Nurses' control over practice is essential to nursing care quality and fosters teamwork at the point of care delivery Nurses' control over practice is essential to nursing care quality and fosters teamwork at the point of care delivery (Mallidou, Cummings, Estabrooks, & Giovannetti, 2011). The ability for nurses to control practice stems from three facets of power: organizational structure, personal belief, and relational leverage (Manojlovich, 2007). Organizational structures that promote empowerment are a characteristic of Magnet hospitals and include nursing opportunity and access to resources and decision-making at all levels of the organization. Control over practice also relies on an individual nurse's value for self-determination and psychological belief, or self-efficacy, that s/he can change and should change practice. Finally, control over practice relies on the empowerment leveraged through professional relationship reciprocity, teamwork, and activities that strengthen personal bonds in the workplace.

Control over practice, or decision-making at the bedside staff nurse level, is an essential moderator of teamwork behaviors (Morey et al., 2002). In turn, effective teamwork is an essential component of nursing care safety and quality (Kohn, Corrigan, & Donaldson, 2000). According to the Agency for Healthcare Research and Quality (AHRQ, 2006) Model of Teamwork Training, effective teamwork contains the following dimensions: operative team structure, leadership, situation monitoring, mutual support, and communication. Team structure includes the team size, members, goals, and purpose. Leadership is assigning, coordinating, and evaluating performance through modeling and setting expectations. Situation monitoring is the awareness and ability to conceptualize the duality of current interrelated functioning of each team member with the overarching desired purpose and movement of the team’s efforts (Salas, Sims, & Burke, 2005). Mutual support is the fluid adjustment to redistribute workflow to support the goals and expected outcomes of team member’s work as well as one’s own workflow (Castner, Foltz-Ramos, Schwartz, & Ceravolo, 2012). Communication is sharing unified perceptions and meanings through information exchange.

In response to the Institute of Medicine’s report (Greiner & Knebel, 2003), Health Professions Education: A Bridge to Quality, several teamwork competencies have been added to undergraduate and graduate healthcare professional programs and curricula (Quality and Safety Educaton for Nurses, 2010). In addition, AHRQ developed a standardized teamwork training curriculum, called TeamSTEPPS^®, for professional development (AHRQ, 2006). Teamwork training alone is not sufficient to enhance teamwork behaviors evident at the point of patient care delivery. Other factors, such as engaged leadership and flattened organizational hierarchies (that contribute to nursing control over practice), are necessary for teamwork behavior to emerge at the bedside (Frankel, Leonard, & Denham, 2006).

Lack of control over practice adds unnecessarily to the hypercomplexity of the nursing work environment. The theoretical and empirical link between nursing control over practice and teamwork is emerging in the safety and quality literature. Perceived care quality, which is enhanced by teamwork, is also elevated when nurses experience higher levels of autonomy in their practice (Mallidou et al., 2011). Theoretically, concepts of teamwork have been linked to a hospital culture with high accountability and improved patient outcomes (Denham et al., 2005; Frankel et al., 2006). For example, by implementing both teamwork training and control over practice, nurses were able to decrease catheter-related bloodstream infections by eliminating physician orders that were not in keeping with current evidenced-based guidelines (Berenholtz et al., 2004). Higher levels of control over practice that minimize power imbalances with other team-members are correlated with lower levels of failure to rescue and urinary tract infections (Boyle, 2004). Disowning responsibility for selected aspects of patient care as a barrier to teamwork has been reported in hospital nursing research, where nursing work design is fragmented by a lack of control over practice (Kalisch, Weaver, & Salas, 2009). Lack of control over practice adds unnecessarily to the hypercomplexity of the nursing work environment (Krichbaum et al., 2007). No research was found that directly linked control over practice to the five dimensions of teamwork: team structure, leadership, mutual support, situation monitoring, and communication. In addition, the impact of years of experience is also unclear and nurses’ experience and tacit understanding may, pragmatically, enhance the sense of control over practice.

Purpose

The purpose of this study was to measure the relationship of control over practice to the five following dimensions of teamwork: team structure, leadership, situation monitoring, mutual support, and communication.

