A Comparative Workforce Study of Midwives Practicing in the State of Texas

Introduction:

Access to quality care is a problem in Texas, an ethnically diverse state with large birth numbers. The state has over 300 areas designated as medically underserved, and a severe lack of obstetricians and midwives. Minimal data exist on midwifery’s contribution, and no known study compares the work environment and clinical practice of the 2 state-recognized midwifery paths, licensed midwives (LMs) and certified nurse-midwives (CNMs). The purpose of this study was to determine the differences in practice by CNMs and LMs, the latter of whom are generally certified professional midwives. The specific aims were to 1) describe the differences in demographic and employment characteristics of CNMs and LMs, 2) identify the geographic areas and population groups served by CNMs and LMs, and 3) compare the nature and scope of CNM and LM clinical practices.

Methods:

Online parallel surveys of Texas LMs and CNMs were conducted in December 2015 and January 2016. The REDCap data management system housed the 123- and 125-item surveys for LMs and CNMs, respectively, addressing demographics, populations served, and clinical practice. A comparative statistical analysis, using Fisher’s exact test, Pearson’s chi-squared test, and Independent Samples t-tests, was performed.

Results:

The survey response rates of LMs and CNMs were 35.4% (n = 75) and 31.9% (n = 143), respectively. Differences in demographics, employment status, workload, scope of practice, risk assessment, time-based care management, and technology use were observed.

Discussion:

Findings represent the first attempt to describe the Texas midwifery workforce. In a large state with health care provider shortages, this step is pivotal in addressing strategies for providing services for women and infants. This groundwork can provide the foundation for including midwifery in a state health plan.

INTRODUCTION

In Texas, there are 316 designated medically underserved areas, 147 counties without an obstetrician, and an estimated deficit of over 600 certified nurse-midwives (CNMs).^1–3 Midwives can fill needs in these areas, particularly in underserved areas where women lack access to basic reproductive, gynecologic, and preventive care as well as birth options with decreased medicalization. The Texas Board of Nursing regulates CNMs as advanced practice registered nurses, and the Texas Department of Licensing and Regulation oversees all non-nurse midwives. No certified midwives currently practice in Texas. Table 1 differentiates the practice status of Texas midwives. Workforce studies are the first step to determining provider availability within a state.

Table 1.

Midwifery Practice in Texas

Descriptor	Certified Nurse-Midwives	Licensed Midwives
State licensure	Advanced practice registered nursea	Licensed midwifeb
National certification	Certified nurse-midwifec	CPR; NRP; certified professional midwife credential is not requiredb
Accreditation and education	Master’s degree or beyond from an accredited midwifery programc	Completion of a state-approved midwifery program or a basic midwifery education course accredited by MEAC; successful completion of the NARM examination, jurisprudence examination, and training in collection of newborn screening specimensb
Regulatory and advisory agencies	Texas Board of Nursinga (APRN Advisory Committee)	Texas Department of Licensing and Regulationb (Midwifery Advisory Board)
Scope of practice	According to the specialty’s professional organization, primary health care for women from adolescence to beyond menopause, including nonreproductive primary care; gynecologic and family planning services; preconception care; care during pregnancy, childbirth, and postpartum; and newborn care through the first 28 days of lifec	Providing necessary supervision, care, and advice to a woman during normal pregnancy, labor, and the 6-week postpartum period; conducting a normal birth of a child; and providing normal newborn care through the first 6 weeks of lifed
Prescriptive authority	Delegated medical act that includes Schedule II drugs and requires a cosigned Prescriptive Authority Agreement between the physician and CNM, with periodic face-to-face meetingse	Prohibited from administering prescription drugs other than the following: oxygen, prophylaxis for prevention of ophthalmia neonatorum; and drugs that are under the supervision of a licensed physician. The actual prescribing of drugs is not addressed.b
National professional bodies	ACNM, AMCB, ACME	MANA, NARM, MEAC
State professional organizations	Consortium of Texas Certified Nurse-Midwives (ie, the Texas affiliate of ACNM)f	Association of Texas Midwivesd
Essential and standard-setting documents that frame practice	Texas Nurse Practice Act; ACNM Definition and Scope of Practice, Standards of Practice, Core Competencies for Basic Midwifery Practice, and Code of Ethics; professional resourcesc	Texas Midwifery Act; Texas Midwifery Board Instructor Manual, MANA Core Competencies for Basic Midwifery Practice Standards of Practice, and Statement of Values and Ethicsg

Abbreviations: ACME, Accreditation Commission for Midwifery Education; ACNM, American College of Nurse-Midwives; AMCB, American Midwifery Certification Board; APRN, advanced practice registered nurse; CNM, certified nurse-midwife; CPR, cardiopulmonary resuscitation; MANA, Midwives Alliance of North America; MEAC, Midwifery Education Accreditation Council; NARM, North American Registry of Midwives; NRP, neonatal resuscitation program.

