Nature and scope of certified nurse-midwifery practice: A workforce study

Abstract

Aims and Objectives:

To describe the nature and scope of nurse-midwifery practice in Texas and to determine legislative priorities and practice barriers.

Background:

Across the globe, midwives are the largest group of maternity care providers despite little known about midwifery practice. With a looming shortage of midwives, there is a pressing need to understand midwives’ work environment and scope of practice.

Design:

Mixed methods research utilising prospective descriptive survey and interview.

Methods:

An online survey was administered to nurse-midwives practicing in the state of Texas (N = 449) with a subset (n = 10) telephone interviewed. Descriptive and inferential statistics and content analysis was performed.

Results:

The survey was completed by 141 midwives with eight interviewed. Most were older, Caucasian and held a master’s degree. A majority worked full-time, were in clinical practice in larger urban areas and were employed by a hospital or physician group. Care was most commonly provided for Hispanic and White women; approximately a quarter could care for greater numbers of patients. Most did not clinically teach midwifery students. Physician practice agreements were believed unnecessary and prescriptive authority requirements restrictive. Legislative issues were typically followed through the professional organisation or social media sites; most felt a lack of competence to influence health policy decisions. While most were satisfied with current clinical practice, a majority planned a change in the next 3 to 5 years.

Conclusions:

An ageing midwifery workforce, not representative of the race/ethnicity of the populations served, is underutilised with practice requirements that limit provision of services. Health policy changes are needed to ensure unrestricted practice.

Relevance to clinical practice:

Robust midwifery workforce data are needed as well as a midwifery board which tracks availability and accessibility of midwives. Educators should consider training models promoting long-term service in underserved areas, and development of skills crucial for impacting health policy change.

1 |. INTRODUCTION AND BACKGROUND

Maternal and child health is a high international priority. Midwives, the largest group of maternity care providers globally (WHO, 2013a), provide care for disproportionate numbers of vulnerable and underserved (Newhouse et al., 2011), and are essential for the delivery of critical health services (WHO, 2016). While birth rates have fallen across the globe since 1960 (The World Bank, 2017), the need for midwifery services has never been greater—particularly in rural and other more marginalised areas where training and retention of midwives have been problematic (WHO, 2013b).

To meet demand for midwifery services, it is crucial to understand the capacity for midwives to deliver comprehensive, quality care. Towards such end, the World Health Organization (WHO) issued a global mandate for nations to clarify midwifery roles and scopes of practice, as well as implement data collection and information systems to enable reliable reporting on the nursing and midwifery workforce status ([WHO] World Health Organization, 2016). This mandate re-emphasised the call to invest in active data collection and monitoring of the midwifery workforce as stated in The State of the World’s Midwifery (SoWMy) report. Ongoing comprehensive data collection was noted to be essential for determining availability, accessibility, acceptability and quality of the midwifery workforce (United Nations Population Fund, 2014a), as well as for workforce planning (Pozo-Martin et al., 2017).

Impressively, the SoWMy report noted progress in collecting midwifery workforce data by nearly three quarters of the 73 low- and middle-income countries examined. However, in many of these and other countries—both moderate and high income, there is lack of clarity regarding the nature and scope of practice, the employment setting, geographic distribution and barriers to practice. Part of the confusion is a lack of differentiation between nursing and midwifery, causing limited understanding of the care midwives provide distinct from nurses. Existing workforce studies often include midwifery under the broader umbrella of “advanced practice nurses” or “nurses” in general. A distinct profession and practice discipline (ICM, 2011), midwifery suffers from a lack of workforce data, hampering legislative, educational and practice efforts to meet population health needs. Robust data are needed to impact health policy change which promotes retention of midwives, as well as improve access to quality care. If the United Nations Sustainable Development Goals are to be met—particularly Goal 3, to ensure healthy lives and promote well-being for all ages (United Nations, 2015), knowledge of the midwifery workforce is crucial.

To date, there have been few national midwifery workforce studies and no known international efforts. In the USA (U.S.), there have been but a handful of comprehensive midwifery workforce studies which have all focused on select regions or individual states (Gordon & Erickson, 1993; Hastings-Tolsma et al., 2015; Jevitt & Beckstead, 2004; Kozhimannil, Henning-Smith, & Hung, 2016). These surveys provided valuable information for shaping midwifery practice, education and health policy reform at the state level. However, national and comparative international data are urgently needed to ensure adequate numbers of midwives to meet the need for sexual, reproductive, and maternal and newborn health services (United Nations Population Fund, 2014b; ten Hoope-Bender et al., 2014). While the U.S.A has initiated a national Midwifery MasterFile to profile nurse-midwives distinct from other advance practice nurses, the data collected are not projected to detail the nature of nurse-midwifery practice (Fullerton et al., 2015). Where robust data are collected using standardised terminology and definitions, interoperability can occur with immediate local impact for the improvement of midwifery services (ten Hoope-Bender et al., 2016).

The purpose of this research was to provide workforce data relevant to certified nurse-midwives (CNMs) practicing in Texas—the largest conterminous state in the U.S.A with the second largest population, as well as to serve as a model for a national survey. Examination of the CNM workforce in Texas was particularly relevant as the state has been ranked in the lower 25% of U.S. states for scope of nurse-midwifery practice (Beal, Batzli, & Hoyt, 2015), has large underserved rural areas and is known to have large areas with no nurse-midwifery services ([ACNM] American College of Nurse Midwives, 2016a, [ACNM] American College of Nurse Midwives, 2016b). Not unlike global trends, the U.S.A is experiencing a shortage in maternal health services (Ollove, 2016) with close to half of U.S. counties lacking a single obstetrician/gynaecologist and 56% without a single nurse-midwife ([ACNM] American College of Nurse Midwives, 2016a, [ACNM] American College of Nurse Midwives, 2016b). With the female population and births in the U.S.A projected to increase sharply over the next decades and as a state with one of the fastest growing population rates (U.S. Census Bureau, 2011), understanding the nature and scope of midwifery practice in Texas has the potential to serve as a bellwether for future access to maternity services.

It is important to note that in the U.S.A, CNMs are graduates of an education programme accredited by the Accreditation Commission for Midwifery Education with professional nursing knowledge and skills a prerequisite to study ([ACNM] American College of Nurse Midwives, 2016a, [ACNM] American College of Nurse Midwives, 2016b). Those who practice as a CNM must pass a national certification exam administered by the American Midwifery Certification Board (AMCB) with requirements for ongoing certification maintenance. In 2016, there were 11,628 CNMs certified by the AMCB with close to 650 first entering the workforce (AMCB, 2016). AMCB also certifies midwives (CMs) who have completed an ACME accredited programme although these midwives have not completed a professional nursing programme, have few numbers (n = 100) and are unable to practice in most U.S. states ([AMCB] American Midwifery Certification Board, 2016). There are also other types of midwives working in the U.S.A, such as direct-entry midwives who are trained through an apprenticeship model comparable to that of traditional birth attendants in other countries, and Certified Professional Midwives (CPMs) who are also direct-entry midwives with education validated by the Midwifery Education Accreditation Council and take a certification examination administered by the North American Registry of Midwives (NARM, 2017). Direct-entry midwives largely practice in out-of-hospital settings and CNMs typically attend births in the hospital setting (Fullerton et al., 2015). Of the more than 3.9 million births in the U.S.A, 98.5% occur in the hospital setting with approximately 8% attended by CNMs (Martin, Hamilton, Osterman, Driscoll, & Mathews, 2017). CNMs and CMs are the only midwives with professional competency expectations to include reproductive and primary care to women of all ages. While CNM services are covered by government and most private health insurance programmes, other midwifery group coverage is highly variable.

