Scholarly article on topic 'Burnout among psychiatrists in the Veterans Health Administration'

Burnout among psychiatrists in the Veterans Health Administration Academic research paper on "Psychology"

CC BY-NC-ND
0
0
Share paper
Academic journal
Burnout Research
OECD Field of science
Keywords
{Burnout / "Veterans Affairs" / Psychiatry / Psychiatrist / Retention / Workplace}

Abstract of research paper on Psychology, author of scientific article — Hector A. Garcia, Cindy A. McGeary, Erin P. Finley, Norma S. Ketchum, Donald D. McGeary, et al.

Abstract Research suggests that mental health workers are at high risk for burnout, and that burnout can increase staff turnover and reduce quality of care. The Veterans Administration (VA) employs over 3000 psychiatrists across the United States, but little is known about burnout in this population. This study was conducted to examine predictors of burnout and intent to leave the VA among a national sample of VA psychiatrists. Participants (N =125) responded to an anonymous online survey. Regression analyses were used to examine relationships between workplace variables, patient characteristics, and burnout as measured by the Maslach Burnout Inventory-General Survey – which includes sub-scales for cynicism, exhaustion, and professional efficacy – as well as intent to leave the VA. Based on established cut-off scores, 90% of the sample reported high cynicism, 86% reported high exhaustion, and 74% reported high professional efficacy. High cynicism predicted the intent to leave the VA (p =.004). Not feeling part of a coherent team predicted greater cynicism (p =.01), and patient characteristics such as suspected malingering showed a positive trend with cynicism (p =.05). Workplace characteristics such as unfair treatment by supervisors (p =.03) and insufficient resources (p =.001) predicted greater exhaustion. The current findings suggest that burnout is prevalent in the VA psychiatry workforce. Specific administrative measures to reduce burnout may have potential to improve the emotional health of that workforce and ensure high quality of care for the veteran population it serves. The size of both the VA psychiatry workforce and patient population underscores the importance of greater understanding of burnout as it occurs in the VA.

Academic research paper on topic "Burnout among psychiatrists in the Veterans Health Administration"

ELSEVIER

Burnout among psychiatrists in the Veterans Health Administration******

Hector A. Garciaa b *, Cindy A. McGearyb, Erin P. Finleya c, Norma S. Ketchumd, Donald D. McGearyb, Alan L. Petersonb

a South Texas Veterans Health Care System, United States

b Department of Psychiatry, The University of Texas Health Science Center at San Antonio, United States

c Department of Medicine, The University of Texas Health Science Center at San Antonio, United States

d Department of Epidemiology and Biostatistics, The University of Texas Health Science Center at San Antonio, United States

ABSTRACT

Research suggests that mental health workers are at high risk for burnout, and that burnout can increase staff turnover and reduce quality of care. The Veterans Administration (VA) employs over 3000 psychiatrists across the United States, but little is known about burnout in this population. This study was conducted to examine predictors of burnout and intent to leave the VA among a national sample of VA psychiatrists. Participants (N =125) responded to an anonymous online survey. Regression analyses were used to examine relationships between workplace variables, patient characteristics, and burnout as measured by the Maslach Burnout Inventory-General Survey - which includes sub-scales for cynicism, exhaustion, and professional efficacy - as well as intent to leave the VA. Based on established cut-off scores, 90% of the sample reported high cynicism, 86% reported high exhaustion, and 74% reported high professional efficacy. High cynicism predicted the intent to leave the VA (p = .004). Not feeling part of a coherent team predicted greater cynicism (p = .01), and patient characteristics such as suspected malingering showed a positive trend with cynicism (p = .05). Workplace characteristics such as unfair treatment by supervisors (p = .03) and insufficient resources (p = .001) predicted greater exhaustion. The current findings suggest that burnout is prevalent in the VA psychiatry workforce. Specific administrative measures to reduce burnout may have potential to improve the emotional health of that workforce and ensure high quality of care for the veteran population it serves. The size of both the VA psychiatry workforce and patient population underscores the importance of greater understanding of burnout as it occurs in the VA.

© 2015 Published by Elsevier GmbH. This is an open access article under the CC BY-NC-ND license

(http://creativecommons.org/licenses/by-nc-nd/4.0/).

Contents lists available at ScienceDirect

Burnout Research

journal homepage www.elsevier.com/locate/burn

CrossMark

ARTICLE INFO

Article history: Received 23 February 2015 Received in revised form 28 September 2015 Accepted 7 October 2015

Keywords:

Burnout

Veterans Affairs

Psychiatry

Psychiatrist

Retention

Workplace

1. Introduction

The Veterans Health Administration (VA) oversees the largest integrated mental health care system in the United States. Psychiatrists comprise a sizeable portion of the VA mental health workforce - at present the administration employs more than 3000

* The research was supported by the Practice Based Research Network, University of Texas Health Sciences Center San Antonio.

** The views expressed in this article are solely those of the authors and do not represent the views of or an endorsement by the U.S. Army, the Department of Defense, the Department of Veterans Affairs, or the U.S. Government.

*** This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

* Corresponding author at: Frank Tejeda Outpatient Clinic, 5788Eckhert Road, San Antonio, TX 78240, United States.