^aSource: 22 Tex. Admin. Code § 222.²⁰

^bSource: Texas Department of Licensing and Regulation.⁸

^cSource: ACNM.²¹

^dSource: Association of Texas Midwives.²²

^eSource: Texas Board of Nursing.¹⁷

^fConsortium of Texas Certified Nurse-Midwives.¹⁶

^gMANA.²³

The Texas Health Professions Resource Center compiles data on its health professionals by county, including licensed midwives (LMs) and CNMs, with CNMs also tracked through the Texas Center for Nursing Workforce Studies. Nonetheless, no comprehensive description and analysis of the employment characteristics, nature, and scope of clinical practice among the midwifery workforces exists. To ensure that Texas women have the option of midwifery care, it was important to examine the care provided by all midwives, the populations cared for, and where gaps in health care occur. The purpose of this study was to determine the differences in practice settings and care provided by CNMs and LMs. The specific aims of this secondary analysis were to 1) describe the differences in demographic and employment characteristics of CNMs and LMs, 2) identify the geographic areas and population groups served by CNMs and LMs, and 3) compare the nature and scope of CNM and LM clinical practices.

METHODS

Design

This secondary comparative analysis was designed to compare findings from parallel online surveys of LMs and CNMs licensed in Texas. One survey focused exclusively on CNMs; the other survey focused exclusively on LMs. The Colorado Multiple Institutional Review Board (#15–1942) and the Baylor University Institutional Review Board (#717761–1) approved the study.

Sample

All legally practicing midwives in Texas were invited to participate. Midwives were initially notified of the upcoming survey at their respective state professional meetings, then by individual email invitations explaining the purpose, format, dates, risks, financial support, and privacy. Prospective participants were also informed that completing the survey implied consent and, as an incentive, that they could enter a drawing for a professional meeting tuition. Surveys were then emailed to each midwife.

CNM Survey

CNMs certified through the American Midwifery Certification Board (AMCB) and listing a Texas address in the AMCB database (N = 449) in December 2015 were identified as potential participants. Prior to initiation of the study, the AMCB Research Committee approved the study and use of its database.

LM Survey

LMs licensed through the Texas Department of Licensing and Regulation (N = 212) in December 2015 were identified as potential participants. Because there was no governmental database of LM email addresses, the email list maintained by the Association of Texas Midwives, the state’s professional LM organization, was used.

Measures and Instrument

The survey designs were based on an existing workforce survey, developed and validated for use with CNMs in another state.⁴ After the receipt of permission for use and adaptation of the original survey, a diverse research team representing LMs and CNMs revamped each survey, customizing them to address specific issues presented by each midwifery group. Each survey item was reviewed and revised for relevance and applicability to the intended group’s nature of practice and workforce data needs. Questions consisted of statements with response options provided. In some cases, respondents could select multiple responses. Likert scale responses were on a 5-point scale, from “strongly agree” to “strongly disagree.” This resulted in 2 separate but comparable versions, one for LMs and one for CNMs. Both focused on demographic, employment, and practice characteristics.

Because of the subjectivity and controversy surrounding terms such as risk and authority, the research team decided not to define these terms in the surveys. The intent was to gather information regarding each group’s perception of what these terms meant. In addition, the researchers wanted to determine how respondents interpreted their legal scope of practice.

Pilot

The research team solicited CNMs to pilot the CNM survey instrument. Five CNMs practicing in various state regions completed the pilot in September 2015, with subsequent revision of 4 survey items. The final CNM survey instrument included a total of 125 items.

The research team also sought LMs for piloting the LM survey instrument. Four LMs practicing in various state regions completed the pilot in October 2015, with subsequent revision based on their feedback. The final LM survey instrument included a total of 123 items.

Data Collection

The surveys were conducted between December 7, 2015, and January 26, 2016. Midwives were asked to participate in a survey to learn more about the nature of midwifery practice in Texas. The study project manager handled all distribution, technical difficulties, and follow-up. Additional reminder emails were sent at the survey period midpoint and during the final survey week. Follow-up of nonresponders was not done. Copies of the surveys are available from the corresponding author.

