APPLICATION OF THE JOB DEMANDS-RESOURCES MODEL TO INVESTIGATE
TURNOVER AND QUALITY OF CARE IN CERTIFIED NURSE-MIDWIVES AND
CERTIFIED MIDWIVES
by
E. BRIE THUMM
B.A., Georgetown University, 1995
M.S.N., Yale University, 2001
M.B.A., Baruch College, 2007
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
College of Nursing
ii This thesis for the Doctor of Philosophy degree by
E. Brie Thumm
has been approved by the
College of Nursing
by
Paula Meek, Chair
Linda Flynn, Advisor
Jonathan Shaffer
Nancy Lowe
Ginger Breedlove
iii Thumm, E. Brie (PhD, College of Nursing)
Application of the Job Demands-Resources Model to Investigate Turnover and Quality of
Care in Certified Nurse-Midwives and Certified Midwives
Dissertation directed by Professor Linda Flynn
ABSTRACT
Aims: U.S. maternity care is riddled with workforce instability and outcomes inconsistent with spending. Certified nurse-midwives and certified midwives are a potential
high-quality, cost-effective solution; however, the current midwifery workforce is
inadequate. Practice climate has been demonstrated to affect quality of care and workforce
stability. The purpose of this study is to investigate the midwifery practice climate by (a)
developing a valid and reliable scale to measure midwifery practice climate and (b) test the
effect of midwifery practice climate in the Job Demands-Resources model, including
mediating relationships of provider burnout and work engagement.
Methods: A cross-sectional survey of U.S. certified nurse-midwives and midwives (n=2,333) was conducted and divided to conduct three distinct phases of testing: establishing
the structure of the midwifery practice climate scale with exploratory factor analysis (n=330);
refinement of the scale structure with structural equation modeling (n=330); and testing the
Job Demands-Resources Model with multiple linear regression analyses (n=1673).
Results: Initial analysis revealed a five-factor structure reflective of the hypothesized structure, accounting for 55.86% of the variance in the items. The structure identified in initial testing was found to have a marginal fit (2=
1019.562, df=411, p<0.001, RMSEA=0.067, CFI=0.903). The revised scale, based upon a hybrid of theoretical and statistical item
iv Midwifery Model of Care), demonstrating less measurement error and more parsimony (2=60.397, df=34, p<0.001, RMSEA= 0.049, CFI= 0.987). There was a significant
relationship (p<0.001) between both dimensions of practice climate (burnout and
engagement) and midwife professional well-being. Burnout mediated relationships between
two dimensions of practice climate and turnover intention (p<0.001). Work engagement
mediated relationships between two dimensions of practice climate and midwife-perceived
quality of care (p<0.001).
Conclusion: Midwifery practice climate can accurately be assessed with 10 items addressing two dimensions of practice: Practice Leadership and Participation and Midwifery
Model of Care. Professional well-being of midwives is related to the practice climate in
which they practice. Workforce and patient outcomes are related to practice climate and
partially mediated by professional well-being of midwives, thereby supporting the health
impairment pathway and the motivational pathway of the Job Demands-Resources model.
The form and content of this abstract are approved. I recommend its publication.
v
DEDICATION
To Dan
Getting my Ph.D. was not my greatest accomplishment in the last four years,
finding you was.
To Missa
Since you never want to do your Ph.D., you can just share this one with me. It’s only fair considering how many hours you spent proof reading, consulting, cheerleading, and listening
to me vent. Do you want the “h” or the “D”?
To my Parents
Thank you for teaching me to value education
and never asking, “Why are you getting yet another degree?”
To all the Midwives
We have been caring for women and families since the beginning of time,
vi
ACKNOWLEDGEMENTS
I want to thank all of the members of my committee who each made a unique and
essential contribution to this work. Ginger Breedlove, thank you for your commitment to this
topic and leveraging your networks to boost awareness and response rates. Nancy Lowe,
thank you for helping me reflect on my methods and midwifery as a profession. Jonathan,
thank you for bringing in an outside perspective to the data. Thank you, Paula Meek, for
having the wisdom in measurement and belief in me to push me to identify a final scale that
honors my investment in this project and the investment of the midwives who participated in
the study. Finally, thank you to my advisor Linda Flynn for your practical approach to
science, keeping me focused on the long term, and asking me “so what?” countless times over the last four years. Linda and Paula, you are both amazing mentors. At every juncture,
you made me feel empowered, supported, and competent.
To Allison Squires, you saw the PhD in me before I did. Thank you for over a decade
of mentorship and friendship. Your commitment to raising others up through your generosity
of knowledge and connections brings all of nursing to a higher level.
I would like to acknowledge the American Midwifery Certification Board, especially Lori Havens, for prioritizing midwives’ professional wellbeing and sending out my survey. I would like to express my gratitude to the staff and administration at the University of
Colorado College of Nursing for all of their behind-the-scenes work, including keeping me
calm every time I freaked out.
Finally, this work would not be possible without the thousands of midwives who
vii
TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION……….1
Statement of the Problem...1
Aims and Hypotheses...2
Aim 1……….2
Aim 2……….3
Aim 3……….3
Aim 4……….4
Definition of Theoretical and Operational Terms...4
Turnover Intention…...4 Quality of Care …...5 Work Engagement …...5 Professional Burnout ...5 Practice Climate……….6 Significance...6
II. REVIEW OF THE THEORETICAL FRAMEWORK AND EMPIRICAL LITERATURE…...………10
Turnover Intention Theories of Turnover Intention………10 Empirical Literature: Organizational Attributes and Midwifery Turnover….14
viii
Quality of Care/Organizational Performance…………..………18
Theories of Organizational Performance and Quality of Care ………...18
Empirical Literature: Organizational Attributes and Nursing and Midwifery Quality of Care……….20
Standard Patient Outcomes as a Measure of Quality of Care………..20
Nurse-Reported Quality of Care as a Measure of Quality of Care…..20
Professional Burnout………22
Theoretical Literature of Professional Burnout………...22
Empirical Literature: Organizational Attributes and Midwifery Burnout…...23
Work Engagement………...30
Theoretical Literature of Work Engagement………...30
Empirical Literature: Organizational Attributes and Work Engagement of Nurses, Physicians, and Midwives………...32
Practice Climate………...34
Theories of Professional Practice Climate………...34
Empirical Literature: Supportive Nurse, Nurse Practitioner and Midwifery Practice Climates and Workforce Outcomes………...37
Conclusion……….…..41
Gap in the Literature………41
Job Demands-Resources Model………...42
Aims and Hypotheses………..45
Aim 1………...45
ix Aim 3………...45 Aim 4………...46 III. METHODS...47 Study Design...47 Sample Population...47 Inclusion Criteria……….49 Exclusion Criteria………49 Sample Size……….49 Limitations………...49 Procedures………50 Data Collection………50 Recruitment………….….………51