As the largest group in the U.S. midwifery workforce, it is important that information about the nature and scope of CNM practice be detailed to meet the healthcare challenges which exist at both the state and national levels. The current absence of workforce data regarding Texas CNMs highlights the need for study to allow for effective policy and decision-making; such data can also further inform policies regarding CNM care at the national level. The aim of this research was to determine for Texas CNMs the (i) demographics and employment characteristics, (ii) nature and scope of practice, (iii) geographic distribution and populations served, (iv) key legislative priorities and practice barriers and (v) future practice plans.

2 |. METHODS

2.1 |. Design

This mixed methods research utilised a prospective descriptive survey, as well as interview of a subset of participants. Survey design is most efficient in gathering data from a large population in a cost-effective, flexible and dependable manner. Additional qualitative interview added greater insight into midwifery practice, augmenting and confirming quantitative findings. This mixed methods sequential design allowed for better understanding of the nature and scope of nurse-midwifery practice than either approach alone would yield (Creswell, 2014).

2.2 |. Participants and Setting

Nurse-midwives certified to practice in Texas (N = 449) were eligible for participation. Other types of midwives (i.e., CPMs, CMs, lay midwives, student midwives) were excluded due to significant variation in education and training, differences and/or lack of certification and licensure, as well as greater restriction in scope of practice and third-party compensation eligibility, making it difficult to develop survey items relevant to all types of midwives. However, a separate and parallel survey of Texas direct-entry midwives was simultaneously conducted and comparative data are reported elsewhere (Burpo et al., in press). All participants were currently certified by AMCB or were in AMCBs retirement classification.

A master email contact list of all certified or retired nurse-midwives is maintained by the AMCB—the national certifying body for nurse-midwives in the U.S.A, with the ability to distribute emails to CNMs by state of practice or residence. Email invitation with a URL link was distributed by the AMCB to all Texas CNMs; 438 (98%) were successfully sent. Once the survey was accessed, directions were provided for survey completion which could be done in more than one sitting. The survey was available for a three-week period with reminders sent at the end of weeks one and two.

On survey completion, participants were asked about their interest in being interviewed to discuss issues related to CNM practice. Those interested were directed to a separate website where they provided contact information. Nonrandom quota sampling was then used to select a small subset of participants (N = 10) who were selected to ensure representation from varied practice types (freestanding birth centre, hospital, private physician-CNM, home birth, military facility) and geographic locations (rural/small-town, suburban, large metropolitan). The interviews were conducted to both deepen understanding of CNM practice issues and provide confirmation of survey results (Harris & Brown, 2010).

Survey participants were eligible for a lottery draw (three paid registrations for the annual American College of Nurse-Midwives meeting); those interviewed were compensated US $25. Following approval by the AMCB Research Committee and Executive Board, the research received ethical approvals from the Baylor University Institutional Review Board (#717761–1) and the Colorado Multiple Institutional Review Board (#15–1942).

2.3 |. Survey Questionnaire

A previously developed workforce survey for CNMs (Hastings-Tolsma et al., 2015) was adapted for use. Items were reviewed for relevancy; irrelevant items were deleted and other items added to meet study objectives. The adapted survey was then piloted with four items subsequently revised. The final 125-item survey focused on eight practice areas: demographics, employment profile, clinical practice, midwifery leadership, consultation/credentialing/liability, prescriptive authority, teaching professional students and professional satisfaction/future trends. Branch logic avoided the need to answer questions that were irrelevant to the respondent’s practice, and respondents could skip items. To view the final survey, see the Data S1.

Study data were collected and managed using REDCap electronic data capture tools hosted at the University of Colorado Denver Anschutz Medical Campus. REDCap (Research Electronic Data Capture) is a secure, web-based application designed to support data capture for research studies, providing (i) an intuitive interface for validated data entry; (ii) audit trails for tracking data manipulation and export procedures; (iii) automated export procedures for seamless data downloads to common statistical packages; and (iv) procedures for importing data from external sources (Harris et al., 2009).

2.4 |. Individual telephone interview

An interview guide related to midwifery practice was developed by the researchers (see Table 1). Because interview participants were from a large geographic area with diverse practice demands, telephone interviews were arranged at a time convenient for the participant. Interviews lasted from 30–60 min. All were taped and uploaded for transcription. The same researcher conducted all interviews (SWF) which took place in the 8 weeks following survey questionnaire completion.

TABLE 1.

Interview guide and themes from interview of Texas CNMs (n = 8)

Interview Guide
Describe the factors that make your midwifery practice a success.
What barriers, if any, keep your practice from growing or reaching its full potential?
How do you think midwifery is viewed in Texas?
What do you think are the biggest challenges for midwifery practice in Texas?
Describe your thoughts on how collaboration between CNMs and CPMs might best be explored.
Themes	Sub-Categories
Practice success	Out-of-hospital practice	Hospital practice
	Patient satisfaction	Ancillary support (e.g., billing, credentialing)
	Shared decision-making with patients	Collaboration between CNMs and physicians
	Business knowledge	Range of provider philosophies
	Judicious use of resources
Practice barriers	Patients unaware of out-of-hospital options	Hospital administrators controlling
	Lack of physician support, hostility	Denied hospital admitting privileges
	Board of Nursing does not understand midwifery	Physician refusal to consultant when colleagues unavailable
	Empanelment restrictions	Physician fear of competition
	Delegated practice authority	Poor knowledge of practice finances
View of Midwives in Texas	Public ignorance or misperceptions regarding midwifery role
Key Legislative challenges	Independent prescriptive authority
	Reimbursement equity
	Inability to admit patients under CNM name
	Monies for midwifery fellowships
	Overcoming TMA barriers to independent practice
CNM-CPM collaboration	Lack of experience working with CPMs
	Improved regular communication
	Doable if there were one regulatory board for all midwives
	Compensation restrictions make collaborative work difficult

CNM, certified nurse-midwife; CPM, certified professional midwife; TMA, Texas Medical Association.

2.5 |. Analysis

Statistical analyses were conducted using SPSS 21 (IBM Corp., 2012). Descriptive analyses included Fisher’s exact Chi-Square tests, independent samples t test/Mann Whitney U test and one-way ANOVA/Kruskal–Wallis test with pairwise comparisons. When sample size was small, continuous variables were assessed using the Shapiro–Wilk’s test to determine whether parametric or nonparametric tests were needed.

Transcribed interviews were read by the same two researchers (MHT, SWF). The structured, detailed analysis of the transcripts used systematic text condensation (Malterud, 2012). This pragmatic, descriptive method of data interpretation includes (i) total impression, (ii) identifying and sorting meaning units, (iii) condensation and (iv) synthesising. Researchers identified commonalities and differences across interviews and then created themes and categories using an iterative process to uncover full meaning.

3 |. RESULTS

3.1 |. Demographics and employment characteristics

The survey was completed by a total of 141 CNMs for a response rate of 32%. Nearly, all respondents were female (n = 139, 98.6%), Caucasian (n = 126, 92.0%) and held a master’s degree (n = 122, 86.5%). Respondents were most often aged 56–65 years (n = 41, 29.3%). Nearly, all were certified by the AMCB (n = 138, 97.9%) with initial certification ranging from 1971–2016 (mode = 2014).