CNM Survey

AMCB maintains email addresses for all CNMs. Using this list, AMCB distributed an email invitation with a survey link to all CNMs. AMCB also sent subsequent reminder emails.

LM Survey

An email invitation with a survey link was distributed by the Association of Texas Midwives to all LMs with an email address in its database. The distribution list was forwarded to the study project manager for subsequent reminder emails.

Data Management and Analysis

Study data were collected and managed using REDCap electronic data capture tools hosted at the University of Colorado Anschutz Medical Campus. REDCap (Research Electronic Data Capture) is a secure, web-based application designed to support data capture for research studies, providing 1) an intuitive interface for validated data entry, 2) audit trails for tracking data manipulation and export procedures, 3) automated export procedures for seamless data downloads to common statistical packages, and 4) procedures for importing data from external sources.⁵

Of the 125 CNM survey items and the 123 LM survey items, a subset of variables was selected for the current analysis. To combine the variables across surveys, descriptive statistics (ie, frequencies) for all variables under consideration were run separately for each survey. Each variable’s response options for both surveys were then compared and categorized into 4 types. In the first type, the questions and response options on both the CNM and LM survey were an exact match and required no transformation or recoding for comparative analysis; 65 of the variables fell into this category. In the second type, the questions and response options differed slightly between the CNM and LM surveys but did not require any transformation for comparative analysis. For example, a CNM survey question referred to “my clinical practice” and the LM survey question referred to “my practice.” Fourteen variables fell into this category. In the third type, the questions and response options between the 2 surveys differed and had to be transformed to be comparable when the 2 surveys were combined for analysis. As an example, the LM surveys asked whether LMs taught certified professional midwifery students, certified nurse-midwifery students, or did not teach students. The CNM survey added a fourth response option of certified midwifery students. Variable transformation included adding the fourth option to the LM survey so that responses from both surveys could be analyzed together. Eighteen variables fell into this category. Finally, there were noncomparable questions. For example, the LM survey included a question that asked “If a woman is carrying twins, I am able to manage independently” with response options of “yes” or “no.” No comparable question existed on the CNM survey. Fourteen variables fell into this category. No data were deleted during this variable transformation; after completion of the process, there were 106 variables deemed appropriate for use in this analysis.

The 2 midwifery groups were compared on each common data point in the surveys, using IBM-SPSS, version 21 (IBM Corporation, Armonk, NY). Comparative analyses included Fisher’s exact test, Pearson’s chi-squared test, and Independent Samples t-tests.

RESULTS

A total of 218 midwives completed the surveys. After the removal of 2 student respondents, 141 CNMs (31.9% response) and 75 LMs (35.4% response) participated.

Demographic and Employment Characteristics

The midwives were asked personal demographics of age, gender, race or ethnicity, education, languages spoken, and number of years certified and licensed. Employment questions included employment status, salary, workload, practice setting, and hospital privileges (Table 2). Overall, LMs were predominately white, female English speakers aged 36 to 45 years. The CNMs were predominately white or African American or black women, and nearly half (45.4%) self-identified as Spanish and English speakers. From an employment perspective, CNMs had been licensed for an average of 4 years longer, earned higher salaries, were more frequently employed in hospitals or medical centers, more frequently attended births in hospitals (75.6% vs 4.2%; χ²(1) = 81.53, P < .001), had a larger monthly birth volume (11–100 births; 85.9% vs 13.9%; χ²(1) = 81.08, P < .001), and had more hospital admitting privileges (30% vs 5.9%; χ²(1) = 14.29, P < .001). In contrast, LMs were more likely than CNMs to be self-employed (Table 2) and attended more births in freestanding birth centers (59.7% vs 30.2%; χ²(1) = 13.85, P < .001) and homes (95.8% vs 12.8%; χ²(1) = 108.12, P < .001).

Table 2.