Protection of Human Subjects……….52
Risks and Benefits………52
Informed Consent……….52
Instruments...53
Practice Structural Characteristics………...53
Midwifery Practice Climate Scale………...54
Maslach Burnout Inventory……….56
Utrecht Work Engagement Scale-9 (UWES-9)………...56
Quality-of-Care Survey………57
x
Personal Demographic Information……….58
Data Analysis Plan ………..58
Describing the Sample and Normality Testing...59
Psychometric Analysis of the Midwifery Practice Climate Scale (Aim 1)...60
Testing Measurement Models of MBI, UWES, and Job Outcomes Scales (Aim 2)...62
Testing the Job Demands-Resources Model (Aims 3 & 4)...62
IV. RESULTS...64
Pilot Study...64
Data Cleaning and Missing Data...65
Participants ………..67
Assumptions of Normality of Scale Items …...68
Splitting of Sample with Modified Random Selection…...69
Aim 1: Psychometric Properties of the MPCS ...70
Initial Exploratory Factor Analysis of the MPCS…...70
Confirmatory Factor Analysis and Item Reduction Results……...72
Further Exploration of Items and Revision of Structure………..75
Revised CFA: Model 3………79
Reliability of the MPCS………...80
Aim 2: Measurement Models of the Additional Variables ...82
Aim 3: Testing the Health Impairment Process of the JD-R ...83
Practice Leadership and Participation...83
xi
Aim 4: Testing the Motivational Process of the JD-R...86
Practice Leadership and Participation...86
Midwifery Model of Care...87
Summary ...88
V. DISCUSSION AND CONCLUSIONS...89
Discussion of Study Findings...89
Sample...89
Aim 1………...90
Hypothesis 1…...90
Hypothesis 2...90
Hypothesis 3...91
Comparison of the Revised MPCS to Other Midwifery Practice Climate Scales………...93 Aim 2 …...94 Aim 3………...95 Aim 4………...97 Summary of Findings...98 Study Limitations...98
Policy and Practice Implications...100
Future Research…...102
Conclusion…...103
REFERENCES...104
xii
A. Pilot Survey Questions...117
B. Survey Cover Letter………..118
C. Recruitment Materials and Results Figure C1: Pre-Notification Postcard……….120
Table C1: Recruitment Response………...12
D. Survey Items Table D1: Practice Demographics………...122
Table D2: Midwifery Practice Climate Scale………124
Table D3: Maslach Burnout Inventory………..126
Table D4: Utrecht Work Engagement Scale-9………..128
Table D5: Quality of Care……….128
Table D6: Turnover Intention………129
Table D7: Personal Demographics………129
E. Measurement Models Figure E1: Midwifery Practice Climate Scale Structural Model ...131
Figure E2: Professional Burnout Measurement Model………...132
Figure E3: Work Engagement Measurement Model……….133
Figure E4: Turnover Intention Measurement Model……….134
F. Sample Demographic Results Table F1: Personal Demographics of Three Subsamples………..135
Table F2: Practice Demographics of Three Subsamples………...138
G. Distribution of MPCS Items Table G1: Distribution of MPCS Items in Full Sample……….144
xiii
H. Exploratory Factor Analysis……….147
I. Models Tested in Confirmatory Factor Analysis Figure I1: Model I………..148
Figure I2: Model II……….149
Figure I3. Model Ia………150
Table I1: Items Covaried in Model Ia……….………...151
Figure I4: Final 10-Item 2-Factor Structure of the MPCS……….155
J. Item Dispersion of Maslach Burnout Inventory Table J: Maslach Burnout Inventory Items………156
xiv
LIST OF TABLES
TABLE
1.Fit indices of three full MPCS models……….75
2. Distribution of MPCS items and rationale for removal………...77
3. Final items of Midwifery Practice Climate Scale………80
4. Reliability of revised MPCS………81
5. Health Impairment Process: relationship between Practice Leadership and Participation and Turnover Intention mediated by Emotional Exhaustion………...84
6. Health Impairment Process: relationship between Midwifery Model of Care and Turnover Intention mediated by Emotional Exhaustion………..85
7. Motivational Process: relationship between Practice Leadership and Participation and Midwife-Perceived Quality of Care mediated by Work Engagement……….87
8. Motivational Process: relationship between Midwifery Model of Care and Midwife-Perceived Quality of Care mediated by Work Engagement………88
xv
LIST OF FIGURES
FIGURE
1. The Health Impairment Process of the JD-R Model………..3
2. The Motivational Process of the JD-R Model………...4
3. The Job Demands-Resources Model………...14
4. The Job Demands-Resources Model with study variables………..44
5. The Health Impairment Process of the JD-R Model………46
6. The Motivational Process of the JD-R Model……….46
7. Diagram of mediation equation………...63
1
CHAPTER I INTRODUCTION Statement of the Problem
Childbirth is the most prevalent reason for hospitalization in the U.S. (Podulka,
2011). Despite this, U.S. maternity care is riddled with workforce shortages (Bushman, 2015)
and patient outcomes inconsistent with spending (Renfrew et al., 2014). The shortage of
maternity providers in the U.S. is so severe that, in 2011, over 40% of counties in the U.S.
had no obstetrical physicians, certified nurse midwives (CNMs), or certified midwifes (CMs)
(Rayburn, 2011; Birth by the Numbers, 2014). Moreover, the rapid growth in the number of women of childbearing age is exacerbating our nation’s shortage of maternity care providers. Over the next 30 years, the number of births is predicted to increase by 14% a year (U.S.
Census Bureau, 2014). Yet over the same time period, the supply of
obstetricians-gynecologists in the U.S. is projected to decline by over 20,000 (Rayburn, 2011). This
shortage comes at a time when the international health community is demonstrating that not
only do midwives provide more cost-effective care but also provide care associated with
better outcomes for the majority of pregnant women (NICE, 2014). Despite the value of
midwifery care, the numbers of CNMs/CMs in the U.S. are insufficient to fill the growing
gap between demand for maternity care and the supply of providers (Bushman, 2015).
Thus, our nation is faced with a critical shortage of maternity care providers that is
reaching crisis proportions. In recognition of the crisis, the U.S. Congress is reviewing the
Maternity Care Act of 2015, which would establish a maternity care shortage designation to
2 Increasing the U.S. supply of practicing CNMs/CMs and retaining current practicing
CNMs/CMs within the active maternity care workforce are imperative to alleviating this
crisis. Regarding professionals in general and nurses in particular, the theoretical and
empirical literatures indicate that key drivers of workforce stability and care quality are
practice climate, burnout, and work engagement (Cimiotti, Aiken, Sloane, & Wu, 2012; Van
Bogaert, Wouters, Willems, Mondelaers, & Clarke, 2013). The influence of these factors,
however, has not been investigated in the CNM/CM workforce. Research is needed to
explore the unique practice environments of CNMs/CMs in order to create evidence-based
practice environments that foster a stable workforce of CNMs/CMs providing high quality
midwifery care. The purpose of this study was to begin to fill this gap in the U.S. midwifery
literature.
Aims and Hypotheses Aim 1
In order to study the unique practice climate of CNMs/CMs, the Midwifery Practice Climate
Scale (MPCS) was developed and underwent initial testing. The first aim of the proposed
study is to further test psychometric properties of the MPCS.