Dual certification (e.g., family nurse practitioner, women’s health nurse practitioner) was held by 19.9% (n = 28). Fewer with dual certification worked in hospitals/medical centres (n = 6, 50% vs. n = 6, 50%) and urban settings (n = 18, 78.3% vs. n = 5, 21.7%) than CNMs without.

Texas CNMs tended to be members of the ACNM—the professional nurse-midwifery association (n = 109; 77.3%). ACNM membership differed significantly for those who worked full-time (n = 70, 76.9%) (X²(1) = 26.385, p < .001) and those who were part-time (n = 14, 82.4%) (X²(1) = 7.118, p < .01). There were no significant differences found in ACNM membership for those working per diem (n = 3, 75%) (p = .32). See Table 2.

TABLE 2.

Demographics of Texas CNMs (N = 141)

Characteristic	n	%	Characteristic	n	%
Gender			AMCB Certification
Female	139	98.6	Current	138	97.9
Male	0	–	Retired	3	2.1
Transgender	0	–	First year Licensed to practice in Texas
No response	2	1.4	1971–1980	11	7.9
Age (in years)			1981–1990	18	12.9
25–35	17	12.1	1991–2000	37	26.4
36–45	34	24.3	2001–2010	32	22.9
46–55	35	25.0	2011–2016	42	30.0
56–65	41	29.3	Other Certificationsa
66–75	11	7.9	FNP	10	7.1
>75	2	1.4	WHNP	11	7.8
Race/Ethnicity			Other	9	6.4
African American	8	5.8	Professional memberships
Caucasian	126	92.0	ACNM	109	77.3
Hispanic	1	0.7	ACOG	11	7.8
Asian	1	0.7	ANA	13	9.2
Native American	–	–	APHA	3	2.1
Other	1	0.7	AWHONN	28	19.9
Educationa			MANA	5	3.5
Associate degree	26	18.4	Other	18	12.8
Bachelors	82	58.2
Masters	122	86.5
DNP	9	6.4
PhD	11	7.8
Other	7	5.0

^aCould answer more than one category.

Percentages may not equal 100 due to rounding.

DNP, doctor of nursing practice; PhD, doctor of philosophy; AMCB, American Midwifery Certification Board; FNP, family nurse practitioner; WHNP, women’s health nurse practitioner; ACNM, American College of Nurse Midwives; ACOG, American Congress of Obstetricians and Gynecologists; ANA, American Nurses Association; APHA, American Public Health Association; AWHONN, Association of Women’s Health, Obstetric and Neonatal Nursing; MANA, Midwives Alliance of North America.

Most respondents were employed full-time (n = 91, 64.5%). In primary midwifery employment, main responsibilities were for antepartum (n = 90, 63.8%), intrapartum (n = 86, 61.0%) and gynaecologic (n = 73, 51.8%) care. Those not employed in midwifery (n = 21, 14.9%) cited inability to find a position (n = 8, 38.1%), preferring to work in another capacity (e.g., Family Nurse Practitioner, registered nurse; n = 6, 28.6%), personal reasons (e.g., family obligations; n = 5, 23.8%), liability issues (e.g., cost of malpractice insurance; n = 4, 19%), not interested in midwifery practice (n = 3, 14.3%) or retired (n = 3, 14.3%), as the reason.

Most CNMs were in clinical practice (n = 103, 73%) with clinical hours per week ranging from fewer than 10 to more than 60 with most working 33 to 40 (n = 44, 31.4%). CNMs reported typically spending 3 to 5 hr a week (n = 40, 28.8%) on nonpatient care (e.g., laboratory report review, completing referral forms). Most CNMs held primary employment with a hospital/medical centre (n = 30, 21.3%) followed by a physician group (n = 25, 17.7%). Prior clinical experience was required for employment by 22.6% (n = 31). Fewer than one-third held additional employment (n = 38, 27.1%).

Midwifery practices typically had a designated practice director (n = 58, 56.3%) who was generally a CNM (n = 49, 49.5%). Most CNMs reported having collaborative clinical practice guidelines (n = 89, 85.6%), usually reviewed annually (n = 58, 65.2%). Consultation was typically with attending physicians (n = 82, 78.8%) who were most often obstetricians/gynaecologists (n = 84, 91.3%) or maternal-foetal medicine specialists (n = 55, 59.8%) and were employed by the same organisation as the CNM (n = 48, 52.2%). Most CNMs believed that a formal practice agreement with physicians was unnecessary (n = 54, 52.9%).

Allied health hospital staff membership (n = 39, 37.1%) was held by most CNMs with 60% (n = 63) unable to admit patients to hospital under their own name (n = 63, 60.0%). Few held voting privileges consistent with physicians (n = 7, 29.2%).

Quality improvement processes were in place in most practices (n = 71, 68.9%) and most carried professional liability insurance (n = 81, 77.9%). Few (n = 13, 12.5%) reported that liability coverage impacted ability to practice midwifery. Where it did affect ability to practice, it influenced the ability to obtain physician consultation (n = 1, 7.7%), hospital privileges (n = 4, 30.8%) and to precept students (n = 4, 30.8%).

U.S. Drug Enforcement Agency (DEA) licensure was required of respondents by most employers (n = 54, 52.4%), as was prescriptive authority (n = 73, 70.9%) which was held by most CNMs (n = 96, 93.2%). Prescriptive authority requirements were believed to limit the scope of midwifery practice (n = 62, 44.0%); few agreed that physician supervision was necessary (n = 5, 4.9%).

Typical CNM income was US $75,001 to US $100,000 (range < US $30,000 to US $200,000). Without prior clinical experience, the typical starting salary was US $80,001 to US $90,000/year (range < US $50,000 to US $120,000); those with prior experience had a typical starting salary of US $90,001 to US $100,000 (range < US $50,000 to US $120,000). See Table 3.

TABLE 3.

Employment characteristics of Texas nurse-midwives (N = 141)