E-mail address: Hector.Garcia2@va.gov (H.A. Garcia).

psychiatrists nationwide (D. Perry [VA], personal communication, April 30, 2014). Occupational burnout is often a concern among mental health providers, and those working in the VA may be particularly at risk because of the VA's unique patient population and bureaucratic demands (Voss Horrell, Holohan, Didion, & Vance, 2011). Given the size and critical health care role of the VA psychiatry workforce, burnout in this population has potentially far-reaching implications.

The most influential model of professional burnout was developed by Christina Maslach based on work among health care and human resource employees. The model includes three factors: (1) exhaustion; (2) cynicism (or depersonalization); and (3) ineffectiveness (Maslach, 1982; Maslach & Leiter, 1997). Exhaustion occurs when a provider's emotional energy is depleted as a result of work demands, such as those from patients, supervisors, or the organization (Maslach, 1988; Maslach, Schaufeli, & Leiter, 2001). Cynicism refers to detachment arising as a means to gain emotional distance

http://dx.doi.org/10.1016/j.burn.2015.10.001

2213-0586/© 2015 Published by Elsevier GmbH. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

from work, while ineffectiveness refers to feelings of incompetence or the inability to fulfill job responsibilities (Maslach & Goldberg, 1998).

International research has consistently found that psychiatrists experience greater burnout than other physicians. Psychiatrists have been found to have: higher levels of burnout compared to other physicians in Italy (Bressi et al., 2009); higher levels of emotional exhaustion (and severe depression) than other physicians in Scotland (Deary, Agius, & Sadler, 1996); higher levels of work-related exhaustion than psychiatric nurses in Sweden (Thomsen, Soares, Nolan, Dallender, & Arnetz, 1999); and less job satisfaction and higher emotional distress than other physicians in Finland (Heponiemi, Aalto, Puttonen, Vanska, & Elovainio, 2014). Little is known about how psychiatrists in the United States fare compared with other professionals, although one study of a managed care setting found that psychiatrists reported greater exhaustion and depersonalization than primary care physicians (Snibbe, Radcliffe, Weisberger, Richards, & Kelly, 1989).

Part of this elevated burnout risk may be explained by the unique demands of the mental health profession. For instance, research has found that occupations requiring "emotion work" - work regularly demanding empathy, dealing with negative emotions, suppressing or displaying emotions, etc. - uniquely contributes to occupational stress (Zapf, Seifert, Schmutte, Mertini, & Holz, 2001). In turn, specific types of patients may require more emotion work. International studies have shown psychiatrists experience increased stress when patients are aggressive and demanding (Bressi et al., 2009), homicidal or suicidal (Rathod et al., 2000), or irritated, disruptive, or anxious (Heponiemi et al., 2014). Voss Horrell et al. (2011) speculate that newly returning veterans from the wars in Iraq and Afghanistan may exhibit aggression levels that prove stressful to clinicians working with them. Little is known, however, about the actual impact of patient characteristics on burnout among psychiatrists in settings such as VA where high numbers of veterans receive psychiatric care.

Studies of professional burnout have also identified organizational characteristics that may increase burnout risk, including: (1) workload, including the availability of resources to meet work demands; (2) control over the manner in which work is conducted; (3) reward for work conducted (e.g. prestige, monetary incentives, etc.); (4) community, or the quality of relationships with leaders and peers; (5) fairness, related to the perception of openness and equity in workplace decision-making; and (6) values, or concordance of values between the employee and the organization (Bressi et al., 2009; Kumar, Hatcher, Dutu, Fischer, & Ma'u, 2011; Maslach & Leiter, 1997,2008; Maslach et al., 2001). Moreover, studies have found that burnout often leads to greater absenteeism and staff turnover (AlbuAlRub & Al-Zaru, 2008; Leiter & Maslach, 2009), which can have the secondary effect of lowering morale among remaining employees, leading to more turnover (Misra-Hebert, Kay, & Stoller, 2004). Turnover itself is associated with high financial costs related to recruitment, training, and lost productivity while positions remain vacant (Berger & Boyle, 1992; Gray, Phillips, & Normand, 1996; Mott, 2000; Stoller, Orens, & Kester, 2001). A study of one major academic hospital found that the annual costs associated with turnover amounted to 5% of the organization's annual operating budget, amounting to $29,303,390 (Waldman, Kelly, Arora, & Smith, 2004). Given that VA is the largest healthcare organization in the United States, with 152 medical centers -and close to 1400 domiciliaries, community based outpatient clinics, and other facilities (VHA, 2014) - a comparable rate of turnover costs at VA could easily run into the billions.

Importantly, burnout can also impact patient care. Nurses experiencing high burnout receive lower ratings on quality of care by their patients (Leiter, Harvie, & Frizzell, 1998). Burnout among psychiatric staff in inpatient and treatment homes is associated with

negative attitudes toward patients (Holmqvist &Jeanneau, 2006). Further, when burnout results in turnover, the loss of a trusted provider can lower patient satisfaction (Misra-Hebert et al., 2004).

Although the literature on psychiatry burnout is diverse and international, few studies have examined burnout among psychiatrists working in the United States. Two studies have examined burnout among mental health providers working with military and veteran populations, and both identified important relationships between provider burnout, patient characteristics, and organizational factors (Ballenger-Browning et al., 2011; Garcia et al., 2015).