Demographics and Employment Characteristics of Midwives in Texas

Characteristics	Licensed Midwives	Certified Nurse-Midwives	P Value
Gender, n (%)
Female	75 (100)	139 (98.6)
Prefer not to answer	0 (0)	2 (1.4)
Age, n (%), y			<.05b
25–35	16 (21.3)	17 (12.1)
36–45	26 (34.7)	34 (24.1)
46–55	20 (26.7)	35 (24.8)
56–65	10 (13.3)	41 (29.1)
66–75	3 (4.0)	11 (7.8)
Over 75	0 (0)	2 (1.4)
Not reported or unknown	0 (0)	1 (0.7)
Ethnicity, n (%)			.08b
White	74 (98.7)	126 (89.4)
African American or black	0 (0)	8 (5.7)
Asian or Pacific Islander	0 (0)	1 (0.7)
Hispanic or Latino	0 (0)	1 (0.7)
Other	0 (0)	1 (0.7)
Not reported or unknown	1 (1.3)	4 (2.8)
Other languages spoken by the midwife, n (%)a,e
Spanish	9 (12.0)	64 (45.4)	<.001c
German	2 (2.7)	1 (0.7)	.28b
French	1 (1.3)	4 (2.8)	.67b
Other	5 (6.7)	3 (2.1)	.13b
Initial Texas midwifery licensure, mean (range), y	2005.19 (1981–2015)	2001.0 (1971–2015)	<.01d
Primary employment, n (%)a			<.001c
Hospital or medical center	1 (1.7)	30 (61.2)
Midwifery group	3 (5)	12 (24.5)
Self-employed	56 (93.3)	7 (14.3)
Income, n (%)			<.001c
Don’t know	19 (25.3)	0 (0)
≤$30,000	16 (21.3)	5 (3.5)
$30,001–$50,000	25 (33.3)	3 (2.1)
$50,001–$100,000	11 (14.7)	63 (44.7)
>$100,000	1 (1.3)	40 (28.4)
Not applicable, not doing clinical midwifery	0 (0)	27 (19.1)
Not reported or unknown	3 (4.0)	3 (2.1)

^aPercent of those who responded.

^bP values are from Fisher’s exact test.

^cP values are from Pearson’s chi-squared test.

^dP values are from Independent Samples t-test.

^eAll participants were English speakers.

Geographic Areas and Populations Served

No significant difference was found between the geographic populations served, with most midwives practicing in large cities and few in rural areas and small towns. However, populations seeking midwifery care reflected the state’s ethnic diversity.⁶ LMs reported serving a majority white population, yet they also served Hispanic and African American or black populations. In contrast, the CNMs reported serving a smaller percentage of white clientele and larger percentages of Hispanic and African American or black clientele (Table 3). In terms of service payment, the following forms of compensation were received: self-pay, Medicaid, commercial insurance, military insurance, and uncompensated care. Women who were self-pay formed the largest segment of the LM client population, whereas women with Medicaid and commercial insurance composed the majority of the CNM client population (Table 3).

Table 3.

Geographic Areas of Practice and Client Ethnicities of Texas Midwives

Characteristicsa	Licensed Midwives	Certified Nurse-Midwives	P Value
Geographic area, n (%)			.36c
Rural	4 (5.3)	5 (3.5)
Town (population 10,000)	3 (4.0)	3 (2.1)
City (population 10,001–50,000)	12 (16.0)	13 (9.2)
City (population 50,001–100,000)	7 (9.3)	24 (17.0)
City (population 100,001–250,000)	8 (10.7)	14 (9.9)
City (population >250,000)	39 (52)	53 (37.6)
Client ethnicity, n (%)b
White	72 (96)	72 (51.1)	<.001d
African American or black	9 (12)	39 (27.7)	<.01d
Hispanic, Latino, or Spanish	26 (34.7)	74 (52.5)	<.05d
Asian or Pacific Islander	1 (1.3)	14 (9.9)	<.05d
American Indian or Alaskan Native	1 (1.3)	6 (4.3)	.43c
Other	2 (2.7)	3 (2.1)	.999c
Not applicable, not in midwifery clinical practice	2 (2.7)	27 (19.1)	<.001d
Clients’ payer source, mean (SD)b
Self-pay	72.3 (26.4)	15.6 (21.7)	<.001e
Commercial insurance	22.7 (22.8)	31.4 (28.5)	<.05e
Medicaid or Medicare	2.3 (7)	45.1 (36.2)	<.001e
Tricare	0.3 (0.95)	8.6 (23.9)	<.01e
Indigent	1.1 (2.2)	3.7 (13.3)	.09e

^aPercent of those who responded.

^bMultiple options could be selected.

^cP values are from Fisher’s exact test.

^dP values are from Pearson’s chi-squared test.

^eP values are from Independent Samples t-test.

Clinical Practice

Questions were asked related to scope of practice, specific clinical practices, and external environmental processes that affect clinical practice. Midwives reported several clinical practice variations.