Characteristic	n	%	Characteristic	N	%
Employment Status			Primary Responsibilitiesa
Full-time	91	64.5	Antepartum	90	63.8
Part-time	17	12.1	Intrapartum	86	61.0
Per diem	4	2.8	Newborn	25	17.7
Retired	17	12.0	Gynecology	73	51.8
Unemployed	7	5.0	Primary Care	46	32.6
Other	5	3.5	Clinical MW education	26	18.4
Clinical Practice-primary			MW educational program	5	3.5
Yes	103	73.0	Administration	15	10.6
No	38	27.0	Research	7	5.0
Primary Employment Site			Other/NA	6	4.3
Hospital	30	21.3	Clinical Hours/week
Military/Federal	5	3.5	<10	6	5.4
Health Science Center	7	5.0	11–20	6	5.4
Community Clinic	3	2.1	21–32	11	9.9
NFP Clinic	3	2.1	33–40	44	39.6
HMO	0	–	41–60	38	34.2
Self	7	5.0	>60	6	5.4
Physician practice	25	17.7	Prior Clinical Experience Required for Employment
Midwifery group practice	12	8.5	Yes	31	22.6
State government	1	0.7	No	67	48.9
Educational institution	19	13.5	Don’t know	9	6.6
Other	15	10.6	N/A (self-employed)	3	2.2
CNM Salary (with experience)			N/A (not in midwifery practice)	27	19.7
<$50,000	3	2.2	CNM Salary (no prior experience)
$50,001–60,000	–	–	<$50,000	5	3.7
$60,001–70,000	5	3.7	$50,001–60,000	5	3.7
$70,001–80,000	7	5.1	$60,001–70,000	6	4.4
$80,001–90,000	23	16.9	$70,001–80,000	16	11.9
$90,001–100,000	32	23.5	$80,001–90,000	35	25.9
$100,001–110,000	16	11.8	$90,001–100,000	21	15.6
$110,001–120,000	2	1.5	$100,001–110,000	–	–
>$120,001	–	–	$110,001–120,000	1	0.7
Unknown	25	18.4	>$120,001	–	–
Income from CNM Practice (gross)			Unknown	22	16.3
<$30,000	5	3.6	Hospital Staff membership
$30,001–50,000	3	2.2	Medical, full	12	11.4
$50,001–75,000	16	11.6	Medical, associate	12	11.4
$75,001–100,000	47	34.1	Allied Health	39	37.1
$100,001–125,000	30	21.7	Don’t know	10	9.5
$125,001–150,000	6	4.3	NA (out-of-hospital)	19	18.1
$150,001–200,000	4	2.9	NA (do not attend births)	17	12.1
>$200,001	0	–	Other	1	0.7
NA (not in midwifery practice)	27	19.6	Hospital Voting Privileges
Prescriptive Authority			Full, like physicians	7	29.2
Yes	96	93.2	No voting privileges	12	50.0
No	7	6.8	Don’t know	4	16.7
Prescriptive Authority Required for Employment			NA (don’t provide intrapartum care)	1	4.2
Yes	73	70.9	Admit Patients to Hospital Under Own Name
No	15	14.6	Yes	27	25.7
Don’t know/other	15	14.6	No	63	60.0
DEA Licensure Required			Don’t know	2	1.9
Yes	54	52.4	NA	13	12.4
No	45	43.7	Professional Liability Insurance
Don’t know	3	2.9	Yes	81	77.9
Other	1	1.0	No	23	22.1
Designated Midwifery Practice Director			Extent of Liability Coveragea
Yes	58	56.3	Occurrence	61	43.3
No	31	30.1	Claims-made	49	34.8
NA (solo practice)	5	4.9	Tail	36	25.5
Other	9	8.7	Liability Coverage Effect Practice Ability
Service Director Background			Yes	13	12.5
CM	1	1.0	No	91	87.5
CNM	49	49.5	Named as Defendant for Malpractice
Registered Nurse	2	2.0	Yes	12	11.5
Physician	16	16.2	No	92	88.5
Business/health Admin	7	7.1	Nature of Physician Consultation
Don’t know	0	–	1 or more attending	82	78.8
NA (solo practice)	14	14.1	Residents	10	9.6
Other	10	10.1	No arrangements	10	9.6
Not nurse midwife	7	7.1	Don’t know	2	1.9
Collaborative Practice Guidelines			Type of Physician Consultationa
Yes	89	85.6	OB/GYN	84	91.3
No	14	13.5	MFM	55	59.8
Don’t know	1	1.0	Family Medicine	8	8.7
Review of Collaborative Practice Guidelines			Other	4	4.3
Annually	58	65.2	Access to Consulting Physicians
Determined by physician	16	18.0	CNM employer	21	22.8
As requested by MW	12	13.5	Employed by same organization	48	52.2
Don’t know	8	9.0	External to MW practice	25	27.2
Other	6	6.7	In-house	20	21.7
Quality Review Process			Off-site, available as requested	9	9.8
Yes	71	68.9	Other	5	5.4
No	24	23.3	Need for Formal CNM-MD Practice Agreement
Don’t know	8	7.8	Yes	35	34.3
Nature of Quality Reviewa			No	54	52.9
M&M review	43	30.5	No opinion	13	12.7
External peer review	25	17.7	Need for MD Prescriptive Authority Supervision
Internal process	73	51.8	Yes	5	4.9
Nursing peer review	14	9.9	No	87	84.5
Risk management	47	33.3	Undecided	11	10.7
Do not participate	5	3.5
Don’t know	14	9.9

^aCould answer more than one category.

Where numbers do not equal 141, there were missing responses. Percentages may not equal 100 due to rounding.

CNM, certified nurse-midwife; OB/GYN, obstetrician/gynaecologist; MFM, maternal–foetal medicine; M&M, mortality and morbidity.

Most CNMs received employer benefits of vacation pay, retirement, sick time, basic life support certification fees, life and health insurance, DEA registration fees, internet access and continuing education funds. Fewer than 10% received support for school loan repayment, shift differential, support for child care or engaged in profit-sharing (see Figure 1). Benefits significantly differed for self-employed, hospital and nurse-midwifery group employed midwives (X²(2) = 8.964, p = .011) such that the average number of benefits was significantly lower for self-employed CNMs (5.14 ± 6.18), compared to those employed by a hospital/medical centre (15.83 ± 9.46) (p = .02). Of note, 40 respondents had completed their midwifery education between 2010–2016 and reported an average debt burden of US $68,737 (range US $0 – US $200,000); only 12 (9.2%) had employer-provided support for loan repayment. There was no difference in employer school loan repayment benefit based on geographic practice area (p = .20).

FIGURE 1.

Employer paid benefits (N = 140)

3.2 |. Nature and Scope of Practice

Practicing CNMs (n = 105) typically worked 9 or fewer days in clinic each month (n = 41, 41%) caring for an average of 16 to 20 patients a day (n = 35, 35.7%). More than one quarter stated the ability to care for greater numbers of patients (n = 29, 28.7%). Finally, close to one-third (n = 49, 47.6%) provided 1 to 10 days of uncompensated midwifery care each year (e.g., international work, volunteer clinics). Overall, total estimated annual uncompensated care was US $80,278 to US $110,139.

Those in clinical practice most often spent a quarter to half of their practice time providing antepartum (n = 45, 45.5%) and intrapartum (n = 34, 39.1%) care; 25% or less time was spent in postpartum care (n = 80, 82.5%). Gynaecologic care was provided 10% to 50% of the time by respondents (n = 46, 49%). Fewer spent more than 10% of clinical time caring for newborns (n = 15, 34.9%) or in primary care (n = 42, 40.7%).

Of those in clinical practice, 81.9% attended births (n = 86); three quarters did so in a hospital setting (n = 65, 46.1%). CNM practices most typically attended 10 to 20 births each month (n = 31, 36.5%). CNMs who took call (n = 66, 62.9%) typically did so out-of-hospital (n = 55, 83.3%) for over 100 hours each month (n = 42, 64.6%). See Table 4.

TABLE 4.

Nature and scope of CNM Care in Texas (N = 141)

Variable	n	%	Variable	n	%
Clinic Days/Month			Attend Births
0–9	41	41.0	Yes	86	81.9
10–19	38	38.0	No	19	18.1
20–29	19	19.0	Birth Setting
≥30	2	2.0	Hospital	65	75.6
Patients Cared for in 8 hr Clinic Day			Birth Center
1–5	2	2.0	Hospital	1	1.2
6–10	20	20.4	Freestanding	26	30.2
11–15	32	32.7	Home	11	12.8
16–20	35	35.7	Monthly Births in CNM Practice
21–25	4	4.1	<10	12	14.1
26–30	4	4.1	10–20	31	36.5
>30	1	1.0	21–35	13	15.3
Current Workload			36–50	6	7.1
Overextended	12	11.9	51–75	7	8.2
At full capacity	60	59.4	76–100	8	9.4
Able to see more patients	29	28.7	>100	8	9.4
Uncompensated (Charity) Care/Year			Take Call
1–2 days	18	17.5	Yes	66	62.9
3–5 days	16	15.5	No	39	37.1
6–10 days	15	14.6	Call Taken From
11–14 days	7	6.8	Out-of-hospital	55	83.3
15–21 days	2	1.9	In-hospital	9	13.6
22–31 days	2	1.9	Other	2	3.0
>31 days	5	4.9	Call Hours/Month
None	39	36.9	<24	4	6.2
Serve as Clinical Preceptor			24–48	4	6.2
Yes	48	46.6	49–72	7	10.8
No	52	50.5	73–100	8	12.3
Don’t Know	3	2.9	>100	42	64.6
Midwifery Students Precepted			Other Types of Students Precepted
Nurse-midwives	63	44.7	WHNP	41	29.1
Professional midwives	9	6.4	FNP	49	34.8
Lay midwives	7	5.0	EMT/paramedic	16	11.3
Do not teach MW students	32	22.7	Basic RN	26	18.4
Percentage of Time Spent Teaching MW Students			Medical students	26	18.4
<25	66	80.5	Physician residents	23	16.3
25–50	11	13.4	Physician assistants	12	8.5
51–75	5	6.1	Other	5	3.5

Where numbers do not equal 141, there were missing responses. Percentages may not equal 100 due to rounding.