The present research represents the first-known study to examine burnout among psychiatrists working in the VA. Study results may be used to improve job satisfaction and retention among VA psychiatrists.

2. Hypotheses

Based on prior studies we hypothesized that three workplace variables would positively predict psychiatrist self-reported burnout: organizational blame for patient suicide, time spent on prescribing/monitoring medication, and administrative tasks. We also hypothesized that less time spent in activities other than prescribing and administration, such as teaching, supervising other staff, or providing psychotherapy, would positively predict burnout in this population.

In addition, we hypothesized that several patient-level variables would positively predict burnout: patient relapse, treatment non-responders, medication non-compliance, threatening patients, assault by patients, patient suicides, and a caseload comprised of a high percentage of severely mentally ill patients. We also hypothesized that the frequency of certain diagnoses/patient presentations would positively predict burnout, including: anger, depression, PTSD, anxiety, substance abuse/dependence, axis II disorders or traits, suspected malingering, psychosis, OCD, cognitive disorder, bipolar disorder, and severe personality disorder.

3. Method

3.1. Participants and procedures

This study was approved by Institutional Review Boards of the South Texas Veterans Health Care System and the University of Texas Health Science Center at San Antonio. Psychiatrists working full time at the VA were eligible for study participation, including those working in inpatient, substance abuse, emergency room, general outpatient, PTSD clinical teams, domiciliary, and telemedicine settings.The researchers used the VA's national directory to obtain the name and e-mail address of VA psychiatrists - as of August 1, 2013, there were 2988 psychiatrists listed in the directory. From this list a random sample of 500 VA psychiatrists was generated. Those 500 were sent an email describing the study and inviting participation in an online survey hosted on Survey Monkey (www. surveymonkey.com).

Subjects were informed that their participation was completely voluntary and confidential, that no identifying information would be collected, and that information gathered from the survey would be reported only in aggregate. To further ensure confidentiality, the researchers chose Survey Monkey's option to decline the record of IP addresses. Two iterations of the invitation e-mail were sent at a three-week interval. Because this research had the opportunity to reach VA psychiatrists experiencing burnout, the survey also included a list of resources for managing distress and seeking help. We received responses from 125 participants for a 25% response rate.

3.2. Measures

This study utilized the Maslach Burnout Inventory General Survey (MBI-GS), a widely used self-report measure assessing burnout (Maslach, Jackson, & Leiter, 1996). The MBI-GS is self-administered and consists of 16 items divided into 3 subscales (exhaustion, cynicism, and professional efficacy), which have been confirmed in factor analyses (Bakker, Demerouti, & Schaufeli, 2002; Schaufeli, Leiter, Maslach, & Jackson, 1996). Five items make up the exhaustion subscale (e.g., "I feel used up at the end of the workday."), five items are included in the cynicism subscale (e.g., "I have become less enthusiastic about my work."), and six items comprise the professional efficacy subscale (e.g., "I have accomplished many worthwhile things in this job."). The MBI-GS uses a 0-6 Likert-type scale to indicate frequency of experiencing a particular feeling regarding work from Never to Daily. Higher scores on cynicism (score range 0-30) and exhaustion (score range 0-30) and lower scores on professional efficacy (score range 0-36) may indicate higher levels of burnout. The cut-off score for high levels of cynicism is 11, and for exhaustion is 16. The professional efficacy subscale is reverse scored, with lower scores indicative of burnout and a cut-off for low levels set at 10. The MBI-GS has been shown to have sufficient validity, reliability, and internal consistency across a variety of occupations, including the human services (Bakker et al., 2002; Maslach & Jackson, 1981; Pines & Maslach, 1978). In prior studies Cronbach alpha coefficients for the MBI-GS ranged from .84 to .90 for exhaustion, .74 to .84 for cynicism, and .70 to .76 for professional efficacy (Bakker et al., 2002).

In addition to the MBI-GS, participants were asked to respond to questions assessing demographics, patient care, patient characteristics, and workplace/organizational factors. Twenty patient care variables were examined, such as time spent: on call, on administrative tasks, providing therapy, prescribing medication, and teaching. Additional patient care variables assessed whether respondents: were allowed protected time for responsibilities other than patient care, hold teaching responsibilities, were given protected time for teaching, or supervise other mental health staff. Also assessed was whether psychiatrists believed their patients were getting better, how often patients experience relapse, percent of caseload comprised of severely mentally ill patients, and how often they are threatened or assaulted by a patient. Whether patients have committed suicide, if psychiatrists fear the organization will blame them, and percentage of patients that are medication non-compliant were also assessed. Thirteen patient characteristics, measured on a 5-point Likert-type scale ranging from Never to Very Often, were examined (anger, depression, PTSD, anxiety, substance abuse, AXIS II disorders, suspected malingering, psychosis, medication non-compliance, OCD, cognitive disorder, bipolar, and severe personality disorder). Workplace characteristics that have been linked to burnout were also assessed regarding: clinical work, staffing, feeling part of a team, administrative workload, organizational politics, emotional support, control, fairness, rewards, and resources. Responses for these ten items were measured on a 5-point Likert-type scale ranging from Strongly Disagree to Strongly Agree. Finally, psychiatrists were asked how likely it was that they would leave their VA job in the next two years, as measured on a 5-point Likert-type scale ranging from Not Likely to Very Likely.