The research questions for this study were:

• Controlling for years of experience in the current role for hospital bedside registered nurses (RNs), does control over practice increase the following five dimensions of teamwork: team structure, leadership, situation monitoring, mutual support, and communication?
• Controlling for years of experience in the current role for hospital bedside RNs, does control over practice interact with TeamSTEPPS® teamwork training to increase the following five dimensions of teamwork: team structure, leadership, situation monitoring, mutual support, and communication?

Methods

Design

Data for this secondary analysis research were gathered from a cross-sectional survey study. Data were collected between February and April of 2011. Previous research using this dataset indicated group differences in teamwork perceptions between those with high control over practice and those without high control over practice (Castner, 2012). Therefore, this analysis was undertaken to focus the statistical analysis on the impact of control over practice and teamwork.

Sample and Setting

A convenience sample of 456 RNs from five hospitals completed the survey administered for this study. A total of 2,056 inpatient RNs were eligible to participate, and a recruitment continued until a target of at least 20 percent of eligible RNs from each hospital participated. The five hospitals were all a part of a single health system in one Northeastern metropolitan area. The hospitals included one children’s hospital, two suburban community hospitals, and two central-city tertiary care hospitals. Over 1,200 of the eligible nurses had attended a 90 minute TeamSTEPPS^® teamwork training session based on AHRQ’s curriculum. The sessions had been offered throughout the hospital system for a year previous to the survey through new employee orientation and hospital-based professional development. Before instituting TeamSTEPPS^®, the hospital system had held optional workshops on strengthening assertive communication, with outcomes reported in Ceravolo, Schwartz, Foltz-Ramos, & Castner (2012). Other teamwork training was offered in a tailored unit-to-unit format, led by the discretion, priority, and effort of individual unit managers. Due to the informal and unit-specific nature of additional teamwork training, only TeamSTEPPS^® training was measured for this analysis.

The research protocol was approved by the University at Buffalo Social and Behavioral Sciences Institutional Review Board and hospital leadership. Participants were recruited through the site’s nurse educator who rounded to nursing unit workstations during a four week period. During usual rounds, the nurse educator distributed a cover letter, explained the purpose of the survey, and instructed interested RNs in how to access the electronic link for the study survey.

Measurement

The survey included demographic information (e.g., years of experience in current RN role); an item assessing control over practice; an item assessing any history of attendance at the institutions TeamSTEPPS^® teamwork training; and the 20-item Brief TeamSTEPPS^® Teamwork Perceptions Questionnaire (Brief T-TPQ). Control over practice has been frequently measured as a single item in nursing studies (Kramer & Schmalenberg, 2003). The item with the stem, “The amount of control I believe I have over my own practice in my current position is,” was included for this study with the response options of high, medium, or low. The Brief T-TPQ includes the following subscales: team structure, team leadership, mutual support, situation monitoring, and communication. The Likert-type responses are coded as 5= Strongly Agree, 4 = Agree, 3 = Neutral, 2= Disagree, and 1= Strongly Disagree. Descriptive results of each dimension of teamwork measured by the Brief T-TPQ, and inferences related to an organizational intervention to deliver TeamSTEPPS^® training, were published in a separate article (Castner et al., 2012). The Cronbach alpha for the Brief T-TPQ subscales ranged from .83-.94, and for the whole survey was .93 (Castner, 2012). Construct validity was supported through factor analysis (Castner, 2012).

Data Analysis

Control over practice was scored as High and Not High (Not High is a composite variable for both low and medium responses), for ease of interpretation and dissemination of the statistical results. TeamSTEPPS^® training attendance was scored dichotomously as attended or not attended. The mean score for the four corresponding items in each subscale of the Brief T-TPQ was calculated, and any subscale with more than 50% missing answers were not utilized (representing less than 1% of the sample). Data were screened using the recommendations in Tabachnick and Fidell (2007). A skew was detected consistently in the Brief T-TPQ responses and subscales (skewness -.381to-.541); otherwise, data screening met the assumptions for the chosen statistics.