H1: The MPCS will display adequate reliability in terms of internal consistency through Cronbach’s alpha coefficient measures and stability through intraclass correlations. H2: The structure of the MPCS will reflect the four-subscale structure of the nurse
practitioner scale upon which it is based plus a fifth midwifery-specific subscale, as
3 H3: The MPCS will demonstrate adequate construct validity, as determined by the
confirmatory factor analysis reflecting a structure similar to the structure determined to be
the best fit in the exploratory factor analysis.
Aim 2
The second aim is to evaluate the measurement models of the Utrecht Work Engagement
Scale, Maslach Burnout Inventory, and turnover intention scale through structural equation
modeling.
Aim 3
The third aim is to test the health impairment process of the Job Demands-Resources (JD-R)
model (see Figure 5).
H4: There will be an inverse relationship between practice climate and professional burnout.
H5: There will be a positive relationship between professional burnout and turnover
intention.
H6: Professional burnout will partially mediate the relationship between midwifery practice
climate and turnover intention.
Figure 1. The Health Impairment Process of the JD-R Model.
Burnout
Turnover
Intention
Practice
Climate
+
-4
Aim 4
The fourth aim is to the test the motivational process of the JD-R model (see Figure 6).
H7: There will be a positive relationship between practice climate and work engagement.
H8: There will be a positive relationship between work engagement and midwifery-reported
quality of care of one’s practice.
H9: Work engagement will partially mediate the relationship between midwifery practice
climate and midwifery-reported quality of care of one’s practice.
Figure 2. The Motivational Process of the JD-R Model.
Definition of Theoretical and Operational Terms Turnover Intention
The two primary types of turnover intention which will be investigated are
occupational and organizational. Occupational turnover intention is movement to a new occupation that differs significantly from one’s previous career (Lawrence, 1980; Rhodes & Doering, 1983). A related term, organizational turnover intention or voluntary turnover
intention, occurs when an individual willingly and voluntarily relinquishes his or her position
in exchange for another position considered more desirable (Vigoda-Gadot & Ben-Zion,
2004). While the general term “turnover intention” will be used to discuss the theoretical Engagement Quality of Care Practice Climate + + +
5 construct in the proposed model, the type of turnover intention will be specified in the
analysis and discussion.
Quality of Care
The Institute of Medicine defines quality of care as safe, effective, patient-centered,
timely, efficient, and equitable patient care (IOM, 2001). Researchers have determined that
nurse-reported quality of care is a valid proxy measure of quality of care (Aiken et al., 2012;
McHugh & Stimpfel, 2012). Freeney and Fellenz (2013a) measured midwife-reported quality
of care using an adaptation of the nurse-reported measure. The variable explored in this study
will be quality of patient care provided by the individual and/or practice as reported by the
individual midwife.
Work Engagement
Work engagement is the energetic and effective connection with one’s work. Kahn defined engagement as attending to task behaviors with one’s “preferred self,” which
encourages connection; cognitive, physical, and emotional presence; and comprehensive role
performance. Some organizational psychologists describe work engagement on a shared
continuum with burnout (Maslach, Schaufeli, & Leiter, 2001). The most empirically tested
definition, and therefore the definition used for this study, is a positive affective-cognitive state of mind related to one’s professional life with the three primary attributes of vigor, dedication, and absorption (Schaufeli, Martinez, Marques Pinto, Salanova, & Bakker, 2002).
Professional Burnout
Professional burnout was coined in the seventies by organizational psychologists as “a state of mental and physical exhaustion caused by one’s professional life” and “the extinction of motivation or incentive, especially where one’s devotion to a cause or
6 relationship fails to produce the desired result” (Freudenberger, 1980, pg. 160). Maslach defined burnout as a psychological condition in which an individual responds to chronic
professional stressors with pathologic levels of (emotional) exhaustion,
cynicism/depersonalization, and a sense of inefficacy/lack of accomplishment (Maslach et
al., 2001). This final definition was used in this study.
Practice Climate
Practice climate is the healthcare providers’ movable perception of their working conditions (Clarke, 2006). Climate research is grounded in psychology and the interaction
between individuals and the more superficial elements of their work environment, such as
policies and procedures (Denison, 1996). Climate differs from the similar concept of culture,
which refers to the underlying, and often unspoken, purpose and values of an organization
(Denison, 1996). Culture is considered more static and trait-like while climate is more fluid
and state-like. Climate is studied quantitatively, while culture is generally studied
qualitatively with a focus on evolution of systems. Climate scales measure the effect of
organizational systems on the individuals and groups that work within those systems
(Schneider, Ehrhart, & Macey, 2013). Characteristics of the practice climate include
teamwork, autonomy, leadership support, feedback, and social support.
Significance
While the American College of Obstetrics and Gynecology refers to the current maternity care provider “crisis,” the shortage of maternity providers in the U.S. is an opportunity for the U.S. healthcare system to invest in an evidence-based maternity care
system. Integral to this is midwifery care. Despite the opportunity for growth of midwifery
7 recommended for low-risk women, the CNM/CM workforce is under strain of attrition. The
midwifery workforce is aging (Hastings-Tolsma et al., 2015; Jevitt & Beckstead, 2004), and
there is an alarming decrease in the rate of CNMs/CMs recertifying since 2006. In 2014, 25%
of CNMs/CMs due to recertify neither recertified nor indicated retirement (AMCB, 2015).
Burnout and work engagement are key variables affecting the vitality of the workforce
through organizational and occupational turnover intention (Laschinger & Fida, 2014; Van
Bogaert, Kowalski, Weeks, & Clarke, 2013). Extending from established relationships in
related professions, this research examined the relationships among practice climate, burnout,
work engagement, turnover intention, and quality of care in the CNM/CM population, to
protect and grow the midwifery workforce.
Work engagement and professional burnout are two opposing concepts that
characterize the relationship one has with her or his work. Burnout and engagement affect
job performance (A. B. Bakker & Heuven, 2006; Halbesleben & Wheeler, 2008), personal
health (Ahola et al., 2008; Eguchi et al., 2015), and longevity in the workforce (Laschinger &
Fida, 2014; Van Bogaert, Wouters, et al., 2013). Specifically, burnout is associated with
adverse patient outcomes, including hospital acquired urinary tract infections (Cimiotti et al.,
2012), longer recovery times (Halbesleben & Rathert, 2008), and failure to rescue (Aiken,
Clarke, Sloane, Sochalski, & Silber, 2002). In contrast, engagement is associated with
improved work performance, such as patient-centered care behaviors (Abdelhadi &
Drach-Zahavy, 2012) and extra-role performance (Salanova, Lorente, Chambel, & Martinez, 2011).
Both burnout and engagement are associated with patient satisfaction (Vahey, Aiken, Sloane,
Clarke, & Vargas, 2004); and overall quality of care (Freeney & Fellenz, 2013a; Van
8 The antecedents to burnout and work engagement are primarily characteristics of the
practice environment, not personal factors (A. B. Bakker, Demerouti, & Sanz-Vergel, 2014).