Scope of Practice

The respondents differed significantly in their primary area of clinical responsibility. LMs more frequently indicated antepartum, newborn, and primary care. The CNMs reported providing gynecologic care more frequently (Table 4). When asked about the perceived risk level of their clients, CNMs and LMs differed in the amount of care they reported providing to self-defined low-, low-moderate-, and moderate-risk women. LMs reported more low-risk care (63% vs 15.5%), less low-moderate-risk care (32.9% vs 40.8%), and less moderate-risk care (0% vs 35.9%). Finally, more CNMs possessed prescriptive authority (93.2% vs 14.1%; χ²(1) = 110.51, P < .001).

Table 4.

Scope of Practice, Specific Clinical Practices, and External Environmental Processes that Affect Clinical Practice for Texas Midwives

Question Responses	Licensed Midwives n (%)	Certified Nurse-Midwives n (%%a	P Value
My primary responsibilities areb
Antepartum care	64 (85.3)	90 (63.8)	.001c
Intrapartum care	65 (86.7)	86 (61)	<.001c
Newborn care	64 (85.3)	25 (17.7)	<.001c
Primary care	44 (58.7)	46 (32.6)	<.001c
Gynecology	21 (28)	73 (51.8)	.001c
Clinical midwifery education	35 (46.7)	26(18.4)	<.001c
Didactic midwifery education	16(21.3)	5 (3.5)	<.001c
Education, other	2 (2.7)	18 (12.8)	<.05c
Administration	35 (46.7)	15 (10.6)	<.001c
Research	11 (14.7)	7(5)	<.05c
Services provided in my midwifery practiceb
Continuous labor support	72 (96)	53 (50.5)	<.001c
Waterbirths	71 (98.6)	31 (35.2)	<.001c
VBAC births	46 (65.7)	34 (44.2)	<.05c
Continuous EFM	1(1.3)	43 (41)	<.001c
I have skiils to d<b
External breech version	38 (54.3)	11 (10.8)	<.001c
Manual removal of placenta	25 (35.2)	68 (66)	<.001c
Limited obstetric ultrasound	19 (26.8)	71 (68.3)	<.001c
Standard ultrasound	9 (12.7)	8 (7.8)	.31c
3rd-degree laceration repair	26 (35.6)	47 (45.6)	.22c
4th-degree laceration repair	3(4.1)	7 (6.9)	.52d
Artificial insemination	4(5.5)	9 (8.9)	.56c
There should be a time constraint on a woman’s labor (time in labor)			<.001c
Strongly disagree	30 (42.3)	19 (21.3)
Disagree	28 (39.4)	31 (34.8)
Somewhat agree	11 (15.5)	15 (16.9)
Agree	1 (1.4)	12 (13.5)
Strongly agree	1 (1.4)	12 (13.5)
In my midwifery practice, I routinely offer postdates IOL, weeks’ gestation			<.001c
39–41	27 (45.8)	75 (77.3)
42 or greater	32 (54.2)	22 (22.7)

Abbreviations: EFM, electronic fetal monitoring; IOL, induction of labor; OB, obstetric; VBAC, vaginal birth after cesarean.

^aPercent of those who responded.

^bMultiple options could be selected.

^cP values are from Pearson’s chi-squared test.

^dP values values are from Fisher’s exact test.

Specific Clinical Practices

CNMs were asked about gynecologic and primary care procedural skills (eg, endometrial biopsy, punch biopsy, gynecologic ultrasound), whereas LMs were asked about management of gynecologic or primary care conditions. These differences prevented comparative analysis. However, both were asked about performing prenatal ultrasounds, external breech versions, circumcisions, manual removals of placentas, and repairs of third- and fourth-degree lacerations. The CNMs were more likely to respond that they possessed the skills of manual placenta removal and limited prenatal ultrasound. Both reported similar rates of repair of third-and fourth-degree lacerations and artificial insemination. The only procedure that no LM performed was circumcision, which 8 CNM participants performed (Table 4).

Comparison of specific intrapartum skills was limited because of survey wording differences. For example, LMs were asked how they managed women presenting with twins and breech presentation. CNMs were asked if they cared for women considered to be “high risk (eg, twins).” Both response options included 1) independently, 2) collaboratively with a physician, 3) unable to manage—must refer, 4) don’t know, and5) not applicable. Most LMs reported independently managing women in labor with twins and with breech presentation (56% and 61.3%, respectively). Each group was asked if they performed external breech versions. Significantly more CNMs than LMs responded affirmatively (Table 4).