3.3. Data analysis

Demographic variables were summarized using descriptive statistics. Internal reliability of the MBI-GS was assessed using Cronbach's alpha. Outcomes of interest consisted of the three MBI-GS subscales (professional efficacy, exhaustion, and cynicism) and the intent to leave the VA within 2 years. First, Spearman

correlations were examined between the four outcomes and all potential predictors of burnout (provider demographics, patient care, patient characteristic, and workplace/organizational). Second, those predictors showing significant correlations (p <0.05) with outcomes were subsequently included in the multivariate least squares regression models. The final multivariate linear regression models were selected using backward elimination. All variables remaining in the models produced F-statistics significant at the 0.10 level. A multivariate linear regression model was also used to assess the relationship between MBI-GS subscale and likelihood of leaving the VA within two years. Post-doctoral fellows and residents were excluded from analyses involving intent to leave the VA. To avoid missing an effect worthy of further study, reported p-values were not corrected for multiple testing. All statistical analyses were performed in SAS (Version 9.3; SAS Institute, Cary, NC).

4. Results

4.1. Demographics and descriptive statistics

Demographic data is summarized in Table 1. Most respondents were full-time VA employees (76%), non-Hispanic Caucasian (71.9%), and working in a general outpatient setting (54.5%). Gender was evenly split (50.4% female). Mean age was 51.7 (SD = 9.7), with mean years in practice of 19 (SD = 10.1) and the majority had worked at VA for greater than 5 years (65.3%).

Sixteen respondents did not complete the MBI-GS items resulting in a final sample of 109 subjects. Mean scores on the three scales of the MBI-GS are presented in Table 2. Professional efficacy, exhaustion, and cynicism showed good internal consistency for this study [Cronbach alpha = 0.77, 0.94, and 0.84, respectively]. Based on established MBI-GS cut-off scores, 86.2% of the sample endorsed high levels of exhaustion (M =24.8; SD = 7.6) and nearly 90% reported high cynicism (M = 21.1; SD = 7.9). Professional efficacy was also high at 74.3% (M = 33.6; SD = 5.8).

4.2. Patient and organizational factors and burnout

In contrast to our hypothesis, less time per week (not more time) spent teaching predicted greater professional efficacy (B = -2.948; p = .03). Receiving fair treatment (B = 1.468; p = .002) and feeling that patients are getting better (B = 2.457; p<.001) also predicted professional efficacy (see Table 3). Unfair treatment (B = -1.337; p = .03), having insufficient resources (B = -3.309; p<.001) and number of times on short call (B = .042; p = .01) predicted greater exhaustion. As hypothesized, percent of caseload comprised of severely mentally ill patients showed a positive trend with exhaustion (B = .411; p = .07). Finally, increased cynicism was predicted by not being part of a coherent team (B = -1.452; p = .01), not having protected time outside of patient care (B = -1.362; p = .004), and feeling patients are not getting better (B = -2.974; p <.001). In contrast to our hypothesis, working with patients exhibiting bipolar disorder resulted in not more, but less cynicism (B = -3.972; p<.001). Having control over work showed a negative trend with cynicism (B = -1.312; p. = .05), while, as hypothesized, working with patients suspected of malingering showed a positive trend with cynicism (B = 1.942; p = .05).

4.3. Patient and organizational factors and intent to leave

No patient characteristics predicted intent to leave (Table 4). Not being part of a coherent team (B = -0.545; p<.001) and not having protected time outside of patient care (B = -0.247; p = .02) predicted likelihood of leaving VA within two years.

Table 1

Demographic characteristics among 125 VHA psychiatrists.

Part-time VA employee, n (%) 34(27.2)

Full-time VA employee, n (%) 95 (76)

Age (years)

Mean (SD) 51.7(9.7)

Median [Q1,Q3] 53 [44,59]

Min, max 32,75

Gender, n (%)

Female 61 (50.4)

Male 60(49.6)

Marital status, n (%)

Married 92(76)

Never married 18 (14.9)

Separated 2 (1.7)

Widowed 9(7.4)

Number of children

Mean (SD) 1.9(1.4)

Median [Q1,Q3] 2[1,3]

Min, max 0,7

Ethnicity, n (%)

African American 6(5)

Caucasian 87(71.9)

Hispanic 6(5)

Native American 2 (1.7)

Other 20(16.5)

Training status, n(%)

Resident 2 (1.7)

Postdoctoral fellow 2 (1.7)

N/A 117(96.7)

Years in practice

Mean (SD) 19(10.1)

Median [Q1,Q3] 20[10,26]

Min, max 2,47

Years at VA, n (%)

1-2 9(7.4)

2-3 10(8.3)

3-4 8 (6.6)

4-5 15(12.4) >5 79(65.3)

Primary appointment, n (%)

Inpatient 8 (6.6)

Substance abuse 14 (11.6)

Emergency room 3 (2.5)

General outpatient 66 (54.5)

PCT 7 (5.8)

Telemed 3(2.5)

Other 20(16.5)

4.4. Burnout and intent to leave

High cynicism predicted likelihood of leaving the VA within two years (B = .067; p = .004). High exhaustion showed a positive trend with likelihood of leaving the VA within two years (B = .041; p = .07).