This means that the engagement or burnout of individual midwives can be fostered or
mitigated by midwives, administrators, and policy makers with adequate data to drive
evidence-based practice environments. The practice environment is comprised of tangible
and intangible factors. The tangible factors include patient acuity, 24-hour service,
compensation, and practice model. The intangible factors are components of the practice
climate. Practice climate refers to the perceptions of providers of the practices and
procedures in their work setting (Clarke, 2006; Denison, 1996). Elements of practice climate
include concepts such as workload, decision latitude, and social capital (Van Bogaert,
Kowalski, et al., 2013).
The international midwifery communities in low-, middle-, and high-income
countries have recognized the significance of the psychological well-being of their midwifery
workforces. Their research has been driven not just by academic curiosity but also by gaps in
the midwifery workforce that demanded investigation into how to maintain a vital workforce
to provide safe maternity care (Curtis, Ball, & Kirkham, 2006; Rouleau, Fournier, Philibert,
Mbengue, & Dumont, 2012).
Despite these findings among our nurse, physician, and international midwifery
colleagues, there has been no research into the relationships among practice environment,
burnout and engagement, and outcomes with CNMs/CMs. The only study that investigated
burnout in CNMs was conducted in 1986 with a sample of 98 members of the American
College of Nurse-Midwives (ACNM; Beaver, Sharp, & Cotsonis, 1986). Midwives use
9 not use evidence to create the environments in which they provide care day-in and day-out
throughout their careers as midwives. Organizational psychologists and nurse researchers
have studied work climate extensively in settings from intensive care units (A. B. Bakker, Le
Blanc, & Schaufeli, 2005) to call centers (Ballard, 2012), and yet, despite knowing the importance of one’s environment, CNMs/CMs have not studied the effect of the practice environment on midwives and professional outcomes. CNMs and CMs are in need of
research that will guide the creation of supportive practice environments in order to provide
10
CHAPTER II
REVIEW OF THE THEORETICAL FRAMEWORK AND EMPIRICAL LITERATURE1
This chapter presents the theoretical literature related to turnover intention,
organizational performance/quality of care, burnout, work engagement, and the primary
independent correlate of practice climate. Following the review of the theoretical literature
related to each of these concepts, a review of midwifery/related professions-based empirical
literature that supports the theoretical propositions will be presented. Summaries will identify
gaps in the existing empirical literatures.
Turnover Intention Theories of Turnover Intention
Occupational and organizational turnover are complex processes in which an
individual willingly withdraws and leaves his or her place of work (Lawrence, 1980; Rhodes
& Doering, 1983; Vigoda-Gadot & Ben-Zion, 2004). Turnover has implications on health
system and organization levels, as well as individual midwives through workforce instability,
increased financial costs, and compromised system performance (Gilmartin, 2013). Leaving one’s career is a complex process involving personal, organizational, and occupational factors. Rhodes and Doering’s (1983) model of occupational turnover begins with organizational factors such as autonomy, compensation, colleagues, responsibilities, and
working conditions. The authors also specify perceptions of growth opportunity, person-work
1
A portion of the material reported in this chapter is reprinted with permission from the author’s previous work “The 5 attributes of a supportive midwifery practice climate: A review of the
literature” published in the Journal of Midwifery and Women’s Health, Wiley Publishing, Copyright 2017.
11 environment congruence, person-organization correspondence, and perceived performance as
drivers of occupational turnover. The lack of these organizational factors drives the
occupational turnover process. Building on this model, Blau (2007) suggested that work
exhaustion is a proxy measure of lack of person-work environment correspondence and
therefore predicts occupational turnover. Additionally, he synonymizes work exhaustion and
emotional exhaustion, a key component of burnout. While this theory is specific to
occupational turnover, the organizational factors are consistent with other, more generic
turnover theories described below.
Rhodes and Doering’s (1983) model posited that a withdrawal process occurs between initial dissatisfaction with organizational factors and occupational turnover. The
withdrawal process can lead directly to occupational turnover but can also loop back and be
the impetus for re-evaluating job opportunities within the person’s career. Gilmartin’s (2013) integrated turnover model proposed a similar process, specifying withdrawal cognitions and
withdrawal behaviors. Withdrawal behavior has significant implications for workforce
stability in that it leads to reduction in clinician productivity. This can occur on an
organizational level with an increase in absenteeism, tardiness, and leaves of absence
(Gilmartin, 2013; Mobley, Griffeth, Hand, & Meglino, 1979). Withdrawal can also occur occupationally through limiting one’s scope of practice, reducing hours, and developing retirement intentions (Hanisch & Hulin, 1990; 1991).
Three theories that explain turnover, occupational and voluntary, are the Nursing
Organization and Outcomes Model (Aiken, Clarke, & Sloane, 2002), the Job Characteristics
Model (Hackman & Oldman, 1980), and the Job Demands Resources Model (JD-R) (A. B.
12 organizational characteristics supporting professional nursing, specifically resource
adequacy, nurse autonomy, nurse control, and nurse-physician relations, determine nurse
outcomes including occupational turnover. The model was derived from sociological theories
suggesting that professionals in general, including but not limited to nurses, seek
employment/practice opportunities within organizations featuring the following
characteristics: inter-professional collaboration and respect, autonomy within scope of
practice or role, supportive managers, adequate resources, opportunity to influence policies
and organizational decisions, effective and horizontal communication channels, opportunities
for continuing education, and a culture of safety and/or quality improvement (Adler, Kwon,
& Heckscher, 2008; Flood & Fennell, 1995; Freidson, 1994). When these attributes are
absent or limited, job dissatisfaction and turnover intentions occur. Similarly, Hackman and Oldman’s (1980) Job Characteristics Theory, from the organizational psychology literature, posits that core job characteristics such as lack of skill variety, autonomy, and feedback are
negatively related to psychological well-being and lead to adverse personal and work
outcomes, specifically turnover.
The JD-R (A. B. Bakker & Demerouti, 2007) is a more universal model that has been
tested in many populations including nursing and midwifery. The JD-R posits that elements
of the practice environment are either demands or resources that drive the health impairment
pathway and motivation pathways, respectively. Demands are aspects of the work
environment that require physical and/or psychological effort and are therefore associated
with a physical and/or psychological price to the employee. Demands initiate the health
impairment process, leading to burnout. The theory also posits that the chronic condition of
13 well as negative organizational outcomes including decreased product/service quality. In
healthcare settings, these impaired organizational outcomes can be indicated by lower
estimates of quality patient care and compromised patient outcomes (Van Bogaert, Kowalski,
et al., 2013). The relationships among demands, burnout, and outcomes are also referred to as
the energetic process in that it requires personal energies to meet chronic demands (Jourdain
& Chênevert, 2010). Since the JD-R’s inception, researchers have further specified challenge demands and hindrance demands. Hindrances (e.g., role conflict) interfere with
accomplishing work goals. Challenges (e.g., responsibility and time pressure), while
requiring sustained effort, have the potential to foster personal growth (Van den Broeck, De
Cuyper, De Witte, & Vansteenkiste, 2010).
In contrast to demands, job resources are defined as aspects in the work environment
that foster personal growth and development and contribute to achieving work goals.
Resources lead to work engagement, which has a positive relationship with outcomes. The
authors of the JD-R model refer to this as the motivational process.