Subsequent questions focused on labor management (eg, physiologic birth, shared decision making, cultural needs, nonpharmacologic labor support, technology use). Both groups strongly agreed that they supported physiologic birth (CNMs 51.6%, LMs 97.1%) and shared decision making (CNMs 53.2%, LMs 100%). Both groups believed that they had the knowledge, skills, and confidence to support women desiring physiologic labor and birth. In practice, LMs provided more continuous labor support and attended more waterbirths and vaginal births after cesareans (VBACs). In contrast, CNMs were more likely to use continuous electronic fetal monitoring. Final questions focused on time-sensitive issues surrounding the length of labor and induction of labor. The LMs were more likely to strongly disagree that there should be constraints on a woman’s time in labor and were more likely to wait until 42 weeks’ gestation or longer before recommending labor induction (Table 4).

External Environment

In this final section, the midwives were asked about environmental impacts on care (eg, practice management, physician consultation arrangements, student teaching, liability insurance procurement, regulatory board complaints, malpractice lawsuits). More CNMs had a designated midwifery practice director, but more than half of LMs were in solo practice (Table 5). Both groups consulted with physicians; however, CNMs were more likely to believe that midwives should have a practice agreement with a physician. All midwives incorporated student teaching into their work. Sixty LM respondents and 63 CNM respondents reported teaching students in their own professions (80% and 44.7%, respectively). This trend reversed with teaching students in other professions, with CNMs teaching more nonmidwifery health care providers (Table 5).

Table 5.

Physician Consultation, Prescriptive Authority, and Liability Insurance AmongTexas Midwives

Characteristics	Licensed Midwives n (%;a	Certified Nurse-Midwives n (%%a	P Value
Designated midwifery practice director			<.001c
Yes	26(36.1)	58 (56.3)
No	8(11.1)	31 (30.1)
Not applicable, in solo practice	37(51.4)	5 (4.9)
Other	1 (1.4)	9 (8.7)
My physician consultant is, yesb
Employed by the same organization	1 (2.6)	48 (52.2)	<.001d
External to my practice	37 (97.4)	25 (27.2)	<.001d
Should have practice agreement with physician to practice			<.01d
Yes	9 (13.0)	35 (34.3)
No	51 (74.0)	54 (52.9)
No opinion	9 (13.0)	13 (12.7)
Teach midwifery students, yesb
Certified midwifery	60 (80)	7 (5.0)	<.001d
Certified nurse-midwifery	5 (6.7)	63 (44.7)	<.001d
Certified professional midwifery	6 (8.0)	9 (6.4)	.78d
Teach other students, yesb
Women’s health nurse practitioner	6 (8.0)	41 (29.1)	<.001d
Family nurse practitioner	5 (6.7)	49 (34.8)	<.001d
EMTs and/or paramedics	2 (2.7)	16(11.3)	<.05d
Basic RN nursing students	7(9.3)	26(18.4)	.11d
Medical students	2 (2.7)	26(18.4)	.001d
Resident physicians	1(1.3)	23 (16.3)	.001d
Physician assistant	0 (0)	12 (8.5)	NA
Other	4(5.3)	5 (3.5)	.72c
Carry professional liability insurance			<.001d
Yes	6 (8.6)	81 (77.9)
No	64 (91.4)	23 (22.1)
Defendant for malpractice			.17d
Yes	3 (4.6)	12 (11.5)
No	62 (95.4)	92 (88.5)
Texas-sanctioned ability to practice			.65c
Yes	1(1.5)	4(3.8)
No	64 (98.5)	100 (96.2)

Abbreviations: EMT, emergency medical technician; NA, statistical test not preformed; RN, registered nurse.

^aPercent of those who responded.

^bMultiple options could be selected.

^cP values are from Fisher’s exact test

^dP values are from Pearson’s chi-squared test.

Professional liability insurance coverage significantly varied, with few LM subscribers. All LMs responded that their liability insurance situation did not affect their ability to practice, whereas 12.5% of CNMs responded that it did. When asked if they had been named as a defendant in a malpractice case, 4.6% of LMs and 11.5% of CNMs acknowledged involvement (Table 5). Disciplinary investigations after a filed regulatory board complaint revealed low rates for CNMs and LMs (3.8% and 1.5%, respectively) and no license suspensions or revocations.

DISCUSSION

This is the first analysis of the demographics and practice of 2 distinct groups of midwives in Texas, even though Texas midwifery precedes the state’s annexation.⁷ Indigenous American Indian and Mexican women gave birth with midwives, as did Anglo and African American women through the 19th century. Similar to national statistics, by the early 1970s, only 2.3% of Texas births were attended by midwives, with approximately two-thirds attributed to Hispanic midwives (parteras).⁷ The same decade brought midwifery practice changes, such as the arrival of the first certified nurse-midwife (CNM), the reappearance of empirically trained non-Hispanic midwives,⁷ and the regulation of CNMs as advanced practice nurses under the Board of Nursing.⁸ In 1984, the state legislature passed the first lay-midwife regulations.⁷ Today, all non-nurse midwives in Texas are regulated by the Texas Department of Licensing and Regulation as LMs.⁸ This historical evolution led to dual midwifery workforces, underscoring the need to profile the characteristics of both groups.

