| b. Personal/family matters | Personal/family issues. Includes children, elderly parents, spouse/ partner, and specific community or geography reasons. “Desire to be closer to family and elderly parents.” | 10.5 | 4.9 |
| c. Retirement | Reason for leaving is retirement; includes returning as a working/ volunteer retiree. “I am returning to work 0.25 FTE. The process of retirement and resuming work at 0.25 FTE is cumbersome.” | 5.7 | 0.0 |
n respondents who provided at least one open-ended response; percentage is of this n, not n of whole study sample.
Discussion
Schools of medicine in the U.S. report a high turnover of academic leaders. For example, from 1993 to 2003, faculty leaders held the position of Pediatrics departmental chair for a mean of 5.6 ± 3.2 years (median: 5 years) (Stapleton et al., 2005). An earlier study showed that the tenure of deans of medicine was even shorter at 4.1 ± 2.7 years during the period 1980–1992, and declining since previous years (Banaszak-Holl & Greer, 1994). It is known that leadership in academic medical institutions has grown increasingly complex as budgetary demands, clinical competition, and regulatory compliance have become increasingly challenging. The rapid turnover in leadership can lead to institutional instability and can be an impediment to institutional progress.
The study findings indicate that exiting faculty leaders cite high-level leaders at a school of medicine as a reason to their leaving. While the evidence gathered in the study does not show direct causal links, the association may suggest the problem may be either characteristics of the leadership styles/personalities/skills of individual high-level leaders or characteristics of the organization that the high-level leaders represent. It is possible that exiting faculty leaders may not be able to differentiate the two factors (high-level leaders vs. organization), when ascribing high-level leaders as a reason to leave.
Characteristics of leadership styles, personalities, and skills of individual high-level leaders are important for workforce stability of an organization. For example, both management skills (e.g., institutional assessment, strategic planning, financial stewardship, and recruitment and retention of talent) and leadership skills (e.g., visioning, maximizing values, and building constituency) are commonly cited characteristics important for deans of medical schools (Rich et al., 2008). Key content knowledge (e.g., academic medical center governance, expectations of clinicians and scientists, and process of medical education) and certain attitudes (e.g., commitment to the success of others, and appreciation of institutional culture) are also noted to be valuable qualities (Rich et al., 2008). A common problem with blaming an individual high-level leader’s style/personality/skill is that it ignores the role that organizational context plays in explaining what high-level leaders do, how they do it, and why. Organizational factors that may contribute to the exit of faculty leaders may relate to organizational culture/climate, or environmental or structural conditions (Banaszak-Holl & Greer, 1994; Sood et al., 2020). This premise is consistent with the organizational theory which emphasizes the importance of the organization in explaining senior administrative turnover (Pfeffer & Salancik, 1977).
The study findings also indicate additional suggestions that might explain why faculty leaders leave the school of medicine. High-level leaders provide mentorship, resources, responsibilities, visibility, role modeling etc. to lower-level leaders. The trend for disproportionate complaints of inadequate ancillary support staff and services by exiting faculty leaders in the study might indicate inadequate provision of these resources by high-level leaders (Sood et al., 2020). Others have indicated that leaders need training in soft skills such as emotional intelligence and communication as well as hard skills such as strategic planning and budgetary management. While organizations often invest funds for training high-level leaders, similar resources are not invested in other faculty leaders.
The strengths of the study include its relatively large sample size, and a significant proportion of women and non-White faculty participants. The weaknesses of the study include the single study site limiting generalizability of study findings and cross-sectional study design that does not establish causality. Furthermore, since only five of the 58 exiting leaders were high-level leaders (i.e., Chairs and Deans), this study does not adequately address why high-level leaders leave.
Conclusion
As compared to non-leaders, exiting faculty leaders more frequently cite high-level leaders as their reason to leave the school of medicine. The mediatory factors for this association are not known. When leaders at any level leave, the consequences can be significant including loss of morale, a decline in productivity, downstream workforce attrition, and financial and reputational risk for organizations. The study emphasizes the need for efforts to ascertain characteristics and design interventions that allow for sustained academic leadership. Our study also indicates the need for further research in differentiating characteristics of the leadership styles/personalities of individual high-level leaders versus those of the organization that drive faculty leaders away.