Personal characteristics are not part of the core JD-R model. Various theories of the
role of personal resources have been tested, but none has been adopted as part of the JD-R
due to inconsistent findings (Demerouti & Bakker, 2011; Xanthopoulou, Bakker, Demerouti,
& Schaufeli, 2007). The health impairment and motivational pathways (Figure 3) interact to
predict outcomes including occupational and voluntary turnover (A. B. Bakker & Demerouti,
2007).
In summary, consistent throughout these turnover theories are (a) distal placement of
organizational characteristics and (b) practice climate unsupportive of professional work,
14 theories of withdrawal and turnover have not been tested among U.S. CNMs/CMs. Given the
relatively small CNM/CM workforce, the shortage of maternity providers in the U.S., and the
evidence supporting expansion of midwifery care, it is important to understand the
relationship between practice climate and turnover intention. Due to the universality of the
JD-R Model and prior application to a variety of professions, it has promise in providing a
useful explanation of burnout, engagement, and outcomes including turnover intention within
the midwifery workforce. A full application of the JD-R Model to the profession of
midwifery is diagrammed in Figure 3.
Empirical Literature: Organizational Attributes and Midwifery Turnover
The research into occupational turnover in midwifery is limited. Of the five studies
cited, only one studied turnover and retention directly; the other studies addressed job
Figure 3. Adapted from “The Job Demands-Resources Model: State of the art,” by A.B.
15 satisfaction. Since the 2004 review, only a small handful of articles have attempted to fill this
gap in the literature, and none of those was conducted in the U.S.
The shortage of midwives in the U.K. has driven a series of studies conducted by the
National Midwifery Recruitment Retention and Return to Practice Project, as well as the
annual State of Maternity Services, published annually by the Royal College of Midwifery.
In a 2001 study, Ball, Curtis, and Kirkham (2002) found that the primary reasons British
midwives provided for leaving the profession were at the health system level, including
dissatisfaction with the organization of midwifery and the role of midwives. Additionally,
midwives who had left the profession reported feeling undervalued, a lack of power over
their work environment, increasing workload, ineffective support, and bullying (Ball et al.,
2002). In a 2006 mixed-methods follow-up study of 978 midwives who had left the
profession, the same authors identified lack of autonomy, flexibility, and support as key areas
of the practice climate that influenced the decision to leave the profession (Curtis, Ball &
Kirkham, 2006a; 2006b). They also found dissatisfaction among midwives regarding how
midwives were required to practice within their healthcare system. These points of
dissatisfaction included not being able to provide the type of care they wanted to provide
(39%), not being able to form meaningful relationships with patients (26%), and feeling
women were not receiving quality care (32%). Additionally, low staffing levels and
unsupportive midwifery managers contributed to their decision to leave the profession
(Curtis et al., 2006b).
An inherent weakness in occupational turnover research is that people who are
leaving a profession generally disengage or withdraw prior to actually leaving, and
16 those who actually leave are frequently unreachable through conventional mechanisms and
therefore not included in samples. The Ceasing to Practice study (Curtis et al., 2006a) was an
exception because the researchers interviewed midwives who had already left the profession.
Sullivan, Lock, and Homer (2011) used mixed methods and an adaptation of the
descriptive scale used by Curtis, Ball, and Kirkham (2002) to investigate the retention of
Australian midwives. In a sample of 209 from a single area health service, they found that
enjoying the work of midwifery (98%) and relationships with women (90%) were reasons
cited for continuing to practice midwifery. Additionally, the respondents reported
interactions with colleagues and a sense of belonging as influencing their continuing to
practice midwifery.
In a study of burnout among 475 Swedish midwives, 30.3% reported considering
leaving midwifery. This proportion is similar to the proportion that reported personal burnout (39.5%), which is consistent with Rhodes and Doering’s (1983) model of occupational turnover and the JD-R (A. B. Bakker & Demerouti, 2007). While there were no trends in
personal characteristics related to turnover intention, common environmental reasons
included lack of staffing and resources in stressful work environments (32.5%), conflicts
with colleagues and managers (24.9%), worries about the future, the content of work (10%),
salary (7.1%), and having experienced a critical patient-care situation (15.4%) (Hildingsson,
Westlund, & Wiklund, 2013).
In a large sample of hospital midwives (n=1190) from seven countries, researchers
explored the intention to leave the profession and current position from the perspective of job
17 intend to leave the profession. The sample was least satisfied with extrinsic rewards,
professional opportunity, and balance between family and work (Jarosova et al., 2016).
Pallant, Dixon, Sidebotham, and Fenwick (2016) investigated intention to leave the
profession of midwifery as a test of construct validity of the Practice Environment Scale:
Midwives (PES:Midwives). The authors measured intent to leave midwifery with a single
dichotomous item asking whether the respondent had considered leaving the profession
within the last six months. While their sample of 600 midwives from New Zealand was
almost equally divided among those that intended to leave the profession and those that did
not, 51 percent and 49 percent, respectively, the authors did find a significant inverse
relationship between four aspects of supportive practice climate measured by the
PES:Midwives and occupational turnover intention. The four aspects were quality of
management, midwife-doctor relations, resource adequacy, and opportunity for development.
Poor quality of management had the strongest influence on intention to leave midwifery (OR
2.6, 95% CI=1.7–3.9, p<.001). This is consistent with theories positing that multiple elements of a supportive practice climate have an inverse effect on turnover intention;
however, the authors did not test for mediation by burnout and engagement as proposed in
the JD-R model.
In summary, the empirical literature regarding correlates of occupational turnover and
turnover intention among midwives is limited. The few studies that do exist support the
theoretical propositions asserting (a) an inverse relationship between adequacy of resources
and a supportive practice environment and turnover and (b) a positive relationship between
18 Bakker & Demerouti, 2007). However, there is no research that specifically examined these
hypothesized relationships among CNMs/CMs in the U.S.
Quality of Care/Organizational Performance Theories of Organizational Performance and Quality of Care
Understanding what improves organizational performance has been fundamental to
the field of organizational psychology since its inception during the Industrial Revolution
(Bowditch & Buono, 2005). In the healthcare field, the most potent measure of
organizational performance is the quality of care provided. Quality of care, therefore, will be
the dependent variable measuring the organizational outcome of organizational performance.
In the theoretical models, the term “job performance” refers to the individual-level construct of organizational performance.
The Institute of Medicine defines quality of care as safe, effective, patient-centered,
timely, efficient, and equitable patient care (IOM, 2001). In recent years, researchers,
managers, economists, and policy makers have measured quality of care by standard patient
outcome indicators, such as central line-associated bloodstream infections, hospital-acquired
urinary tract infections, and patient satisfaction. Although there are not any external
standardized patient outcome indicators for midwives in the U.S., the ACNM benchmarking
project specifies clinical outcomes such as prematurity rates, episiotomy rate, and cesarean
rate (ACNM, 2016).