The demographic and employment characteristics in this analysis parallel the statistics presented in the Health Professions Resources Center and the Texas Center for Nursing Workforce.^3,9 The CNMs more closely reflected the ethnic and linguistic preferences of the state’s population; however, the percentage of African-American or black and Hispanic CNMs did not match the proportion of these ethnicities in the population at large.⁶

The geographic distribution and populations served by midwives are noteworthy. Neither group tended to practice in underserved rural areas or small towns. However, the CNMs’ client populations and payer sources suggest care for those who are medically underserved in urban areas. Texas CNMs cited hospital bylaws and the prescriptive authority requirement for periodic face-to-face meetings with the collaborating physician as a deterrent to practicing in areas without a strong physician network.¹⁰ A state-sponsored initiative could facilitate midwifery services in rural areas.

Although the results cannot be generalized to all midwives, the data demonstrated that CNMs work predominantly in medical practices with hospital privileges, suggesting that they would be able to meet provider workforce demands in rural and safety-net hospital systems. It is beyond this analysis to discuss future health care models; however, the home birth and birth center settings of the LMs are low-resource options for exploration. With nearly 70 licensed birth centers in Texas (primarily operated by LMs),¹¹ the state is well positioned for integration of this care model into workforce strategies.

The lower-volume work environment reported among surveyed LMs is congruent with the role description on LM professional websites.¹² The North American Registry of Midwives states that LMs “carry a relatively low client (average 3–6 births per month) load which allows for more personalized and comprehensive care than typical obstetrical practices.”¹² The higher salaries reported by CNM respondents also match their higher volume work environments. Furthermore, the costs associated with hospital credentialing requirements and insurance panel membership require increased revenues; however, cost analysis was beyond the scope of this comparison. In addition, reasons for selecting self-employment were not addressed in either survey. Future qualitative surveys exploring entrepreneurial interests, preferences for personal control over one’s practice, and a willingness to assume more financial risks would provide a better understanding of the differences between these 2 groups’ work practices.

When asked about clinical practice, more LMs responded that they spent most of their time on maternity-related care. However, the broader scope of CNM practice may dilute the percentage of time spent on each type of service. The finding that LMs see fewer clients may also account for the differences here. Further analysis is needed to determine if the 2 groups actually spend different amounts of time providing a particular service, or if the variance is associated with scope and workload.

In soliciting information about pharmacologic practices, the research questions were about the connotative use of prescriptive authority among the state’s midwives. Legally, delegated prescriptive authority is available to CNMs, with LMs authorized to administer selected prescription medications (eg, neonatal ophthalmic prophylaxis) through standing physician orders.⁸ However, previous unclear regulatory wording caused confusion. LMs were authorized to procure oxygen and ophthalmic ointment. Because both require prescriptions, some LMs believed they had independent prescriptive authority and that CNMs had delegated prescriptive authority. Two years ago, the regulations were clarified, stating that it is the administering of these pharmacotherapeutics under standing physician orders that is within the scope of LM practice.⁸ By maintaining the term prescriptive authority on the LM survey, it was possible to determine if LMs understood the difference between prescribing and administering. The LMs’ low response rate for having prescriptive authority (14.1%) may reflect understanding of the current regulations. Because there are LMs who possess prescriptive authority through other licenses, a study examining role delineation in medication management is needed.

Specific clinical practices varied the most. The finding that more LMs than CNMs were in solo practices could be associated with more influence and control over clinical care policies. Asking questions about the amount of influence each had over clinical care decisions would have enhanced this information, as clinical site and practice group policies influence care.¹³ Some policy questions were asked (eg, “I am able to care for women who desire a trial of labor after a cesarean”); however, nothing was asked about one’s ability to influence or change that policy. Policy and political knowledge is imperative if the midwifery workforce is to participate in policy-making processes throughout health care systems.