MW, midwifery; CM, certified midwife; CNM, certified nurse-midwife; WHNP, women’s health nurse practitioner; FNP, family nurse practitioner; EMT, emergency medical technician; RN, registered nurse.

3.3 |. Care practices

Most CNMs providing intrapartum care (n = 102) felt the birth setting (n = 67, 68.3%), and consulting physicians (n = 60, 60.6%) were supportive of physiologic, normal birth. Most agreed there was adequate time for shared decision-making with patients (n = 59, 61.5%), that intermittent auscultation for foetal monitoring was used unless medically contraindicated (n = 48, 52.2%), and patients could give birth in the position of their choice (n = 54, 62%). Fewer than half agreed that patients could eat/drink in labour as desired (n = 42, 44.2%). Intravenous access was routinely provided (n = 57, 54.3%), as was continuous electronic foetal monitoring (CEFM) (n = 43, 41.0%), and 27% (n = 24) agreed time constraints were imposed on a woman’s labour. Two-thirds (66.3%) agreed that women desired physiologic birth.

Nonpharmacologic and pharmacologic methods commonly offered during labour included ambulation (n = 78, 74.3%), epidurals (n = 65, 61.9%), opioids (n = 63, 60%), continuous midwife labour support (n = 53, 50.5%), hydrotherapy (n = 51, 48.6%) and music (n = 52, 49.5%). Fewer offered acupressure (n = 18, 17.1%), nitrous oxide (n = 19, 18.1%), sterile water injections (n = 17, 16.2%), transcutaneous nerve stimulation (TENS) (n = 12, 11.4%) or paracervical block (n = 4, 3.8%). Although most midwives believed water birth should be offered (n = 46, 80.7%), it was only offered by one-third (n = 31, 35.2%).

Elective labour induction was not routinely offered (n = 72, 69.9%); only 13 (18.1%) believed it should be. Earliest gestational age where elective induction was offered was 39.0 weeks (n = 22, 84.6%; range 39.0–41.0). Postdates induction was typically offered at 41.0 of weeks gestation (n = 34, 33.7%), followed by 42.0 (n = 16, 15.8%), 41.5 (n = 13, 12.9%), 41.6 (n = 4, 4.0%) and greater than 42 weeks (n = 6, 5.9%).

Vaginal birth after Caesarean (VBAC) option was available for most CNM patients (n = 77, 79.4%), although 25% (n = 5) felt that it should not be offered. Most CNMs needed to collaboratively manage VBAC patients with a physician (n = 33, 42.9%) or transfer care (n = 12, 15.6%). Primary elective Caesarean birth was offered in 16.5% (n = 17) of midwifery practices; 8.6% (n = 7) believed that it should be.

3.4. |. Primary care and advanced midwifery skills

Advanced practice midwifery skills are defined as those not routinely taught in basic education programmes in the U.S.A (ACNM, 2012). More than half of CNMs performed limited ultrasound (i.e., confirm intrauterine pregnancy and viability, early dating, foetal presentation) (n = 71, 68.3%) and manual removal of the placenta (n = 68, 66%). Few performed external version for breech presentation (n = 11, 10,8%), medical abortion (n = 11, 10.7%), gynaecologic ultrasound (n = 9, 8.8%), colposcopy (n = 9, 8.7%), male circumcision (n = 8, 7.8%), standard/detailed ultrasound (e.g., foetal biometry) (n = 8, 7.8%), fourth degree perineal laceration repair (n = 7, 6.9%) or paracervical block (n = 7, 6.8%).

Respondents were also asked about the frequency of primary care by body system. Care was frequently provided for psychological (52.9%), renal (39.4%), endocrine (36.5%) and reproductive (26.9%) conditions. See Figure 2.

FIGURE 2.

Frequency of primary care conditions managed by CNMs (N = 104)

3.5 |. Teaching involvement

Most CNMs did not teach students clinically (n = 52, 50.5%). Where clinical teaching was performed, it was usually for student nurse-midwives (n = 63, 44.7%), typically spending less than 25% of time in doing so (n = 66, 80.5%). Comments regarding why midwifery students were not clinically taught largely reflected employer restrictions (n = 7), being in practice less than one year (n = 6), and not in full-scope practice (n = 5). Few who clinically taught were compensated (n = 19, 21.1%); where compensated, it was through financial remuneration to the practice (n = 12, 63.2%) or the individual CNM (n = 7, 36.8%), library access (n = 3, 15.8%) or continuing education credit award (n = 6, 31.6%).

Table 4 details key aspects of the nature and scope of CNM practice.

3.6 |. Geographic distribution and populations served

Primary populations cared for included Hispanic (n = 74, 52.5%), White (n = 72, 51.1%) and African American/Black (n = 39, 27.7%) with compensation largely from government Medicaid (n = 98, μ = 45.12, SD = 36.23), commercial insurance (n = 97, μ = 31.44, SD = 28.51) and self-pay (n = 99, μ = 15.57, SD = 21.73). Patient risk status was judged to be primarily low-moderate (n = 42, 40.8%) to moderate (n = 37, 35.9%) risk. There were no significant differences in geographic practice area and providing care for higher risk patients, where respondents attended births, the ability to admit women to a hospital under the CNMs name or in having liability insurance.

Most CNMs worked in large urban areas with a population > 250,000 (n = 53, 47.3%); few did so in more rural areas (population ≤ 50,000) (n = 21, 18.8%) (see Table 5 and Figure 3). CNMs with prior clinical experience were not more likely to work in rural areas (n = 5, 16.2%) than those without (n = 12, 17.9%). Prescriptive authority requirements limited where some CNMs practiced geographically (n = 23, 16.3%). Average distance travelled between personal residence and work was 15.56 miles (25 km) (range: 0–272 miles or 0–437.7 km). See Table 5.

TABLE 5.

Geographic distribution and populations served (N = 141)

Characteristic	n	%	Characteristic	n	%
Primary populationsa			Patient risk level
African American/Black	39	27.7	Low	16	15.5
Asian/Pacific Islander	14	9.9	Low-moderate	42	40.8
Hispanic/Latina	74	52.5	Moderate	37	35.9
Native American	6	4.3	High	8	7.8
White	72	51.1	Paymenta
Other	3	2.1	Medicaid (government)	98	69.5
Geographic practice area			Medicare (government)	79	56.0
Rural	5	4.5	Commercial insurance	97	68.8
Town (<10,000 population)	3	2.7	Military	82	58.2
City			Self-pay	99	70.2
10,001–50,000	13	11.6	Uncompensated	81	57.4
50,001–100,000	24	21.4
100,001–250,000	14	12.5
>250,000	53	47.3

^aCould answer more than one category.