Alternatively, quality of care can be measured as reported by the provider. This
means of measuring quality of care was demonstrated to be valid in a study (n=16,241)
comparing nurse-reported outcomes to 30-day inpatient mortality and failure to rescue in 396
19 increase in nurses reporting care on their unit as “excellent” was associated with 5% lower odds of mortality (OR=0.95, 95% CI=0.92-0.98) and failure to rescue (OR=0.95, 95%
CI=0.92-0.98) for surgical patients. Aiken et al. (2012) had similar findings in a cross-national study of over 60,000 nurses. According to Rhodes and Doering’s model (1983) of occupational turnover, perception of performance is a key driver of occupational turnover.
This further supports the use of nurse- or midwife-perceived quality of care as a predictor of
occupational turnover, as well as an indicator of actual quality of care.
The theories that posit an inverse relationship between a supportive practice climate
and occupational turnover also posit a positive relationship between positive practice climate
and better quality of care. The Job Characteristics Theory proposes a positive relationship
between (a) the psychological states resulting from skill variety, task identity, task
significance, autonomy, and feedback and (b) superior work performance (Hackman &
Oldham, 1980). According to the Nursing Organization and Outcomes Model, organizational
characteristics consistent with a supportive practice environment are positively related to a
higher level of quality of care and better patient outcomes. The model likewise suggests that
positive nurse outcomes, such as job satisfaction and retention, promote higher levels of
quality nursing care, resulting in better patient outcomes. Conversely, according to the
model, poor nurse outcomes such as turnover would adversely affect the quality of the care
provided and thus negatively affect patient outcomes. Additionally, qualifications of staff and
the process of care contribute to quality of care. Work environment remains distal to nursing
care processes, the quality of care delivered, and patient outcomes (Aiken et al., 2002).
Moreover, the JD-R proposes that the resources in the work environment, such as
20 posit that work engagement is one operant mechanism that mediates the positive relationship
between resources and organizational performance (A. B. Bakker & Demerouti, 2007).
In summary, a supportive work environment has a positive relationship with quality
of care. The positive relationship between supportive practice climate and quality of care can
be mediated by the positive psychological state of job satisfaction and work engagement.
Additionally, care providers can accurately assess quality of care.
Empirical Literature: Organizational Attributes and Nursing and Midwifery Quality of Care
There exist a limited number of empirical studies investigating the relationships
between organizational correlates such as organizational attributes or practice climate and the
quality of care provided by midwives; therefore, both the nursing literature and the scant
midwifery literature will be referenced.
Standard patient outcomes as a measure of quality of care. In large and diverse samples, nurse researchers have found a correlation between attributes of the practice
environment and patient outcomes indicators. Friese, Lake, Aiken, Silber, and Sochalski
(2008) found that a poor and unsupportive nurse practice environment was associated with
increased odds of 30-day mortality (OR=1.37; 95% CI=1.07-1.76) and failure to rescue
(OR=1.48; 95% CI=1.07-2.03) among surgical oncology patients when controlling for
hospital and patient characteristics. A major gap in the empirical literature is that there are no
studies of midwives examining the relationships among work environment, midwifery
professional psychological well-being, and standard patient outcomes measures.
Nurse-reported quality of care as a measure of quality of care. There have been many large studies corroborating the theory that a more supportive nurse practice
21 environment is associated with higher levels of nurse-reported quality of care. In a study of
33,659 nurses and 11,318 patients in Europe and 27,509 nurses and more than 120,000
patients in the U.S., Aiken et al. (2012) found that nurses who reported better work
environments (i.e., work environments characterized by more supportive attributes) were half
as likely to report poor or fair patient care. In a study of 1,201 acute care nurses from six
Belgian hospitals, Van Bogaert, Kowalski, Weeks, van Heusden, and Clarke (2013) found
that in a model explaining 47% of the variance of nurse-reported quality of care, (a) the
supportive organizational characteristics of more autonomy as indicated by decision latitude,
(b) a supportive unit-level nurse manager, and (c) higher levels of social capital were
associated with higher levels of nurse-reported quality of care (CFI=0.904, IFI=0.904,
RMSEA=0.43). Additionally, decision latitude mediated the relationship between
nurse-physician relations and nurse-reported quality of care.
Freeney and Fellenz (2013a) conducted a study of midwife-reported quality of care
with a sample of 158 Irish midwives from two hospitals using an instrument similar to the
instrument tested by McHugh and Stimpfel (2005) and Aiken el al. (2012). The authors
found that work engagement was positively associated with quality of care. They also found
that supervisor support and organizational support were positively related to quality of care with partial mediation by engagement (χ2[67]=113, p<0.01, CFI=0.961, RMSEA=0.06) .
In a qualitative study, Carolan-Olah, Kruger and Garvey-Graham (2015) examined
Australian midwives’ perceptions of practice environment factors that led to normal birth, which is widely considered a positive patient outcome of midwifery care (ACNM, 2012).
22 midwife, emotional support, and promotion of skill development and safety promoted the
positive patient outcome of normal birth.
These findings indicate that organizational attributes (e.g., support) are correlated
with quality of midwifery care. In order better to understand this relationship and translate it
into patient care in the U.S., similar studies need to be conducted with a sample of
CNMs/CMs practicing in the U.S.
Professional Burnout Theoretical Literature of Professional Burnout
Professional burnout is defined as a psychological condition in which an individual
responds to chronic professional stressors with pathologic levels of emotional exhaustion,
cynicism/depersonalization, and a sense of inefficacy/lack of accomplishment (Maslach et
al., 2001). Kristensen, Borritz, Villadsen, and Christensen (2005) expanded on this definition,
adding that symptoms of exhaustion can be broken down into three categories: general
symptoms; symptoms related to work; and symptoms related to working with clients. In the
empirical literature, burnout is evaluated as a single construct, as the three symptoms, or as
the three areas of burnout.
Maslach, Schaufeli, and Leiter (2001) proposed that burnout is the result of chronic
job-person incongruence. They delineated six areas of work-life mismatch: workload,
control, reward, community, fairness, and values. Moreover, this focus on incongruence is reflective of Rhodes and Doering’s model of occupational change.
Within the context of the Nursing Organization and Outcomes Model, burnout is
23 as a mediator between an unsupportive practice environment and negative patient outcomes
(Aiken et al., 2002).
Burnout is a specified construct in the JD-R. The health impairment pathway of the
JD-R posits that when demands (e.g., hindrances) are high and resources are low, there are
higher rates of burnout. Additionally, the theory posits that burnout can have a mediating role
between unsupportive practice climate and negative outcomes (A. B. Bakker & Demerouti,
2007; Demerouti & Bakker, 2011). All of these theories propose that unsupportive workplace
attributes are the primary cause of burnout.
Empirical Literature: Organizational Attributes and Midwifery Burnout
Only one 1982 study has investigated burnout among CNMs/CMs in the U.S. Beaver,
Sharp, and Cotsonis (1986) conducted a cross-sectional survey of a convenience sample of
98 members of the ACNM. The independent variables were personal and environmental
characteristics. The environmental characteristics were predominantly structural, such as
number of full-time-equivalents and delivery location. Two exceptions were practice climate
items about sources of support and sources of problems. The authors measured burnout using
a version of the Maslach Burnout Inventory (MBI; Maslach & Jackson, 1981) that included
frequency and intensity of emotional exhaustion, depersonalization/cynicism, and personal
accomplishment; personal accomplishment is reverse-scored to indicate perceived lack of
personal accomplishment. Given that only the frequency scores were reported in the other
literature cited in this review and frequency is the standard of MBI scoring, only the
frequency scores are reported numerically below.
