Association of Clinical Nursing Work Environment with Quality and Safety in Maternity Care in the United States

Abstract

Purpose:

Maternal outcomes in the United States are the poorest of any high-income country. Efforts to improve the quality and safety of maternity care are frequently reported by individual hospitals, limiting generalizability. The purpose of this study is to describe maternity care quality and safety in hospitals in four states.

Study Design and Methods:

This cross-sectional study is a secondary analysis of the Panel Study of Effects of Changes in Nursing on Patient Outcomes data. Registered nurses reported on maternity unit quality, safety, and work environment. Descriptive statistics and clustered linear regressions were used.

Results:

The sample included 1,165 nurses reporting on 166 units in California, New Jersey, Pennsylvania, and Florida in 2015. One-third of nurses, on average, gave their units an overall safety grade of “excellent,” but this decreased to less than one-sixth of nurses in units with poor work environments. Overall, 65% of nurses reported that their mistakes were held against them. A good work environment, compared with poor, was significantly associated with fewer nurses grading safety as poor (β −35.6, 95% CI −42.9 – −28.3).

Clinical Implications:

Our research found that the nurses in the majority of hospitals with maternity units in four states representing a quarter of the nation’s annual births felt their units do not provide excellent quality care and have a less than optimal safety climate.

Birth is the most common reason for hospitalization in the United States (Martin et al., 2019). Due to persistently high rates of maternal morbidity and mortality, measuring and improving the quality and safety of maternity care has become a matter of national concern (American College of Obstetricians and Gynecologists [ACOG], 2017; National Academies of Sciences Engineering and Medicine [NASEM], 2020). Nurses provide the majority of hands-on care for nearly every woman who gives birth in the United States (NASEM). They have primary responsibility for many of the recommended improvements in the quality and safety of maternity care, such as fetal and maternal monitoring, labor support, and ambulation or repositioning. Research on the association of nursing care with maternal outcomes is limited (NASEM).

Qualitative research on maternity nursing care, and its association with maternal outcomes, has highlighted importance of nursing organizational factors for maternity outcomes. Nurses have reported that inadequate staffing affects both the quality and quantity of their care (Simpson & Lyndon, 2017a). When there is insufficient staffing, the first elements of nursing care to be missed are bedside presence and providing labor support, which are both associated with increased rates of vaginal birth (Simpson & Lyndon, 2017a). Nurses report that their care has an impact on birth outcomes and their physician colleagues and new mothers have agreed (Lyndon et al., 2017; Simpson & Lyndon, 2017b). The limited quantitative research on the association of nursing with maternal outcomes has reported heterogenous findings (Edmonds et al., 2017; Hodnett et al., 2002). Researchers have found that the hospital in which the birth takes place may explain more of the variation in birth outcomes than variation in care practices or the patient population (Grobman et al., 2014).

Although many studies have identified hospitals as a major source of variation in maternal outcomes, a hospital’s structural characteristics, such as bed size and teaching status, do not explain these variations. Nursing organizational factors are one possible explanation for between-hospital variation in maternal outcomes. The RN4CAST survey data are uniquely suited to address such questions, which are critical to improving maternity care and outcomes in the United States.

A robust body of literature, much of it stemming from Linda Aiken’s RN4CAST study, has shown that organizational nursing factors, such as staffing, education, and the work environment, are directly associated with patient outcomes (Aiken et al., 2008, 2011). The RN4CAST, a rich, unique, population survey of registered nurses, has been administered in 1999, 2005, 2015, and 2019. Although the first iteration of the survey was conducted in Pennsylvania, it was expanded to include California, New Jersey, and Florida in 2005 and 2015, thus including some of the most populous states in the country. In 2019, the survey was conducted in New York and Illinois. The survey uses nurses as primary informants on the hospitals and offices where they work and asks questions related to education, staffing, quality, safety and the work environment, among other things. This survey dataset serves many nursing and patient outcome related questions on the individual, unit, hospital, and multihospital level.