Where numbers do not equal 141, there were missing responses. Percentages may not equal 100 due to rounding.

FIGURE 3.

Primary clinical practice location by county in Texas (N = 141)

3.7 |. Legislative issues and practice barriers

Key legislative priorities noted by respondents were the need for independent prescriptive authority (n = 107, 75.9%), elimination of physician supervisory language in hospital bylaws (n = 73, 51.8%), and obtaining hospital privileges and admitting privileges (n = 71, 50.4%) (see Table 6).

TABLE 6.

Priority legislative issues of Texas CNMs (N = 141)

Legislative Issue	n	%
Independent prescriptive authority	107	75.9
Elimination of physician supervisory language in hospital bylaws	73	51.8
Obtaining hospital privileges	71	50.4
Admitting privileges	71	50.4
Midwifery representation on agencies creating health policy	66	46.8
Support for physicians who collaborate with midwives	58	41.1
Support for independent midwifery practice	57	40.4
Need for independent midwifery board	49	34.8
Reimbursement rates	46	32.6
Empanelment by managed care organisations	29	20.6
Obtaining malpractice insurance	28	19.9
Timely response to application for application for hospital privileges	25	17.7
Need for independent advanced practice nursing board	23	16.3
Monies to support graduate midwifery education	21	14.9
Support for ACME nurse-midwifery education	18	12.8
Obtaining medical consultation services	11	7.8
Regulations from the Texas Board of Nursing	11	7.8
Legalisation of Certified Midwife licensure	6	4.3
Other (e.g., loan forgiveness, practice reciprocity in other states)	4	2.8

More than half of CNMs felt they could influence health care in Texas (n = 87, 62.6%). Legislative participation was primarily carried out by following issues through ACNM (n = 92, 65.2%) or contact with legislators (n = 52, 36.9%). Fewer than 10% were members of a legislative committee to promote midwifery or women’s health (n = 11, 7.8%) or engaged in activities like inviting legislators to visit practice sites (n = 5, 3.5%) or education programmes (n = 1, 0.7%). Nearly 25% of CNMs reported no legislative participation (n = 31, 22%); most did not feel competent to influence legislation (n = 86, 62.8%).

CNMs typically felt that regulation should be by either an independent board of midwifery regulating all types of midwives (n = 48, 34.3%) or a board of midwifery regulating CNMs only (n = 34, 24.3%). Most (n = 119, 89.5%) did not believe direct-entry midwives should have the same recognition and legal standing as AMCB credentialed midwives, nor that they should be able to practice independently (n = 82, 61.7%).

3.8 |. Future practice plans

Certified nurse-midwivess (n = 96, 69.1%) generally felt somewhat or very positive about the current state of midwifery practice in Texas, and for the future of midwifery in the U.S. (n = 130, 92.2%). If CNMs had their career to do over again, most would again choose nurse-midwifery (n = 115, 83.3%) and would recommend it as a career (n = 115, 83.3%).

When asked about practice plans for the next 3 to 5 years, more than half (n = 74, 53.9%) planned to change from current practice and hours, including a reduction in hours (n = 18, 12.8%), working part-time (n = 16, 11.3%), retirement (n = 20, 14.2%) or seeking employment in another practice, community, or outside midwifery altogether (n = 58, 41.8%). CNMs in more rural practice locations were more likely to plan to seek a nonclinical position within health care than those working in urban areas (p = .02). Benefits received were significantly related to plans to stay in current practice and hours (μ = 15.22, p < .001), reduce practice hours (μ = 17.61, p = .001), retire (μ = 7.8, p = .04), seek employment in a public facility (μ = 0.75, p = .01), find a nonclinical position outside health care (μ = 23.50, p = .06) and work full time (μ = 15.64, p = .008). CNMs who made US $50,000 or less did not differ from CNMs who made more than US $50,000 related to future plans (ps = 0.07–1.0). CNMs who made more than US $50,000 were more likely to have plans to volunteer midwifery care in the next 3–5 years (p = .03). Where change was planned, it was less likely for those who worked full-time than for those who worked part-time in midwifery (p = .01).

Most CNMs were satisfied with current midwifery practice (mean = 7.20, SD = 2.015, range 2–10); satisfaction did not significantly differ by age group (X²(4) = 10.122, p = .04). In general, all age groups had relatively high average satisfaction ranging from 6.57 (25–35 years)–8.17 (66–75 years) on a scale of 1–10 with a higher rating indicating greater satisfaction. Similarly, practice satisfaction did not differ for those working in rural areas or towns <10,000 population, those working in areas of 10,000–250,000 population and those working in areas >250,000 population (X²(2) = 2.458, p = .29). However, those working in areas with >250,000 population had the lowest average practice satisfaction (mean = 7.10) and those working in rural areas or towns with <10,000 population had the highest average practice satisfaction (mean = 8.25).

3.9 |. Qualitative data: Individual interviews

Eight CNMs were telephone interviewed; two were unavailable due to practice demands. Nurse-midwifery experience ranged from 1–35 years. All seven geographic regions of Texas were represented, with varied practice types and settings.

Themes related to practice success differed by practice location (i.e., out-of-hospital or in-hospital). CNMs working out-of-hospital reported increased patient satisfaction with greater ability for shared decision-making, a better sense of the “business of midwifery” and recognising the need for judicious use of resources. Those working in-hospital identified ancillary support for billing and credentialing, easier physician access for collaboration and appreciation for a range of provider philosophies, as factors promoting practice success.

Practice barriers noted by out-of-hospital CNMs included patients being unaware of care options, physician hostility and lack of support, empanelment restrictions by managed care organisations and issues related to regulation by the Texas Board of Nursing. CNMs working in-hospital identified administrative control, denial of hospital privileges, physician refusal to consultant, physician fear of competition for patients and poor knowledge of practice finance, as barriers. Midwives across settings and geographic areas identified public ignorance and misperceptions about the role of midwives—particularly differing types, as a concern. Key legislative challenges mirrored those identified on survey. See Table 1.

4 |. DISCUSSION

4.1 |. Demographics and employment characteristics

Nearly all respondents were female, typically older and, of note, was lack of racial diversity with more than 90% Caucasian, consistent with national data (Fullerton et al., 2015). Lack of racial diversity is concerning as almost two-thirds of respondents cared for women of colour. Lack of midwifery diversity is a barrier to people of colour, and increased numbers of midwives of colour are needed to reduce racial health inequalities (Guerra-Reyes & Hamilton, 2017). Despite U.S. efforts to improve racial diversity in midwifery, there has been relatively little progress (Dawley & Walsh, 2016). One potential solution is to consider models which integrate internationally educated midwives into the existing workforce, providing greater diversity and cultural sensitivity for the populations midwives serve (Tyson & Wilson-Mitchell, 2016).

A gender-skewed preponderance of females within midwifery was noted in this research and has also been noted nationally (Fullerton et al., 2015). Gender imbalance contributes to a lack of midwives in senior healthcare administration and policymaking roles (Dussault & Franceschini, 2006).

Most CNMs (86.5%) held a master’s degree, reflecting the fact that most nurse-midwifery education programmes in the U.S.A offer master’s degree preparation (ACNM, 2014). Few held a DNP—the clinical nursing doctorate (6.4%) or PhD (7.8%). CNMs with clinical doctorates are urgently needed to address the leadership and evidence-based practice challenges required to transform maternity care ([AACN] American Association of Colleges of Nursing, 2006), in concert with knowledge discovery by those with a research doctorate.