In Beaver and colleague’s study of CNMs, those practicing in diverse (rural and urban) communities had lower rates of burnout than those practicing in exclusively rural
24 settings. Full-scope practice was associated with higher average rates of personal
accomplishment in comparison to midwives providing exclusively antepartum care (40.7
versus 31.1, α=.05). Further, the estimated percentage of patients “on welfare” was
significantly inversely correlated with all three dimensions of burnout (emotional exhaustion
β=0.28, p=0.003; depersonalization β=0.27, p=0.005; personal accomplishment β= -0.32, p=0.001). The inverse was true of percentage of patients who had participated in childbirth classes with all three dimensions of burnout positively correlated with childbirth preparation
(emotional exhaustion r= -0.25, p=0.006; depersonalization r= -0.24, p=0.01; personal
accomplishment r=0.25, p=0.007). The number of deliveries monthly, both by the individual
midwife (r=0.20, p=0.03) and the practice (r=0.29, p=0.005), and hours worked per week
were slightly positively correlated with emotional exhaustion (r=0.19, p=0.04). Higher
salaries were positively associated with personal accomplishment (r=0.26, p=0.01). From
these findings, the authors made assumptions about elements of practice climate, such as
assuming that higher number of deliveries was associated with working in tertiary care
centers and consequently having less peer support and more interaction with interdisciplinary
colleagues (Beaver et al., 1986).
In the same 1986 study by Beaver et al., the two practice climate items measured
revealed that physician support was inversely related to emotional exhaustion intensity and
that nurse support was inversely related to lack of personal accomplishment intensity. These
findings emphasize the role of interdisciplinary teamwork in burnout prevention and are
consistent with findings in the nursing literature showing an inverse correlation between
25 Beaver and colleague’s study is limited by a small sample size and by having
investigated only two elements of practice climate. Additionally, the scales used to measure
support of practice and sources of problems were not psychometrically tested. Finally, the
healthcare system has changed dramatically since the data were collected in 1982. Emphasis
on quality-of-care indicators and cost containment reflected in the Triple Aim (IHI, 2015)
and implementation of the Affordable Care Act have changed the practice climate and its
effect on midwifery burnout.
Because burnout is related to the environment in which midwives practice and
because practice environment is related to the greater healthcare system, the remainder of the
midwifery burnout literature will be limited to high-income countries that more closely
reflect the U.S. Midwifery burnout literature in high-income countries is dominated by
Australia and the U.K. In addition to being the most prolific in midwifery burnout research,
midwifery in these two countries most closely resembles the role of CNMs/CMs in the U.S.
healthcare system.
Mollart, Skinner, Newing, and Foureur (2013) conducted a study among Australian midwives similar to Beaver, Sharp, and Cotsonis’ (1986) study. The independent variables included individual and work environment characteristics but did not include any measures
of practice climate. Overall, the researchers found moderate levels of emotional exhaustion,
lack of personal accomplishment, and low levels of depersonalization. More specifically, the
authors found a significant relationships between low personal accomplishment and years in
the profession (F(5,49)=2.7, p=0.031), the percentage of patients with multiple psychosocial
and emotional exhaustion (F(4,51)=2.756, p=0.039), and shifts worked and personal
26 (n=56) and lack of variation in practice setting, which was two hospital maternity units in the
same region of Australia.
Researchers have identified relationships between traumatic birth experiences and
burnout. In the U.K., Sheen, Spiby, and Slade (2015) found a positive relationship between
posttraumatic stress symptoms related to a traumatic patient perinatal experience and
increased rates of emotional exhaustion (r(385)=0.420, p<0.001). They found a lesser
positive association between posttraumatic stress symptoms and depersonalization
(r(385)=0.247, p<0.001), and no significant association between posttraumatic stress
symptoms and lack of efficacy. Researchers in Israel found that the level of burnout among
their sample of 93 midwives was generally low (97.8%) but that burnout was positively
correlated to the number of exposures to traumatic births (Cohen, Leykin, Golan-Hadari, &
Lahad, 2017).
R. H. Bakker et al. (1996) investigated burnout among Dutch midwives, focusing on
workload as defined by number of procedures, visits, and births. They found that working
hours did not affect rates of emotional exhaustion or depersonalization and led to higher rates
of personal accomplishment. Social support was significantly negatively correlated to
emotional exhaustion (β= -0.28, P= <0.05) and depersonalization (β= -0.21, P= <0.05) and positively correlated with lack of accomplishment (β=0.26, P= <0.05). The study also found that burnout was lower for midwives attending a higher percentage of homebirths. However,
the study lacks applicability to the U.S., where (a) homebirth constitutes only 2.5% of U.S.
births (Hamilton, Martin, Osterman, Curtin, & Matthews, 2015), (b) only 4.8% of CNMs/CMs attend homebirths (Fullerton et al., 2015), and (c) homebirth does not have universal support (ACOG, 2011).
27 In a small study of burnout prevalence and its relationship with demographic
characteristics among Australian midwives (n= 58), Jordan et al. (2013) investigated
different types of burnout using the Copenhagen Burnout Inventory (CBI). Their findings
were significant in that they found high rates of burnout in the personal burnout (57%) and
work-related (57%) domains but not in the client-related domain (9%). The authors
concluded that these findings indicated midwives are susceptible to work
environment-related burnout but not the burnout environment-related to caring for women that is inherent to midwifery.
They also found that the type of burnout experienced was related to age and years in practice.
Midwives aged 35 or younger and with less than 10 years of experience scored highest in
rates of personal and work-related burnout, while midwives over 35 years old scored highest
in client-related burnout. The findings were limited by a small sample from a single
maternity unit, leading to potentially little variance in work environment.
Researchers have identified this pattern of burnout in other samples. In a much larger
sample of Australian midwives (n=1,037), Creedy and colleagues (2017) found a similar
pattern of burnout: 64.9% of midwives reported personal burnout and 43.8% work-related
burnout, but only 10.4% reported client-related burnout. In a study of Norwegian midwives
(n=598), researchers found a 20.1% rate of personal burnout, with 19.1% work-related and
4.2% client-related (Henriksen & Lukasse, 2016).
In contrast, in a study comparing burnout of 31 human service professions, Danish
researchers found that midwives (n=41) had the highest rate of work-related burnout,
consistent with Jordan et al. (2013), but also had the second highest rate of burnout related to
client interaction (Borritz et al., 2006). This indicated that both the work environment and the
28 single maternity unit in Denmark (n=50) published in 2017 reported lower rates of burnout in
all three domains of the CBI than found by Borritz and colleagues (2006): personal burnout,
22% (95% CI=12–36%); work-related burnout, 20% (95% CI=11–34%); and client-related burnout, 10% (95% CI=4–22%; Jepsen, Juul, Foureur, Sørensen, & Nøhr, 2017).