In other patient populations, from infants to the elderly, the clinical work environment is associated with numerous outcomes, including patient safety and the quality of care (Lake et al., 2016; Lasater, Sloane, et al., 2019). Other studies using the RN4CAST survey have found that the clinical work environment explains much or all of the variance in patient safety climate (Lake et al., 2018; Olds et al., 2017). Good clinical work environments may be a prerequisite for safe, high-quality care (Olds et al.).

Current bundles and consensus statements directed at improving the quality and safety of maternity care do not address the clinical work environment and this gap is beginning to receive recognition (Atallah & Goffman, 2019; NASEM, 2020). Such research in a maternity setting, however, is sparse (Lyndon et al., 2017). To address this gap, the purpose of this study is to describe the clinical work environment in U.S. maternity units and its association with quality and safety.

Study Design and Methods

This retrospective, cross-sectional study examined the nursing work environment, safety climate, and quality rating in 275 hospitals via a secondary analysis of the 2015 RN4CAST survey data and the American Hospital Association’s (AHA) 2015 Annual Survey (AHA, 2016). The RN4CAST study methodology has been described in detail elsewhere (Lasater, Jarrín, et al., 2019). In 2015, surveys were sent to a 30% random sample of licensed nurses in California, Florida, New Jersey, and Pennsylvania. The survey had a response rate of 26%. A random sample of nurses who did not respond to the initial survey request were resurveyed in a more intensive manner in order to evaluate nonresponse bias (achieving a response rate of 90%), which showed no bias in key variables (Lasater, Jarrín, et al.). The AHA’s Annual Survey provided hospital information. The University of Pennsylvania’s institutional review board approved the study protocol.

For this study, survey respondents were included in the sample if they were a 1) currently employed 2) staff nurse in a 3) maternity/newborn unit. Hospitals were included if they had at least four nurse respondents to obtain a stable aggregate measure of the work environment (Glick, 1985). The hospitals in the sample were nonfederal, adult, general acute-care hospitals.

The RN4CAST survey included the National Quality Forum-endorsed Practice Environment Scale of the Nursing Work Index (PES-NWI), a tool to measure the clinical work environment (Lake, 2002). The PES-NWI includes five domains comprised of traits associated with professional nursing practice: 1) Nurse Participation in Hospital Affairs; 2) Nursing Foundations for Quality of Care; 3) Nurse Manager Leadership, Ability and Support of Nurses; 4) Staffing and Resource Adequacy, and; 5) Collegial Nurse-Physician Relations. The five subscales had an overall Cronbach’s α of 0.85 with the individual subscales ranging from 0.79 to 0.89. Nurses respond whether the trait is present at their primary job via a four-point Likert scale ranging from 1 (Strongly Agree) to 4 (Strongly Disagree). The intraclass correlation coefficient(1,k) was 0.62, demon-strating sufficient reliability to aggregate to the unit level (Glick, 1985). The composite PES-NWI score was divided into quartiles with the top quartile defined as having good work environments, the bottom as poor, and the two middle quartiles combined as mixed.

The RN4CAST survey includes seven items from the Agency for Healthcare Research and Quality’s Hospital Survey on Patient Safety Culture (Sorra et al., 2016). Each item was measured via a five-point Likert scale ranging from 1 (Strongly Agree) to 5 (Strongly Disagree). Nurses also gave an overall safety grade ranging from A (Excellent) to F (Failing). Quality was rated on a four-point Likert scale from “excellent” to “poor.” We calculated the percentage of nurses on each unit who gave a safety grade of excellent, a safety grade of poor (C, D, or F), and an excellent quality rating.