Respondents were largely employed by hospitals/medical centres, physician practices or an educational institution, consistent with national data (Fullerton et al., 2015). Of note, 13.5% were in solo or group midwifery practice. Low numbers in independent midwifery practice may reflect the modest autonomy granted Texas CNMs (Phillips, 2013). Limited prescriptive authority, inability to admit patients to hospital under the CNM name, lack of hospital staff membership commensurate with physicians and failure to have full voting privileges impacted full scope of practice and where respondents worked.

Approximately 20% of CNMs did not work in any midwifery capacity. Given the projected increased demand for women’s health services and the looming shortage of midwives, understanding the reasons for nonpractice is needed. Strategies are also needed to engage certified but nonpracticing midwives back into the workforce such as flexible work arrangements and workloads, accessible professional development (Pugh, Twigg, Martin, & Rai, 2013), and structured educational and healthcare agency programmes which would promote success on practice re-entry (McMurtrie, Cameron, Oluanaigh, & Osborne, 2014).

Typical CNM income was US $75,001 to US $100,000. Other advanced practice providers in the U.S.A—nurse practitioners and physician assistants, average $101,621 and US $112,344, respectively (Advance Healthcare Network, 2014). Given lifestyle differences with on-call demands and liability expenses, CNM earnings are worth noting. Additionally, there was relatively little salary increase with midwifery experience and few benefitted from profit-sharing—all potential incentives in the recruitment and retention of midwives. Finally, salaries did not differ by geographic location, a concern since 126 of 254 Texas counties have been designated as Health Provider Shortage Areas (U.S. Health Resources and Services Administration, 2014).

Most CNMs had liability insurance (77.9%). Having lability insurance, however, made no difference in geographic work location although nearly 13% reported that liability coverage had impacted the ability to practice. It has been well documented that risk assessment and assumptions of birth as abnormal affect how maternity services are governed and increases rates of intervention for low-risk women (Healy, Humphreys, & Kennedy, 2016). Health system restructure, focus on birth as normal and legislative efforts to cap medical malpractice awards are all important in curtailing liability costs for midwives as well as reducing unnecessary interventions performed when practicing so-called defensive medicine.

4.2 |. Nature and scope of practice

Births were attended by 81.9% of CNMs which is higher than the national average (64%), and most did so in the hospital setting (75.6%) although 30% attended in free-standing birth centres and 12.8% in the home—rates also much higher than noted nationally (Fullerton et al., 2015). These differences may reflect growing consumer desire for less intervention and for greater shared decision-making, or response bias. Birth in out-of-hospital settings with the midwifery model of care demonstrates higher rates of physiologic birth and lower rates of intervention without increase in adverse outcomes (Cheyney et al., 2014).

Certified nurse-midwives largely provided maternity and well-woman services; less than 10% of time was spent providing care to newborns or in primary care. Close to one-third were able to see greater numbers of patients. Restrictive regulatory policies and attempts by physician organisations to limit CNM practice, insurance restrictions, lack of consumer knowledge regarding midwifery services and confusion regarding the role of different types of midwives are likely factors. Strategies are needed to improve access to CNM care and determine reasons for underutilisation. Marketing campaigns designed to inform the public of the role of midwives, and study of patient perception of midwifery care, would be important adjuncts.

4.3 |. Care practices

Certified nurse-midwives providing intrapartum care largely felt that birth settings and consulting physicians were supportive of physiologic birth and that care practices which promoted physiologic birth (e.g., intermittent auscultation, birth in position of choice) were performed. However, fewer than half agreed that patients could eat/drink in labour as desired and more than one quarter agreed time constraints were imposed on labour. While a wide variety of nonpharmacologic and pharmacologic methods were provided, few CNMs offered measures which would further promote physiologic birth such as water birth, acupressure, nitrous oxide analgesia, sterile water injections and TENS. Such methods should be encouraged and practices such as food/fluid restrictions and use of CEFM reconsidered (ACNM, n.d.). Overall, findings underscored setting restriction and the need to reconsider both birth setting and the importance of the midwifery model of care. Further, the fact that one-third believed that the women they cared for did not desire physiologic birth suggests that birth is believed to be inherently risky. Midwives have an obligation to promote the normalcy of birth.

Approximately one-third of CNMs routinely offered elective induction as early as 39.0 weeks of gestation; close to 20% believed that it should be offered. Interestingly, the rate of labour induction by U.S. midwives (20.7%) nearly mirrors that of physicians (19%) (Martin, Kochanek, Strobino, Guyer, & MacDorman, 2005), and survey of U.S. women found that 44% indicated the provider tried to persuade them to have labour induction because the baby was full or close to term (Declercq, Sakala, Corry, Applebaum, & Herrlich, 2013). With Caesarean birth rates in the U.S.A at 32% (Martin et al., 2017), elective labour induction should be strongly discouraged given the association with operative birth. Further, while VBAC option was available for most midwifery patients, physician involvement or transfer of care was typically required, limiting patient choice. And, a quarter of CNMs felt that a trial of labour after Caesarean birth should not be offered. This later finding, coupled with the fact that 16.5% offered primary elective Caesarean birth and 8.6% believed that it should be offered, is disturbing and trends should be carefully tracked. Midwifery beliefs regarding practices which disrupt normal birth processes likely influence patient counselling, impacting cost and outcomes.

Overall, midwifery practices which do not promote physiologic birth were noted. While some practices may reflect the nature of physician consultation and hospital policies, further examination is warranted.

4.4 |. Primary care and advanced midwifery skills

One-third of respondents held primary care responsibilities although complexity of management was unknown. CNMs frequently provided care for patients with psychological, renal and endocrine issues; fewer than 10% reported never providing care across all conditions. While ACNM core competencies identify the need for midwives to apply knowledge and skills in primary care ([ACNM] American College of Nurse-Midwives, 2012), there is little direction regarding those conditions and management strategies. Further clarification of primary care expectations is needed.

A significant number of advanced practice skills were performed (e.g., basic ultrasound, manual removal of the placenta) and, given the frequency, these skills should be considered for inclusion in basic midwifery education programmes. Possessing these skills would both improve midwifery marketability and provide important access for underserved populations.

4.5 |. Teaching involvement

It was concerning that most CNMs did not teach students clinically. For those who did clinical teaching, it was most often nurse-midwifery students but for less than a quarter of CNM time. The reason most commonly cited for not precepting midwifery students was employer restrictions; financial compensation was not identified as a factor consistent with other studies (Germano, Schorn, Phillippi, & Schuiling, 2014; Hastings-Tolsma et al., 2015).

Student nurse-midwives are clinically taught primarily by CNMs in the workforce, and growing a new generation is heavily dependent upon their voluntary supervision of students, as well as the willingness of agencies to allow access for care. A shortage of preceptors, and a reluctance of agencies to allow students, has created a bottleneck preventing midwifery education programmes from accepting qualified applicants (Germano et al., 2014).

Clinical education of midwives is costly and labour intensive and has been recognised as a significant problem across the globe where resources are often scarce and new midwives are pressed into precepting students (Lori, Stalls, & Rominski, 2015). While greater numbers of midwives are needed, even more pressing are those who possess competencies consistent with advanced midwifery education. These challenges require significant investment into clinical midwifery education.