Risk interpretation varied between the groups. For example, all LMs described themselves as providing care for low- and low-moderate-risk women, whereas CNMs described themselves as providing care for low-, low-moderate-, and moderate-risk women. The American College of Nurse-Midwives home birth guidelines lists multiple gestation and breech presentations as increased-risk situations necessitating a hospital setting.¹⁴ Yet more than half of LMs responded that they independently manage the care of women with twins and breech presentations. These risk assessment and birth location differences lead to a logical deduction that LMs consider twin gestation and breech presentation to be low risk, eligible for independent management outside of the hospital. This reflect show an individual’s risk interpretation affects perception of one’s legal scope of practice.

Other practice differences included labor induction, labor support, and hydrotherapy use. LMs may refrain from use of continuous electronic fetal monitoring and induction prior to 42 weeks’ gestation because of their practice settings. As predominantly self-employed practitioners working in birth centers and homes, LMs have more freedom to refrain from practices they may not support. Generally, Dopplers are used by LMs, and although some birth centers have electronic fetal monitors, continuous electronic fetal monitoring and pharmacologic induction of labor are not the standard of care for out-of-hospital birth in Texas.¹¹ Likewise, the LMs’ significantly increased provision of labor support and waterbirth services may be associated with nonrestrictive facility policies, the patient population, or less workload. It is not possible to determine a cause and effect from the findings. However, the beliefs about time in labor revealed practice differences. Future studies are needed to address the significance of these different clinical approaches on women’s health outcomes.

Interprofessional teamwork improves coordination and communication, leading to improved quality; nurse-midwifery standards of practice underscore the importance of collaborative relationships with physicians.^15,16 The belief by many CNMs that a collaborative agreement with a physician is needed may be a reflection of the legal status of nurse-midwifery in Texas, which requires a collaborative agreement for delegated medical acts.¹⁷ Further investigation is needed to differentiate philosophical beliefs from legal requirements. For LMs, external physician consultation and collaboration present coordination challenges. Efforts to develop collaborative mechanisms supportive of broader midwifery care are needed. With CNMs, the education of nonmidwifery students for interprofessional teamwork is noteworthy. The Texas midwife and physician shortage necessitates further exploration of educational approaches.

Liability insurance protects against the financial risk associated with potential lawsuits or claims. Anecdotally, many LMs consider this unnecessary because of the strong relationships they form with their low-risk clients.¹⁸ The high percentage of CNMs carrying liability insurance, however, may be associated with the broader scope of practice, employer requirements, and insurance empanelment policies. Differences in scope of practice, worksite requirements, and professional ideology appear to drive this variation. Questions regarding beliefs about liability insurance would have provided more insight into this topic.

Limitations

There were several limitations of this study’s administration and design. First, the survey was dependent on computer and email access. Also, there was no verification if email addresses were correct or if emails were received and read.

In an effort to adapt the surveys to the scope of practice and terminology used by each group, the wording of some questions varied slightly between the 2 surveys, thereby limiting comparative analysis. Furthermore, questions regarding specific practices asked only if the practices were performed. Few questions sought to identify the effect of beliefs and policies on respondents’ practices.

Conformity, prestige, and response bias affect all surveys. These response rates surpassed the average 4.7% personalized internet survey response rates reported by Sinclair and associates.¹⁹ Because of financial restraints and satisfactory initial response rates, evaluation or follow-up of nonresponders was not initiated.

CONCLUSION

This initial comparative examination of the Texas midwifery workforce can serve as a model for other states. Attention to available workforce resources is key to developing a quality health care system. No midwives are categorized as health care providers in the Texas State Health Plan.²⁰ By describing the differences in demographics and employment characteristics, the geographic areas and populations served, and the differences in the nature and scope of CNM and LM clinical practice, this study provides information that has policy implications for influencing state health planners to consider the midwifery workforce when developing strategies to provide reproductive, gynecologic, and preventive care for all Texas women.

To improve midwifery workforce knowledge, future research could focus on seeking additional information in the following areas: workforce diversity, educational preparation, risk perception, scope of practice, conformity to professional standards, work setting preferences, influence over policy creation, liability insurance, exploration of CNM precepting, and women’s decision-making process when choosing a health provider.

• Midwifery has always been legal in Texas, but health workforce analysis is needed to develop strategies for including these services in strategic health plans for the state.

• Few Texas midwives practice in rural areas and small towns.

• Licensed midwives (LMs) are predominately white women who care for a diverse but largely white population, whereas certified nurse-midwives (CNMs) are more ethnically diverse and care for an increased number of African American or black and Hispanic women.

• CNMs and LMs differ significantly in their personal perceptions of health risks of women.

• LMs provide more nonpharmacological labor management; CNMs employ more technological and pharmacotherapeutic labor management.