5. Discussion

5.1. Burnout overall

This is the first known study to investigate occupational burnout among VA psychiatrists. Burnout in this population has important implications for both the health ofVA providers and the subsequent care they provide to the veterans they serve. On a positive note, a high percentage of VA psychiatrists reported feeling that the work they are doing is effective; 74.3% scored in the high range on the MBI-GS professional efficacy scale. However, an alarmingly high percentage of the sample also reported high levels of occupational burnout, with 86% reporting high exhaustion and 90% reporting high cynicism. A similar pattern of high professional efficacy (88%

Table 2

MBI sub-scale scores and likelihood of leaving position within 2 years.

MBI professional efficacy subscore

Mean (SD) 33.6 (5.8)

Median [Q1,Q3] 35 [29,38]

Min, max 15,42

MBI exhaustion subscore

Mean (SD) 24.8 (7.6)

Median [Q1,Q3] 27 [20,30]

Min, max 5,35

MBI cynicism subscore

Mean (SD) 21.1(7.9)

Median [Q1,Q3] 21 [15,27]

Min, max 5,35

MBI professional efficacy subscore category, n(%)

Low 6(5.5)

Moderate 22(20.2)

High 81(74.3)

Total 109

MBI exhaustion subscore category, n (%)

Low 9(8.3)

Moderate 6(5.5)

High 94(86.2)

Total 109

MBI cynicism subscore category, n (%)

Low 2(1.8)

Moderate 9(8.3)

High 98(89.9)

Total 109

How likely is it that you will leave this position in two years? n (%)

Not likely 30(26.8)

Somewhat unlikely 14(12.5)

Neutral 16(14.3)

Somewhat likely 26(23.2)

Very likely 26(23.2)

Total 112

Table 3

Factors associated with MBI sub-scale scores among 109 VHA psychiatrists.

Variable Estimate SE p-Value

MBl-professional efficacy

Intercept 22.297 2.238 <0.001

Fair treatment 1.468 0.451 0.002

Time per week spent teaching -2.948 1.342 0.03

Feel patients are getting better 2.457 0.634 <0.001

MBl-exhaustion

Intercept 37.038 2.670 <0.001

Fair treatment -1.337 0.612 0.03

Have resources -3.309 0.668 <0.001

Times per year on short call 0.042 0.016 0.010

Percent ofcaseload comprised of 0.411 0.224 0.07

severely mentally ill patients

MBl-cynicism

Intercept 46.251 4.503 <0.001

Coherent team -1.452 0.584 0.01

Control over work -1.312 0.672 0.05

Allowed protected time outside patient -1.362 0.459 0.004

Feel patients are getting better -2.974 0.791 <0.001

Patients suspected of malingering 1.942 0.957 0.05

Patients exhibit bipolar disorder -3.972 1.059 <0.001

a Associations assessed using a linear regression model with variables selected using backward elimination.

Table 4

Factors associated with likelihood ofleaving position within 2 years among 107 VHA psychiatrists.

Variable Estimate SE p-Valuea

Likelihood of leaving position in 2 years

Intercept 5.292 0.418 <0.001

Coherent team -0.545 0.113 <0.001

Allowed protected time outside patient care -0.247 0.100 0.02

a Associations assessed using a linear regression model with variables selected using backward elimination.

moderate-high) occurring alongside high exhaustion (50%) and cynicism (47%), was found in a study examining burnout among non-prescribing VA mental health clinicians (Garcia, McGeary, McGeary, Finley, & Peterson, 2014). It is concerning that psychiatrists in the current sample reported high exhaustion and cynicism much more frequently than non-prescribing clinicians in the prior work.

5.2. Burnout sub-scales 5.2.1. Cynicism

Notably, cynicism levels were the highest among the burnout scales and cynicism was the only scale that significantly predicted intent to leave the VA within two years. Given that a high number of VA psychiatrists (46.4%) reported their intent to leave as Likely or Very Likely (by comparison, in the non-prescriber study by Garcia et al. (2014), that figure was 32%), these findings suggest cynicism may have a far-reaching negative impact - not only because psychiatrists may vacate positions due to occupational burnout, but because staff departure often places emotional strain on patients who must adjust to new providers, and on remaining staff who must absorb the workloads of their vacating colleagues (Misra-Hebert et al., 2004).

Our study found that several workplace variables predicted cynicism, including not having a sense of control over how work is conducted. Lower perceived control has consistently been found to predict burnout across occupations (Maslach et al., 2001), including among non-prescribing PCT clinicians (Garcia et al., 2014). Other researchers have expressed concern that, as an enormous federal bureaucracy, the VA may inadvertently increase burnout risk by reducing mental health providers' sense of control through mandates for scheduling and practice and performance goals (Voss Horrell et al., 2011). Interestingly, not being allowed time outside of patient care also predicted cynicism in the current sample. It may be that returning some administrative control to providers - by allowing greater flexibility in scheduling or time outside of direct patient care (e.g., for research, serving on committees, etc.) - can reduce burnout. Current concerns about understaffing at the VA and the impact on timely access to mental health care for veterans (IOM, 2012), however, may sustain scheduling pressure on clinics, making it difficult to achieve greater flexibility for providers.