The conclusion drawn by Jordan et al. (2013) regarding burnout as an outcome of
organizational and administrative characteristics and not of actual care of women is
supported by a subset of midwifery burnout literature that addresses two practice models
referred to as continuity or shift models. Continuity models, referring to models in which
midwives follow women throughout the course of their pregnancy and birth, are referred to
as caseload and community midwifery. Practice models in which midwives work
pre-assigned shifts and do not follow specific women are referred to as standard care or hospital
midwifery (Newton, McLachlan, Willis, & Forster, 2014).
In a longitudinal study of hospital-based and caseload midwives in Australia,
midwives working in caseload models, in which they provided continuity of care,
experienced lower levels of personal burnout (14% compared to 49%, p<0.01) and
work-related burnout (5% compared to 40%, p<0.01) measured by the CBI (Newton et al., 2014).
The authors concluded that although the caseload model led to working more hours, it did not
lead to higher rates of burnout because of improved practice climate (e.g., increased levels of
autonomy), flexibility of work schedule, and opportunity to build relationships with women.
These findings were corroborated in a large sample (n=862) of Australian midwives that
found midwives practicing in a continuity model had significantly lower levels of burnout in
all three burnout domains measured by the CBI (CBI Personal Mean=50 vs. 58.3 p= 0.002;
29 Sidebotham, Gamble, & Creedy, 2017). The Danish study conducted by Jepsen and
colleagues (2017) referenced above also supported the findings that caseload midwives have
lower rates of burnout than other models of care.
In a study comparing levels of burnout of community midwives versus hospital
midwives in the U.K., Yoshida and Sandall (2013) found that working longer hours, being
unable to meet conflicting demands, manager approachability about flexible work, stress
recognition, and intention to leave were associated with burnout. They also found that
increased involvement in work-based decision-making, being able to decide how to do work, improved perception of the organization’s commitment to work-life balance, and satisfaction with opportunity to use abilities were protective against burnout. Their final model included
nine independent factors (control [24%], teamwork [14%], demographics [13%], work
management perceptions [10%], job satisfaction [6%], and work pattern [5%]) and predicted
73% of emotional exhaustion.
Kalicinska, Chylinska, and Wilczek-Rozyczka (2012) compared midwives (n=59) and hospice nurses (n=58) in Poland. They found that midwives had higher levels of
emotional exhaustion and depersonalization than hospice nurses (t(1,115)= -3.51, P< 0.01). Additionally, they found that poor managerial support was more strongly associated with emotional exhaustion (r= -0.58, p<0.01) and depersonalization (r= -0.43, p< 0.01) than poor social support (r= -0.46, p<0.01) and that social support was more strongly associated with alleviating emotional exhaustion and depersonalization among midwives than among hospice nurses.
Finally, there has been no research into midwifery burnout as an antecedent to
30 of inpatient hospital nurses (n=7,076) working at 161 hospitals, Cimiotti, Aiken, Sloane, and
Wu (2012) found that nurse burnout was positively correlated with higher incidences of
healthcare-associated urinary tract infections (β=0.82; P=0.03) and surgical site infections (β= 1.56, P<0.01) when controlling for nurse and hospital characteristics.
In summary, the scant evidence that does exist supports the theoretical propositions that supportive practice characteristics lead to less burnout among CNMs/CMs. However, most of this handful of studies have been conducted in Europe and not in the U.S. There has been increased attention paid to burnout in the U.S., with educational sessions at professional meetings (Breedlove, 2014) and self-help articles (Wright, 2016); still, as of yet, there has been no empirical data to explain this phenomenon in CNMs/CMs practicing in the present-day U.S. healthcare climate.
Work Engagement Theoretical Literature of Work Engagement
The concept of work engagement was born out of the positive psychology movement
that aims to move away from the pathological perspective of mental health and focus on
happiness and satisfaction (Seligman & Csikszentmihalyi, 2000). Kahn (1990) first
introduced engagement as a specific psychological state resulting from elements of the job
and driving outcomes. In a qualitative study he defined engagement as attending to task behaviors with one’s “preferred self,” which encourages connection; cognitive, physical, and emotional presence, and comprehensive role performances. Kahn determined that three
psychological conditions predicted engagement: meaningfulness, safety, and availability.
31 positive affective-cognitive state of mind related to one’s professional life, with the three primary attributes of vigor, dedication, and absorption (Schaufeli, Martinez, et al., 2002).
Some organizational psychologists have defined work engagement as the opposite of
burnout or on the spectrum with burnout (Maslach & Leiter, 1997). These researchers
assigned its attributes as the opposite of those of burnout: exhaustion versus energy;
depersonalization versus involvement; and lack of efficacy versus efficacy (Maslach &
Leiter, 1997). These suppositions included the belief that the two constructs can be measured
with the same instrument. For example, a high score on the MBI indicates high level of
burnout, and a low score indicates work engagement.
There are several theories from the field of organizational psychology that posit the
causes of engagement. The theory of burnout and engagement as antipodal constructs on a continuum extends to Maslach, Schaulfeli, and Leiter’s (2001) theory of the job-person mismatch or match, predicting engagement. They posited that congruence of expectations
and reality of workload, control, reward, community, fairness, and values leads to
engagement. Similarly, Kahn’s (1990) theory of engagement is built upon contractual imagery between the worker and the job. Kahn posited that people engage to the degree that
they feel (a) the benefit of the contract; (b) the meaningfulness of the contract; or (c) the
guarantee of safety provided by the contract. Following the same theory of a relationship
between worker and job, Saks (2006) proposed application of the social exchange theory to
explain work engagement. Saks proposed that as a job provides resources to a worker, the
worker reciprocates with varying degrees of engagement. The motivational process of the
JD-R posits that job resources predict engagement (A. B. Bakker & Demerouti, 2007). The
32 and organizational outcomes. Demands generally are associated with decreased engagement,
with the exception of those demands that are considered to be challenges (not hindrances),
which are theorized to increase engagement (Van den Broeck et al., 2010).
Empirical Literature: Organizational Attributes and Work Engagement of Nurses, Physicians, and Midwives
Due to the limited empirical research conducted with midwives and engagement, this
review will include nurses and physicians. In an integrative review of articles about nursing
engagement published between 1990 and 2013, García-Sierra, Fernández, and
Martínez-Zaragoza (2016) identified four organizational antecedents most studied by nurse
researchers: leadership; the six areas of work-life specified by Maslach, Schaufeli, and Leiter
(2001); structural empowerment; and social support.
In nurse engagement studies not included in García-Sierra, Fernández, and
Martínez-Zaragoza’s review, researchers have found relationships between a variety of elements of supportive practice environments and work engagement. Testing the mediating role of work
engagement, Santos, Chambel, and Castanheira (2016) found that engagement fully mediated
the positive relationship between (a) relational job characteristics (defined as nurses’ perceived impact on clients’ lives), commitment to clients, and perceived social worth and (b) organizational commitment. In a study of 300 clinical nurses in China that employed the Utrecht Work Engagement Scale and Lake’s Practice Environment Scale, researchers found supportive practice environments directly and positively affected work engagement and
indirectly affected work engagement through psychological empowerment (Wang & Liu,
2015). Finally, in a study of 2,115 medical residents, Prins et al. (2009) found that higher