Hospital characteristics, including bed size, technology, obstetric level of care, and teaching status were obtained from the AHA’s 2015 Annual Hospital Survey published in 2016. Bed size was defined as small (≤250 beds), medium (251–500 beds), and large (>500 beds). Technology status was defined as high or low based on whether the hospital has capabilities for performing open heart surgery and organ transplants. Obstetric level was defined as only uncomplicated maternity and newborn care (1), uncomplicated and most complications with special neonatal services (2), and all complications and presence of maternal–fetal medicine (3; AHA, 2018). Teaching status was based on the ratio of resident physicians in training to beds: major (1:4); minor (less than 1:4); and nonteaching (no residents).

Descriptive statistics were used to describe nurses and hospitals included in the study. Missing data on key variables at the nurse level were handled by listwise deletion. To calculate the unit-level mean work environment, a minimum of four nurses with complete data for each subscale were required. Eleven hospitals were missing data on one or more hospital characteristics and were included in regression models via dummy variables. Linear regression models at the unit level were used to examine association of the clinical work environment with nursing unit quality and safety. Dependent variables were the percentage of nurses that gave a grade of A for safety and excellent for quality. Hospital organizational characteristics, which might be associated with safety, were controlled for in the regression models. StataIC 15 was used for data analyses.

Results

The sample was comprised of 1,165 nurses reporting on 166 hospitals in California, New Jersey, Pennsylvania, and Florida in 2015. Births in these four states comprised 25% of all births in the United States in 2015 (Hamilton et al., 2016). Almost half of the hospitals were medium size (250–500 beds) with two-thirds being high-technology and about two-fifths nonteaching. Few of these hospitals (14.2%) reported caring for only uncomplicated maternity patients. Nurse respondents had a mean age of 47.8 years (SD 12.1) and were predominantly white (78.8%) and female (99.7%). Most had a Bachelor of Science in Nursing degree or higher (58.6%) and reported an average of 20.7 years of work experience (SD 12.6). Table 1 presents demographic detail for the nurses and the structural characteristics of hospitals in the study.

Table 1.

Characteristics of Hospitals (n_total = 166) and Nurses (n_total = 1,165)

			N (%)	Mean (SD)
Hospital
	Bed size
		≤250 beds	41 (24.7)
		>250 & ≤500 beds	79 (47.6)
		>500 beds	46 (27.7)
	Teaching†
		Non	67 (40.4)
		Minor	67 (40.4)
		Major	28 (16.9)
	Technology†
		Low	52 (31.3)
		High	110 (66.3)
	Obstetric level‡	Uncomplicated maternity and newborn	20 (14.2)
		All uncomplicated and most complicated	56 (39.7)
		All serious illnesses and abnormalities	65 (46.1)
Nurse
	Female		1,161 (99.7)
	Age			47.8 (12.1)
	Race	White	903 (78.8)
		Black	74 (6.5)
		Filipino	58 (5.1)
		Asian	50 (4.4)
	Years of Registered Nurse experience			20.7 (12.6)
	Bachelor degree in nursing or higher		682 (58.6)

^†Four hospitals missing teaching and technology status

^‡Twenty-five hospitals missing obstetric level

Average unit-level PES-NWI composite score was 2.8 (SD 0.3) and ranged from 1.8 to 3.6. Average PES-NWI score varied significantly across work environment categories, from 2.4 (SD 0.2) for poor environments to 3.2 (SD 0.2) in good environments (P < 0.001).

One-third of nurses gave their hospital an overall patient safety grade of “excellent” or “good.” Safety grade and quality varied significantly by the work environment. In units with poor work environments, an average of 42.4% of nurses gave a safety grade of C, D, or F (in some units, all the nurses gave a grade of C, D, or F). In good work environments, by contrast, only 7.8% of nurses, on average, gave a safety grade of C, D, or F. The highest percent of nurses in a unit with a good work environment giving poor safety grades was 27.3%. In poor work environments, 29.4% of nurses rated care quality as “excellent,” compared with 74.6% of nurses in good work environments.