Not feeling part of a coherent team also predicted cynicism. Research across various occupations (c.f. Burke, Greenglass, & Schwarzer, 1996; Cherniss, 1980; Firth & Britton, 1989), to include international studies of psychiatry (Bressi et al., 2009; Kumar et al., 2011), have consistently noted the role of working relationships in the experience of burnout. Similarly, prior research on non-prescribing VA providers found that lack of team coherence predicted taking more mental health days off work, an indicator of burnout (Garcia et al., 2014). Peer supervision and consultation have been recommended to mitigate the stress of providing mental health care (Foa, Hembree, & Rothbaum, 2007), and future research should consider how best to strengthen and maintain team coherence within VA settings.

Patient characteristics also predicted cynicism - specifically, working with patients suspected of malingering showed a positive (non-significant) trend with cynicism (p = .05). Suspected malingering has been found to predict cynicism in VA non-prescribers (Garcia et al., 2015). Potik, Feldinger, and Schreiber (2012) argue that malingering patients lack incentive to respond to treatment, and that providers working with patients they suspect of malingering may experience increased feelings of anger, despair, or hopelessness. Taylor, Freuh, and Asmundson (2007) point out that cynicism may arise in providers working in settings where malingering is prevalent, who may begin to suspect malingering in a greater subset of their patients than are actually malingering. Other scholars have raised concerns that malingering related to disability compensation may be prevalent in the VA (cf. Frueh et al., 2005). In the current study, providers who reported feeling their patients were not getting better also reported increased cynicism, which may speak to psychiatrists' concerns about possible malingering and/or sense of failure and frustration when a patient does not respond to treatment (Kumar, 2007). More research on the complex relationship between treatment gains, suspected malingering and cynicism in VHA settings is needed.

We also found that working more frequently with bipolar sufferers was associated with less cynicism. The reasons for this finding are unknown and more research is needed to understand this association.

5.2.2. Exhaustion

Over 86% of this sample endorsed high exhaustion, with workplace variables appearing to play a significant role. A lack of sufficient resources to perform job duties was associated with greater exhaustion, as was the number of times per year that psychiatrists were on short call. It seems likely that hiring more frontline psychiatry staff could reduce such workload pressures, and possibly burnout. Given the costs of recruitment and lost productivity associated with turnover, the cost of hiring more staff may be offset by the savings associated with improved staff retention. A perceived lack of fair treatment from supervisors also predicted exhaustion, consistent with other research showing that negative relationships with superiors increases burnout (cf. Bressi et al., 2009; Kumar et al., 2011; Maslach et al., 2001).

Patient characteristics also impacted exhaustion, particularly for those providers treating a high percentage of patients with severe mental illness (SMI). Patients with SMI often require intensive treatment and prior research has also suggested treating difficult patients can lead to burnout (Bakker, Schaufeli, Sixma, Bosveld, & Van Dierendonck, 2000).

5.2.3. Professional efficacy

Despite exhibiting high levels of cynicism and exhaustion, VA psychiatrists also reported high levels of professional efficacy (74.3%). Work variables that impacted professional efficacy included time spent teaching and receiving fair treatment. Unexpectedly, more time spent teaching was associated with lower professional efficacy. While the reasons for this finding are unclear, it may be that teaching performance may become impaired if protected time is not granted for teaching (e.g., by limiting preparation). As in prior research (Maslach et al., 2001), providers who reported being treated fairly by supervisors reported greater professional efficacy.

The only patient variable that affected professional efficacy was whether providers felt patients were getting better; this finding was also consistent with prior research (Garcia et al., 2015).

5.2.4. Workplace factors and intent to leave

Finally, we examined whether workplace and patient variables predicted providers' intent to leave their jobs within two years. No

patient variables predicted intent to leave. However, both not feeling part of a coherent team and not having protected time outside of patient care were associated with intent to leave. The fact that not feeling part of coherent team was associated with both cynicism and intent to leave in regression models suggests it may have both direct and indirect (via cynicism) effects on staff retention. At least one prior meta-analysis has demonstrated the importance of social support on staff retention (c.f. Barak, Nissly, & Levin, 2001). The results highlight the importance of the social environment for ensuring staff well-being and retention over time.

The lack of protected time was also associated with both cynicism and the intent to leave. Although the VA has significantly increased the size of their mental health workforce in recent years, widely acknowledged staffing shortages (1OM, 2012) may inhibit providing protected time for non-clinical activities. However, turnover is, as noted above, associated with enormous financial costs involved in training, recruitment, and lost productivity. While financial costs are incurred when an employee is engaged in non-clinical (i.e., non-billable) activities, our study results suggest that supporting such activities may result in less cynicism and turnover among VA psychiatrists, which in the long term could produce substantial cost savings.

5.3. Limitations

Interpretations of the current findings should be qualified by several methodological limitations. Sampling bias may have occurred such that providers experiencing the most burnout were more likely to respond to our survey. Conversely, providers administering the largest panels, and thus at greater risk of burnout, may have been less likely to complete our survey due to time constraints. In addition, the study's cross-sectional design does not allow us to establish causality in relationships identified between variables. Finally, some of the questions used in this survey utilized single-item indices developed by the authors and had not been previously validated.

