Educational Preparation and Nurse Turnover Intention From the Hospital Bedside

Mary Jane K. DiMattio, PhD, RN; Adele M. Spegman, PhD, RN


Online J Issues Nurs. 2019;24(2) 

In This Article

Study Methods

Design and Sample

Data for this study were taken from a larger study of career trajectories of nurses at an integrated health system based in central Pennsylvania. The parent study had a response rate of 25%. The current study employed a cross-sectional, comparative design to address the research questions, using a non-probability sample of registered nurses who identified their primary work setting to be at the point of care (i.e., bedside) in the setting of the acute care hospital. Nurses in managerial, administrative, or support roles were not included.

Data Collection and Human Subjects

The health system's institutional review board deemed this study as exempt research. Self-report data were collected using an electronic survey that included 60 items and required 15–20 minutes to complete. Submission of the survey implied consent. Responses to the survey were stored on an SQL database behind a corporate firewall and then downloaded to a spreadsheet. There was no ability to link study data with specific participants.


Turnover Intent. The dependent variable for this study was assessed using a question about nurses' intention to remain at or to leave bedside nursing in the acute care hospital setting. For the larger study, this question had nine categories, which were collapsed into three: "I intend to remain as a bedside nurse for the remainder of my career; I intend to leave bedside nursing within the next 1–2 years; I already left bedside nursing." This study only examined results from nurses who selected either of the first two of the three categories.

Educational Preparation. Nurses were asked to identify their highest nursing degree from seven choices (diploma, associate's degree, baccalaureate, master's, DNP, PhD, EdD) and to respond to a dichotomous variable assessing whether or not they were currently enrolled in a nursing education program. If they responded in the affirmative, they were then prompted to indicate the program type.

Clinical Expertise. As was reported by Kutney-Lee et al. (2009), nurses were asked to rate their nursing expertise according to definitions of Benner's (1984) categories of advanced beginner, competent, proficient, and expert.

Years of Experience. Nurses were asked to choose from one of 6 categories indicating years of experience as an RN (0–2, 3–5, 6–10, 11–15, 16–20, 21 or more).

Practice Environment. This study used the Conditions for Work Effectiveness Questionnaire [CWEQ II] (Laschinger, 2012) to examine nurses' perceptions of their practice environments in terms of structural empowerment. Structurally empowering practice environments afford access to opportunity and power to "get things done" (Laschinger, 2012). The CWEQ II has 12 items that assess access to opportunity, resources, information, and support on four subscales, which are conceptually similar to the following three subscales of the Practice Environment Scale of the Nursing Work Index, used by Kutney-Lee at colleagues (2009): nurse participation in hospital affairs; staffing and resource adequacy; and nurse manager ability, leadership, and support of nurses. CWEQ II's developer established its validity, and many studies have confirmed the instrument's reliability (Heather K. Laschinger Research Tools, 2017). In this study, Cronbach's alphas were used to assess reliability, and all were consistent with the literature. They were as follows: overall structural empowerment, a = 0.89; Resources, a = 0.84; Support, a = 0.84; Opportunity, a = 0.85; and Information, a = 0.93.


Data were analyzed using SPSS® software version 22 and SAS® version 9.4. Descriptive, parametric, and non-parametric statistics were used to describe study variables and to explore bivariate relationships. Logistic regression analysis was used to regress the dependent variable, treated as dichotomous (intent to leave bedside nursing or intent to remain as a bedside nurse) on the independent variables of educational preparation, years of experience, clinical expertise, and practice environment. In addition, interaction effects were tested in the regression analyses modeling the odds of leaving predicted by the independent variables.


Four hundred twenty-four RNs identified their primary area of nursing practice as at the bedside. Nearly 63% of these nurses (n = 266) intended to remain at the bedside for the duration of their careers, whereas 30.4% (n = 129) indicated an intention to leave bedside nursing in 1–2 years. Twenty-three (5%), however, reported they had already left bedside nursing. Three of these nurses were also employed in other settings, suggesting they maintained their hospital practice at least part of the time. It was not clear, however, why the remaining 20 answered this way, so data for these 23 nurses along with 6 who did not answer the intention question were removed, leaving 395 cases for analysis.

RN Characteristics

Table 1 displays RN characteristics, including demographics, years of experience, and clinical expertise. Over a quarter of the sample had 21 or more years of experience, and most nurses self-reported as proficient. Nearly 43 percent (42.8%) had a baccalaureate degree, whereas 55% had less than a baccalaureate degree as their highest level of educational preparation (21.5% diploma and 33.2% associate's degree). Two percent had a master's degree, and none reported higher than a master's degree. Twenty-two percent (n = 85) of the sample were enrolled in a nursing education program. Analysis of enrollment in an educational program by highest educational preparation revealed that associate degree nurses represented the highest proportion of nurses enrolled in an educational program (48.2%). The next highest proportion of nurses enrolled was those with a baccalaureate degree (37.6%), followed by nurses with a diploma (9.4%) and master's degree (3.5%).