Less than half of the nurses reported feeling free to question decisions of those in authority. Only 25% of nurses, on average, disagreed that their mistakes were held against them. About half of the respondents disagreed that important information was often lost during change of shift or at transfer. Safety items related to actions were better rated, with about three-quarters of nurses reporting that error prevention was discussed on their unit and roughly two-thirds acknowledged receiving feedback about changes based on event reports. Further information on quality and safety is reported in Table 2.

Table 2.

Maternity (Unit-Level) Safety Climate Percent Positive Responses

		Mean (%; SD)
Quality Grade (Excellent)		49.4 (27.3)
Safety Grade (Excellent)		33.3 (24.2)
Safety Grade C, D, F		23.9 (19.8)
Safety Culture Items
Staff feel free to question the decisions or actions of those in authority		43.7 (24.6)
Staff feel like mistakes are held against thema		26.7 (21.0)
Important patient care information is often lost during shift changesa		49.7 (22.5)
Things “fall between the cracks” when transferring patients from one unit to anothera		44.8 (22.6)
In this unit, we discuss ways to prevent errors from happening again		80.0 (18.2)
We are given feedback about changes put into place based on event reports		58.2 (22.7)
The actions of hospital management show that patient safety is a top priority		63.9 (24.2)

^aItem is reverse coded

Fewer nurses graded safety as poor in a good work environment, as compared with a poor environment (adjusted β −35.6, 95% CI −42.9 – −28.3). A similar pattern was noted for mixed, as compared with poor environments (adjusted β −19.2, 95% CI −25.3 – −13.1). In a similar manner, the percentage of nurses who rated care quality as excellent was significantly higher in better work environments than in poor (adjusted β 43.5, 95% CI 33.1–53.9) as well as in mixed, compared with poor environments (adjusted β 15.3, 95% CI 6.6–24.0). Table 3 presents the unadjusted and adjusted regression results.

Table 3.

Regression Results, the Association of Work Environment by Safety and Quality at the Unit-Level (N = 166)

	Percentage of Nurses Who Graded Safety as Poora		Percentage of Nurses Who Graded Quality as Excellentb
	Coefficient (95% CI) Unadjusted Model	Coefficient (95% CI) Adjusted Model†	Coefficient (95% CI) Unadjusted Model	Coefficient (95% CI) Adjusted Model†
Good environments	−36.1 (−43.4 – −28.8)‡	−35.6 (−42.9 – −28.3)‡	45.7 (35.2–56.2)‡	43.5 (33.1–53.9)‡
Mixed environments	−20.4 (−26.5 – −14.3)‡	−19.2 (−25.3 – −13.1)‡	17.4 (8.7–26.1)‡	15.3 (6.6–24.0)‡
Poor environments	Reference	Reference	Reference	Reference

^aPoor is defined as a safety rating of fair, poor, or failing.

^bExcellent is defined as a quality rating of excellent.

^†Adjusted model controls for hospital size, technology, and teaching status.

^‡p ≤ 0.001

Clinical Implications

In light of the prevalence of poor maternal outcomes in the United States, we were interested in studying the profession responsible for most of the care for almost all laboring women in this country; nurses (NASEM, 2020). Specifically, we looked at the clinical work environment and nurses’ assessments of quality and safety. We found that roughly one-quarter of nurses gave their unit a safety grade of C, D, or F. In units with poor work environments, 42.4% of nurses gave a poor safety rating, compared with 7.8% of nurses in good work environments. Similarly, 29.4% of nurses in units with poor work environments rated quality of care as excellent, compared with 74.6% of nurses in good work environments. Quality of the clinical work environment predicted safety and quality. As in research in other patient populations, the better the work environment, the better the safety and quality of care (Lake et al., 2018; Olds et al., 2017). The impact of the clinical work environment on the quality and safety of maternity care suggests that the context in which care is given and received deserves attention from nursing, medical, and hospital leaders.