6. Conclusions

Overall, this study found that both workplace factors and patient characteristics impact burnout among VA psychiatrists and their intent to leave the organization. While patient characteristics are difficult to control, it appears that the VA could benefit by mitigating workplace factors associated with burnout and intent to leave by, for example: restoring a sense of administrative control to providers; enhancing team/relationship-building; ensuring fair treatment by supervisors; reducing frequency of short call; hiring additional staff; and allowing protected time for activities like research, teaching, and supervision. Such steps may not only help to improve the health of a large segment of the VA workforce, but also save taxpayers and the federal government substantial costs associated with staff attrition. Most importantly, it seems likely that reducing provider burnout has potential to improve quality of work life for VA psychiatrists, and improve the quality of the mental health care they provide.

Conflict of interest

The authors declare that there are no conflicts of interest References

AlbuAlRub, R. F., & Al-Zaru, 1. M. (2008). Job stress, recognition, job performance

and intention to stay at work among Jordanian hospital nurses. Journal of

Nursing Management, 16, 227-236. http://dx.doi.org/10.1111/j.1365-2834.

2007.00810.x

Bakker, A. B., Demerouti, E., & Schaufeli, W. B. (2002). Validation of the Maslach Burnout Inventory - Genera survey: An internet study. Anxiety, Stress, and Coping, 15(3), 245-260.

Bakker, A. B., Schaufeli, W. B., Sixma, H. J., Bosveld, W., & Van Dierendonck, D. (2000). Patient demands, lack of reciprocity, and burnout: A five-year longitudinal study among general practitioners. Journal of Organizational Behavior, 21,425-441.

Barak, M. E. M., Nissly, J. A., & Levin, A. (2001). Antecedents to retention and turnover among child welfare, social work, and other human service employees: What can we learn from the past research? A review and metaanalysis. Social Service Review, 625-661.

Ballenger-Browning, K. K., Schmitz, K. J., Rothacker, J. A., Hammer, P. S.,

Webb-Murphy, J. A., & Johnson, D. C. (2011). Predictors of burnout among military mental health providers. Military Medicine, 176, 253-260. http://dx. doi.org/10.7205/MILMED-D-10-00269

Berger, J. E., & Boyle, R. L., Jr. (1992). How to avoid the high costs of physician turnover. Medical Group Management Journal, 39, 80-84, 86.

Bressi, C., Porcellana, M., Gambini, O., Madia, L., Muffatti, R., Peirone, A., et al. (2009). Burnout among psychiatrists in Milan: A multicenter survey. Psychiatric Services, 60,985-988.

Burke, R. J., Greenglass, E. R., & Schwarzer, R. (1996). Predicting teacher burnout overtime: Effects of work stress, social support, and self-doubts on burnout and its consequences. Anxiety, Stress & Coping: An International Journal, 9(3), 261-275.

Cherniss, C. (1980). Professional burnout in the human service organizations. New York: Praeger.

Deary, I. J., Agius, R. M., & Sadler, A. (1996). Personality and stress in consultant psychiatrists. International Journal of Social Psychiatry, 42,112-123.

Firth, H., & Britton, P. (1989). Burnout, absence and turnover amongst British nursing staff. Journal of Occupational and Organizational Psychology, 62(1), 55-59.

Foa, E. B., Hembree, E. A., & Rothbaum, B. O. (2007). Prolonged exposure therapy for PTSD: Emotional processing of traumatic experiences: Therapist guide. Oxford: Oxford University Press.

Frueh, B. C., Elhai, J. D., Grubaugh, A. L., Monnier, J., Kashdan, T. B., Sauvageot, J. A., et al. (2005). Documented combat exposure of US veterans seeking treatment for combat-related post-traumatic stress disorder. The British Journal of Psychiatry, 186, 467-472.

Garcia, H. A., McGeary, C., McGeary, D. D., Finley, E. P., & Peterson, A. (2014). Burnout in Veterans Health Administration providers in posttraumatic stress clinics. Psychological Services, 11(1), 50-59.

Garcia, H. A., McGeary, C., McGeary, D. D., Finley, E. P., Ketchum, N., McGeary, D. D., et al. (2015). The influence of trauma and patient characteristics on provider burnout in VA PTSD specialty programs. Psychology and Psychotherapy: Theory Research and Practice, http://dx.doi.org/10.1111/papt.12057

Gray, A. M., Phillips, V. L., & Normand, C. (1996). The costs of nursing turnover: Evidence from the British National Health Service. Health Policy, 38,117-128. http://dx.doi.org/10.1016/0168-8510(96)00854-8

Heponiemi, T., Aalto, A. M., Puttonen, S., Vanska, J., & Elovainio, M. (2014).

Work-related stress, job resources, and well-being among psychiatrists and other medical specialists in Finland. Psychiatric Services, 65(6), 796-801.

Holmqvist, R., & Jeanneau, M. (2006). Burnout and psychiatric staffs feelings

towards patients. Psychiatry Research, 145, 207-213. http://dx.doi.org/10.1016/ j.psychres.2004.08.012

IOM (Institute of Medicine). (2012). Treatment for posttraumatic stress disorderin military and veteran populations: Initial assessment. Washington, DC: The National Academies Press.

Kumar, S. (2007). Burnout it psychiatrists. World Psychiatry, 6,186-189.