Acknowledgments
The NIH/NIGMS U01GM132175 (Sood, PI) and 2U54GM104944 (Sy, PI); and NIH/NCATS UL1TR001449 (Pandhi/Campen, PIs); and HRSA 1 D34HP45723-01-00 (PI Romero-Leggott) funded this study.
References
- Abbasi J (2022, Apr 19). Pushed to their limits, 1 in 5 physicians intends to leave practice. JAMA, 327(15), 1435–1437. 10.1001/jama.2022.5074 [DOI] [PubMed] [Google Scholar]
- Alexander H, & Lang F (2008). The long-term retention and attrition of US medical school faculty. AAMC Analysis in Brief, 8, 1–2. [Google Scholar]
- Banaszak-Holl J, & Greer DS (1994). Turnover of deans of medicine during the last five decades. Acad Med, 69(1), 1–7. 10.1097/00001888-199401000-00001 [DOI] [PubMed] [Google Scholar]
- Greenberg N, Myers O, Lawrence E, & Sood A (2021). Factors related to faculty work life balance as a reason to leave a school of medicine UNM Mentoring Conference, Albuquerque, NM. [PMC free article] [PubMed] [Google Scholar]
- Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, & Conde JG (2009). Research electronic data capture (REDCap) - A metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform, 42(2), 377–381. 10.1016/j.jbi.2008.08.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Liu CQ, & Morrison E (2014). U.S. medical school full-time faculty attrition. Analysis in Brief, AAMC, 14(2). [Google Scholar]
- Magdaleno Y, Rishel Brakey H, Greenberg N, Myers O, & Sood A (2022). A qualitative review of comments by faculty who cite work-life balance as a reason to leave. Chronicle Mentoring Coaching, 6(15), 587–593. i [PMC free article] [PubMed] [Google Scholar]
- Pfeffer J, & Salancik GR (1977). Organizational context and the characteristics and tenure of hospital administrators. Academic Management Journal, 20(1), 74–88. https://www.ncbi.nlm.nih.gov/pubmed/10305921 [PubMed] [Google Scholar]
- Rich EC, Magrane D, & Kirch DG (2008). Qualities of the medical school dean: Insights from the literature. Academic Medicine, 83(5), 483–487. 10.1097/ACM.0b013e31816becc9 [DOI] [PubMed] [Google Scholar]
- Shanafelt T, Goh J, & Sinsky C (2017). The business case for investing in physician well-being. JAMA Intern Med, 177(12), 1826–1832. 10.1001/jamainternmed.2017.4340 [DOI] [PubMed] [Google Scholar]
- Shanafelt TD, & Noseworthy JH (2017). Executive leadership and physician well-being: nine organizational strategies to promote engagement and reduce burnout. Mayo Clinic Proc, 92(1), 129–146. 10.1016/j.mayocp.2016.10.004 [DOI] [PubMed] [Google Scholar]
- Sood A, Parker T, Nieto N, Maloney J, Greenberg N, Rishel Brakey H, Myers O, Sigl D, & Wilson B (2018). Medicine faculty consider mentoring as the critical issue for development and retention (2018) Chronicle of Mentoring and Coaching, 2(Special Issue 1), 377–383. [Google Scholar]
- Sood A, Rishel Brakey H, Myers O, Greenberg N, Tigges B, Sigl D, & Wilson B (2020). Exiting medicine faculty want the organizational culture and climate to change. Chron Mentor Coach, 4(SI13), 359–364. https://www.ncbi.nlm.nih.gov/pubmed/33244484 [PMC free article] [PubMed] [Google Scholar]
- Stapleton FB, Jones D, & Fiser DH (2005). Leadership trends in academic pediatric departments. Pediatrics, 116(2), 342–344. 10.1542/peds.2004-2426 [DOI] [PubMed] [Google Scholar]