4.6 |. Geographic distribution and populations served

Large portions of Texas, particularly rural and small-town areas, were underserved by CNMs. Almost half of CNMs worked in large metropolitan areas; fewer than one-fifth worked in areas with populations ≤50,000. Such maldistribution between urban and rural areas is an important issue influencing health care. Adequate numbers of midwives and their appropriate distribution are critical as density of maternity care providers is associated with maternal and infant survival. In Texas, where maternal mortality rates are on the rise (MacDorman, Declercq, Cabral, & Morton, 2016), midwifery-delivered care would be key in reducing current trends.

The disparity between CNMs practicing in urban and more rural areas may reflect restrictions on autonomous practice (Kozhimannil et al., 2016). Further, most CNM training occurs in large urban areas—a practice identified as one of the most important factors driving practitioners away from rural areas (Ricketts, 2005). Student recruitment from rural areas, placement in carefully matched rural communities for clinical training, financial incentives, training opportunities for interprofessional education and strategies to support long-term employment are all crucial in decisions to practice in such areas (RHI Hub, 2017). Models with demonstrated success in placement of students in smaller rural communities (Smith, Lloyd, Lobzin, Bartel, & Medlicott, 2015) should be considered, along with strategies which attract midwives and encourage retention.

Of note, CNMs who worked in rural areas (population ≤50,000) were more likely to consult with family practice physicians. More than half of these physicians routinely work with CNMs, nurse practitioners or physician assistants—particularly those working in rural areas (Peterson, Phillips, Puffer, Bazemore, & Petterson, 2013). However, there is a steeply declining percentage of family practice physicians providing maternity care (Tong et al., 2013) with significant repercussion for CNMs in need of collaborative practice arrangements. Creative strategies for CNM and physician collaboration are needed.

Finally, although there is no agreed upon definition of “risk,” respondents typically provided care to patients deemed to be of low-moderate to moderate risk. Texas CNMs provide care to large numbers of minority and underserved patients as well as those enrolled in Medicaid; high in-migration of both international and domestic patients (Office of the State Demographer, 2016) and high poverty rates in Texas (15.6%) fuel these numbers (Center for American Progress, 2018). Findings likely reflect socio-economic and psychosocial risk factors often noted in vulnerable and underserved populations.

4.7 |. Legislative issues and practice barriers

Nurse-midwives use less technology yet provide safe, effective care when compared to physicians (Newhouse et al., 2011). Despite these facts, significant barriers exist for midwifery practice in Texas. Since 2000, the professional practice environment of CNMs in the U.S.A has been regularly examined regarding legal status, reimbursement and prescriptive authority. Texas scores increased 21.3% from 2000–2015 with smallest gains made in legal status and prescriptive authority; greatest increase was in reimbursement (Beal et al., 2015) demonstrating but modest improvement in midwifery practice autonomy.

Texas regulations for CNMs allow for collaborative rather than autonomous practice, mandating contracts with physicians. Similarly, the ability to prescribe is under regulatory supervision of a physician. Findings here demonstrate that these restrictions create barriers for CNMs in providing services where there is lack of medical care, as well as preventing practice commensurate with education. In states with regulations which support autonomous midwifery practice, there is a larger CNM workforce and a greater proportion of CNM-attended births (Yang, Attanasio, & Kozhimannil, 2016). Strategies which mitigate practice barriers for midwives are needed.

4.8 |. Future practice plans

High job satisfaction levels were reported overall though more than half planned to change current practice and hours in the next 3 to 5 years. While partially explained by respondent age and retirement, the attrition rate is alarming and may indicate employment-related problems such as overwork, noncompetitive pay and supervisory restrictions (Castro Lopes, Guerra-Arias, Buchan, Pozo-Martin, & Nove, 2017). This study did not examine reasons for attrition plans, although physical demands and poorer mental well-being have been cited as reasons for intention to leave employment (Perry et al., 2017) and may have been influential in findings here.

Midwifery is a high stress profession with significant burnout (Perry et al., 2017). Research is needed to delineate factors contributing to dissatisfaction with midwifery as the antecedents to burnout and work engagement are primarily characteristics of the practice environment not personal factors (Bakker, Demerouti, & Sanz-Vergel, 2014). Workload, autonomy and decision latitude are central components in creating a positive practice environment (Van Bogaert, Kowalski, Weeks, & Clarke, 2013).

5 |. STUDY LIMITATIONS

This research had several limitations. The response rate of 32% raises questions about bias and generalisability. Reduction in survey length, inclusion of a completion bar and reconsideration of participant compensation may be useful to increase response rates on future survey. Further, there was no follow-up of nonrespondents although email reminders were sent. The differences between those who responded and those who did not are unknown although it is likely that time limitations along with survey burden, lack of interest and a desire for privacy may have been factors.

6 |. CONCLUSION

Midwifery, generally poorly understood and often at the bottom of the healthcare system ladder ([WHO] World Health Organization, 2013a, [WHO] World Health Organization, 2013b), has a projected shortfall. Understanding the midwifery workforce is crucial in planning future development of the pipeline—as well as in the development of retention strategies, and is a global midwifery research priority (Soltani, Kane Low, Duxbury, & Schuiling, 2016). As the largest group of maternity care providers across the globe, midwives are needed to provide leadership at point of care, as well as on interprofessional teams and in health policy positions (Thompson, 2016). Understanding the nature and scope of midwifery practice can have a substantial impact on access to health care, particularly for vulnerable and underserved populations (Dawson, Nkowane, & Whelan, 2015; Sonenberg, 2010).

There is clear need to examine whether health systems across the globe are providing effective coverage for women (United Nations Population Fund, 2014a). Detailing the existing healthcare workforce is essential in the development of strategies to ensure that women can receive the care they want and need. Where there are restrictions to midwifery care, the need to reconsider health policies which promote supply and access is compelling and will require significant public investment (Day-Stirk & Fauveau, 2012). Nations looking to conduct midwifery workforce studies may find the work reported here useful in guiding such survey.

6.1 |. Relevance to Practice

Workforce is a global issue, and midwifery visibility needs to be strengthened at the WHO level (Wong et al., 2015), as well as within the healthcare policymaking bodies of individual countries. It is crucial to engage midwives in the establishment of healthcare policies and resource distribution which allow for effective delivery of care (Lori et al., 2015; Sanders, Hunter, & Warren, 2016) and allow practice consistent with education. Professional midwifery associations can play a key role through more active involvement in workforce planning and decision-making processes, although globally, participation has tended to be limited (Lopes, Titulaer, Bokosi, Homer, & ten Hoope-Bender, 2015). Similarly, education and health-care institutions need to make a concerted effort to promote the upscaling of midwives demonstrating an equitable standing with other health team members and integrated inclusion within the health system (Renfrew et al., 2014). Policies emphasising the midwifery model of care as standard of practice for low-risk women are crucial in meeting the United Nations Sustainable Development Goals.

What does this paper contribute to the wider global clinical community?

• Midwifery workforce studies are lacking globally with most surveys examining midwives within the broader rubric of “advanced practice nurses” or “nurses” in general. A unified midwifery oversight board—distinct from medicine or nursing, is needed to track the availability and accessibility of midwives for workforce planning.

• Workforce data from this study confirm that nurse-midwives provide disproportionate care to vulnerable and underserved populations but are underutilised with restrictive practice policies that are unwarranted given educational preparation.

• Nurse-midwives provide care that is comparable to physicians, yet lack equitable pay and recognition, and are largely excluded from decisions regarding health policies. To improve midwifery service access and coverage, midwives with skill in affecting legislation are urgently needed to influence health policy changes.