Kumar, S., Hatcher, S., Dutu, G., Fischer, J., & Ma'u, G. (2011). Stresses experienced by psychiatrists and their role in burnout: A national followup study. The International Journal of Social Psychiatry, 57,166.

Leiter, M. P., & Maslach, C. (2009). Nurse turnover: The mediating role of burnout. Journal of Nursing Management, 17,331-339.

Leiter, M. P., Harvie, P., & Frizzell, C. (1998). The correspondence of patient satisfaction and nurse burnout. Social Science & Medicine, 47,1611-1617. http://dx.doi.org/10.1016/S0277-9536(98)00207-X

Maslach, C. (1982). Understanding burnout: Definitional issues in analyzing a complex phenomenon. In W. S. Paine (Ed.), Job stress and burnout (pp. 29-40). Beverly Hills, CA: Sage.

Maslach, C. (1988). A multidimensional theory of burnout. In C. L. Cooper (Ed.), Theories of organizational stress (pp. 68-85). Oxford, England: Oxford University Press.

Maslach, C., & Goldberg, J. (1998). Prevention of burnout: New perspectives. Applied and Preventive Psychology, 7,63-74. http://dx.doi.org/10.1016/S0962-1849(98)80022-X

Maslach, C., & Jackson, S. E. (1981). The measurement of experienced burnout. Journal of Occupational Behavior, 2, 15.

Maslach, C., Jackson, S. E., & Leiter, M. P. (1996). Burnout inventory manual (3rd ed.). Palo Alto: Consulting Psychologist Press.

Maslach, C., & Leiter, M. P. (1997). The truth about burnout: How organizations cause personal stress and what to do about it. San Francisco, CA: Jossey-Bass.

Maslach, C., & Leiter, M. P. (2008). Early predictors ofjob burnout and engagement. Journal of Applied Psychology, 93, 498-512. http://dx.doi.org/10.1037/0021-9010.93.3.498

Maslach, C., Schaufeli, W. B., & Leiter, M. P. (2001). Job burnout. Annual Review of Psychology, 52(1), 397-423.

Misra-Hebert, A. D., Kay, R., & Stoller, J. K. (2004). A review of physician turnover: Rates, causes, and consequences. American Journal of Medical Quality, 19, 56.

Mott, D. A. (2000). Pharmacist job turnover, length of service, and reasons for leaving, 1983-1997. American Journal of Health-System Pharmacy, 57,975-984. Retrieved from http://www.ajhp.org/

Pines, A., & Maslach, C. (1978). Characteristics of staff burnout in mental health settings. Hospital Community Psychiatry, 29(4), 233-237.

Potik, D., Feldinger, E., & Schreiber, S. (2012). Issues in psychotherapeutic encounters of patients suspected of post-traumatic stress disorder malingering. European Journal of Psychotherapy & Counselling, 14(3), 261-277. http://dx.doi.org/10.1080/13642537.2012.713185

Rathod, S., Roy, L., Ramsay, M., Das, M., Birtwistle, J., & Kingdon, J. (2000). A survey of stress in psychiatrists working in the Wessex Region. Psychiatric Bulletin, 24, 133-136.

Schaufeli, W. B., Leiter, M. P., Maslach, C., & Jackson, S. E. (1996). MBI-general survey. In C. Maslach, S. E. Jackson, & M. P. Leiter (Eds.), Maslach burnout inventory. Palo Alto, CA: Consulting Psychologists Press.

Snibbe, J. R., Radcliffe, T., Weisberger, C., Richards, M., & Kelly, J. (1989). Burnout among primary care physicians and mental health professionals in a managed health care setting. Psychological Reports, 65, 775-780.

Stoller, J. K., Orens, D. K., & Kester, L. (2001). The impact of turnover among respiratory care practitioners in a health care system: Frequency and associated costs. Respiratory Care, 46, 238-242.

Taylor, S., Freuh, B. C., & Asmundson, J. G. (2007). Detection and management of malingering in people presenting for treatment of posttraumatic stress disorder: Methods, obstacles, and recommendations. Journal of Anxiety Disorders, 21, 22-41. Thomsen, S., Soares, J., Nolan, P., Dallender, J., & Arnetz, B. (1999). Feelings of professional fulfillment and exhaustion in mental health personnel: The importance of organizational and individual factors. Psychotherapy and Psychosomatics, 68,157-164. Veterans Health Administration. (2014). Where do ¡get the care 1 need? Retrieved

from http://www.va.gov/health/findcare.asp Voss Horrell, S. C., Holohan, D. R., Didion, L. M., & Vance, G. T. (2011). Treating traumatized OEF/OIF Veterans: How does trauma treatment affect the clinician? Professional Psychology, Research and Practice, 42, 79-86. http://dx. doi.org/10.1037/a0022297 Waldman, J. D., Kelly, F., Arora, S., & Smith, H. L. (2004). The shocking cost of

turnover in health care. Health Care Management Review, 29, 2-7. http://dx.doi. org/10.1097/00004010-200401000-00002. Retrieved from http://journals. lww.com/hcmrjournal/pages/default.aspx Zapf, D., Seifert, C., Schmutte, B., Mertini, H., & Holz, M. (2001). Emotion work and job stressors and their effects on burnout. Psychology^ Health, 16,527-545. http://dx.doi.org/10.1080/08870440108405525