A two-way between subjects multivariate analysis of covariance (MANCOVA) was conducted on the five dependent variables composing teamwork: team structure, leadership, situation monitoring, mutual support, and communication after screening for appropriate mild to moderate correlation of the dependent variables. The independent variables were control over practice (high/not high) and TeamSTEPPS^® training attendance (attend/not attend). The covariate of years of experience in current role was included as a variable of theoretical interest.

Results

Description of the Study Sample

The dataset included 456 surveys, representing 22 percent of eligible participants. The RNs had an average of 14 years (SD=12) in their current position, with a range of 0-44 years. Forty-three percent (n=194) reported having high control over practice. Half of the participants (n=226) had attended a 90 minute TeamSTEPPS^® training presentation offered as professional development by their hospital system. Table 1 lists the means and standard deviations of each of the scale divided by both control over practice and teamwork training groups.

Table 1. Mean and Standard Deviation for Dimensions of Teamwork by Control Over Practice and Teamwork Training Attendance

Correlations and Test Assumptions

There was no relationship between years of experience in the current role and belonging to the group with high control over practice (F _{(1, 428)} =2.5, p=.104) or the group that attended TeamSTEPPS^® training (F _(1,428) =.8, p=.376), as evidenced by a non-significant result using one-way ANOVA. Table 2 lists the correlations among the five dependent variables. The moderate correlations of the five dependent variables, ranging from .33 to .71, are appropriate for the using the chosen statistic, MANCOVA (Meyers, Gamst, & Guarino, 2006). With a Box’s M test statistically insignificant (p>.05), the variance-covariance matrices of the dependent variables were equal across the independent variables. Thus, the variance within and between the responses of each group, or category, was appropriate for the chosen statistical test. Years of experience demonstrated no independent correlation with any dimension of teamwork (.01-.07, p>.05). The interaction terms for years of experience with both control over practice and TeamSTEPPS^® attendance were not significant (p>.0) and the final results were similar when the model was run with or without this covariate. The final model reported includes the years of experience covariate due to the impact of global teamwork.

Table 2: Correlation Coefficients for the Five Dimensions of Teamwork

Note: **p < .01.

Group Differences

The global measure of teamwork, or weighted composite of all five measured dimensions of teamwork, demonstrated a relationship with high control over practice and years of experience in the current role, but not with teamwork training (Figure 1 & Table 3). Figure 2 depicts the results when broken down into relationships with the specific dimensions of teamwork; all five dimensions of teamwork were perceived as better for those who had a high level of control over practice compared to those who did not. Years of experience in the current role did not demonstrate a relationship with any one specific dimension of teamwork. Only leadership was perceived as better for those who attended Teamstepps^® (M=3.5, SD=1.1) than those who did not (M=3.1, SD=1.1), F(1,425)=8.1, p=.005, partial η2=.019. However, the Situation Monitoring subscale demonstrated an interaction effect between TeamSTEPPs^® attendance and control over practice F_(1,425)=5.2, p=.023, partial  η²=.012. The multivariate interaction of TeamSTEPPS^® attendance and control over practice was not statistically significant for Team Structure, Leadership, Mutual Support, and Communication.

Table 3. Multivariate and Univariate Analysis of Variance for Teamwork Measures

[see full size Table 3 pdf]

Figure 1: Statistically Significant Relationships with Teamwork Composite

Figure 2: Statistically Significant Relationships with Five Teamwork Domensions

Discussion

High control over practice exerted a positive, statistically significant relationship with all five teamwork skill and behavior perceptions... High control over practice exerted a positive, statistically significant relationship with all five teamwork skill and behavior perceptions with small effect sizes. Additionally, there was an interaction effect between high control over practice and TeamSTEPPS^® training to improved situation monitoring perceptions, suggesting training effectiveness is contingent upon nursing control over practice. Between nurses who had attended TeamSTEPPS^® and those who did not, the only difference was the perception of leadership. Given the expectation that nurses who attended the workshop shared what they learned with co-workers who did not attend (or expectation of a diffusion of intervention), this is not a surprising finding. It is unclear if nurses with a greater sense and awareness of leadership would have self-selected to attend the training or if the exposure to education outside of the patient care environment enhanced perceptions of leadership.