Although there is always room to improve safety culture, even in the best units, two areas require immediate attention. Only 25.2% of nurses in the study disagreed that their mistakes were held against them. Even in the three units in the sample where all nurses rated safety as excellent, only one-third disagreed that their mistakes were held against them. This number stands in stark contrast to the recommendations of the Institute of Medicine’s To Err is Human report of 20 years ago, suggesting that these recommendations have not been realized in maternity care (Institute of Medicine, 1999). Hospital administrators and nurse managers should work to build and support a safety culture where mistakes are recognized, reported, studied for ways to prevent recurrence, and actions are taken to mitigate harm to the patient, while not penalizing the nurse or provider who reports the error. Maternity nursing instructors and clinical preceptors can teach their nursing students to report errors appropriately and build students’ confidence to do so by helping them to take necessary action when they make an error, not penalizing them.

The other concerning finding is that less than half of nurses reported feeling free to question the decisions of those in authority. Poor communication underlies two-thirds of preventable maternal mortality and is a focus of safety initiatives (Centers for Disease Control and Prevention, 2019; Lyndon et al., 2015). Ability to question decisions of those in authority is a marker of high-reliability organizations that have better outcomes in critical situations (Lyndon et al., 2012). Nurse managers and charge nurses can work to improve healthcare team dynamics and communication, including advocating for the importance of “see something, say something” as critical for patient safety and team performance. Conducting simulations with the multidisciplinary healthcare team is another way to build communication overall and the capacity to question decisions appropriately, specifically. Safety scale items related to action were better rated by nurses than items related to either safety culture or communication. This suggests, as has been noted elsewhere, that quality and safety initiatives in a poor or mediocre work environment are less effective (Lake et al., 2018).

A good clinical work environment may be a prerequisite to a strong safety culture and high-quality care (Lake et al., 2016, 2018). Excellent quality and safety ratings vary significantly across clinical work environments. Far fewer nurses mark quality and safety as “excellent” in units with poor work environments. Research in other populations has found the clinical work environment predicts quality grade (Lake et al., 2016). Much of the research and recommendations on quality and safety in maternity care does not account for the organizational context in which care is delivered (ACOG, 2017; Howell & Zeitlin, 2017; Howell et al., 2014).

Future research examining hospital variation in maternal outcomes, and the contribution of nurses to maternal outcomes, should take the organizational context of nursing care into account. Longitudinal studies should be conducted to further our understanding of the association of nursing organizational factors with maternal outcomes and should specifically identify labor and birth versus postpartum nurses.

Use of cross-sectional data for this study limits our ability to infer causality, indicating need for longitudinal studies. Survey respondents identified themselves as maternity/newborn nurses and did not include neonatal intensive care nurses, who were able to identify themselves separately. Although this study was not able to examine labor and birth nurses separately from postpartum and well-newborn nurses, these units tend to be closely connected organizationally because they are serving the same specialized patient group.

Safety and quality in maternity settings is uneven and a substantial fraction of direct care nurses in this study considered the care on their units to be unsafe or of poor quality. There is significant room for improvement in the quality and safety of maternity care. Attention to the work environment has potential to improve the quality and safety of maternity care, thereby improving outcomes. ✜

Suggested Clinical Implications.

• Nurse managers can work to improve the environment on their unit through their leadership, supporting and advocating for nurses, providing opportunities for professional development, working to ensure adequate nurse staffing, and nurturing collegial relations between physicians, nurse midwives, other perinatal team members, and nurses.

• Nurture a safety culture that recognizes “to err is human” recognizing that doing so should not be penalized. Demonstrate this safety culture by highlighting and valuing those who share these types of events and concerns.

• Build a safety culture where decisions may be questioned without fear of penalization by the questioner.

• Consider the role of organizational characteristics on nurses’ ability to provide care.

• Work for collegial relations with physician colleagues and other members of the perinatal team. Consider simulations as opportunities to practice team communication.