Practice Environment

The average score for overall structural empowerment on the CWEQ II was 19.43 (SD = 3.64), representing "moderate" empowerment (Laschinger, 2012). Subscale means were as follows: Resources (M = 3.07, SD = 0.84); Information (M = 3.15, SD = .91); Opportunity (M = 3.91, SD = 0.74); and Support (M = 3.11, SD = 0.88).

Research Question 1

Nearly a third of the sample (32.7%) reported the intention to leave hospital nursing within the next 1–2 years. Nurses whose highest preparation was a baccalaureate degree or more were significantly more likely to intend (42.9%) to leave hospital nursing than those with less than a baccalaureate degree (24.5%) (χ2 (1, n = 393) = 14.94, p < .001). Similarly, nurses who were currently enrolled in school were also significantly more likely to intend to leave (58.8%) than those who were not enrolled (25.5%) (χ2 (1, n = 395) = 33.72, p< .001). Table 2 illustrates the degrees to be awarded. Of the nurses (n = 46) enrolled in a baccalaureate degree program, 21 (45.7%) intended to leave in 1–2 years.

Research Question 2

In terms of experience, the category of 0–2 years had the highest proportion of nurses who intended to leave (31%), and the category of 11–15 years had the lowest (7%). Nurses with 0–10 years of experience were significantly more likely to intend to leave (71.3%) than those with 11 to >20 (28.7%) years of experience (χ2 (1, n = 395) = 20.15, p < .001). The category of clinical expertise with the highest proportion of nurses who intended to leave was Proficient (35.7%), but more of the Proficient nurses intended to stay (43.2%). Slightly more Expert nurses intended to stay (27.4%) than to leave (25.6%), but there were no statistical differences on this variable.

After Bonferroni correction for multiple comparisons, nurses who intended to leave reported a significantly lower mean score for Opportunity (M = 3.75, SD = 0.75) than those who intended to stay (M = 3.99. SD = 0.73), t (393) = 3.07, p = .002. The same was true for Support where nurses who intended to leave had a lower mean score (M = 2.95, SD = 0.88) than those who intended to stay (M = 3.18, SD = 0.88), t (393) = 2.51, p = .01). Mean scores were also lower for nurses who intended to leave on Resources and Information, but these differences were not significant after correction. Finally, nurses who intended to leave reported significantly lower overall structural empowerment (M = 18.49; SD = 3.70) than those who intended to stay in hospital nursing (M = 19.88, SD = 3.53), t (393) = 3.63, p < .001.

Research Question 3

Variables on which there were significant differences between nurses intending to leave and those intending to stay (highest degree in nursing; enrollment in school; years of experience, and overall structural empowerment) were entered into the binary logistic regression model. Nurses with a baccalaureate degree or more and those who were enrolled in school had significantly higher odds of intention to leave the hospital bedside than those with less education and who were not enrolled in school. Nurses with 11 or more years of experience had significantly lower odds of intention to leave than those with less experience, and every unit increase in structural empowerment was significantly associated with lower odds of intention to leave.

Because of the strong relationship between satisfaction and turnover intent demonstrated in the literature, as well as the relationship between job satisfaction and the practice environment (Al Maqbali, 2015), four items assessing job satisfaction collected for the larger study were combined and added to the model as a predictor of turnover. The four satisfaction items included a global assessment of satisfaction on a 4-point scale and three items assessing whether or not the respondent would do the following: take the same type of job again given current knowledge of the job; choose the type of job again given the choice of any job; and recommend the job to a friend. When entered into the model, every unit increase in satisfaction predicted lower odds of intent to leave, and structural empowerment was no longer a significant predictor. Further, the removal of structural empowerment as a predictor and addition of job satisfaction improved the overall model fit (AIC = 419.44) compared to the original model (AIC = 427.31).

Table 3 shows the final predictive model in which nurses with a baccalaureate degree or higher had almost 3 times greater odds of intention to leave than nurses with less than a baccalaureate degree, and nurses enrolled in school had almost 5 times greater odds of intention to leave than nurses who were not enrolled in school. Nurses with 11 or more years of experience had lower odds of intention to leave, as did nurses with higher scores on the composite of the 4 items assessing job satisfaction.

Research Question 4

There were no differences on overall structural empowerment between nurses with a baccalaureate degree or more (M = 19.60, SD = 3.55) and those with less than a baccalaureate degree (M = 19.34, SD = 3.69), t (391) = -0.72, p = 0.47. There were also no differences between the groups on any of the subscales assessing the dimensions of structural empowerment. Nurses with a baccalaureate degree or more had a higher score for job satisfaction (M = 10.10, SD = 1.93) than nurses with less than a baccalaureate degree (M = 9.81, SD = 2.08), but this difference was not statistically significant t (391) = -1.45, p = 0.15). Interaction effects on intention to leave were examined between educational preparation and structural empowerment, as well as between educational preparation and satisfaction. Neither interaction was a significant predictor of intention to leave.