High performing hospitals demonstrate not only higher levels of nursing control over practice at the point of care, but also within governance of the organization.Control over practice is not only an inherent employee characteristic, but a formalized organizational structure (Kramer & Schmalenberg, 2003). In nursing, professional autonomy is often synonymous with control over practice. When designing, performing, or scheduling any aspect of work or self-management, autonomy is the amount of independence and judgment an employee is afforded to complete the work (Cummings & Worley, 2009). Self-managed employees, across professions and workplaces, are more likely to be invested in the organization and feel responsible for the outcomes of their work. Empowered decision-making over team structure and resources needed in clinical practice are key to nursing control over practice (Kramer & Schmalenberg, 2003). High performing hospitals demonstrate not only higher levels of nursing control over practice at the point of care, but also within governance of the organization (Aiken, Clarke, & Sloane, 2002). This study contributed to the understanding of control over practice as a variable that has a positive relationship with teamwork skill and behavior perceptions.

Years of experience in the current role had a relationship with the global composite of all five dimensions of teamwork, but not with any one specific dimension. Thus, years of experience exerts a gestalt impact on elevating teamwork, without a statically significant relationship with any one dimension of teamwork. Further research is needed to uncover specifically how years of experience contributes to teamwork skills and behaviors and determine how this impact can be leveraged to enhance teamwork skills and behaviors.

The effect size of control over practice on the dimensions of teamwork is small, accounting for 2 to 4% of the overall variability in subjects’ responses on teamwork. Likewise, years of experience accounted for 4% of the variability. This study supports the idea that the foundations and situations that lead to teamwork in healthcare are complex and multi-dimensional. While nursing control over practice is an essential component to elevating teamwork, additional contributors to teamwork must continue to be studied and supported.

As a single, cross-sectional design, the relationships do not demonstrate causation, only correlation. Selection bias may have impacted results as the survey relied on self-report and convenience sampling. Control over practice was measured with a single item and not a subscale with confirmed validity and reliability. The generalizability of the results is limited by the sample recruitment from one hospital system in one metropolitan area of the United States. The statistical assumption of normality was violated, as data was positively skewed. However, we utilized statistics that were robust to violation of the normality assumption when all variables are skewed in the same direction.

Implications for Practice and Research

Downplaying organizational power structures and hierarchies is a crucial step towards enhancing the patient safety impact potential for the nursing profession. Focus on nursing control over practice, nurse self-management, and shared governance is an essential variable to enhancing teamwork skills and behaviors for nurses. Downplaying organizational power structures and hierarchies is a crucial step towards enhancing the patient safety impact potential for the nursing profession. The cause and effect relationship of control over practice on teamwork cannot be ascertained by this study’s cross-sectional design. Further study is required to explore control over practice as a potentially moderating or mediating variable of other precursors to successful teamwork.

Conclusion

Nursing control over practice demonstrates a positive relationship with effective teamwork. This study contributes preliminary understanding of control over practice as a variable that should continue to be addressed in policy, practice and research with a goal to enhance teamwork.

Acknowledgement: The authors would like to acknowledge Jin Young Seo for her assistance in preparing this manuscript, the nurses who participated in the study, and the nurse educators who collected the data.

Authors

Jessica Castner, PhD, RN, CEN
E-mail: jcastner@buffalo.edu

Jessica Castner, PhD, RN, CEN is a research assistant professor at the University at Buffalo’s School of Nursing. Dr. Castner’s background includes a decade of bedside emergency nursing and over six years of nursing education at the undergraduate and graduate level. Dr. Castner has led efforts to enhance teamwork training initiatives into regional nursing curricula, professional education, and nurse residency programs. Her publications on nursing teamwork include the organizational development results of teamwork training, testing for the validity and reliability of the Brief T-TPQ survey, and a pre-and post-test of lateral violence with a communication initiative.

Diane J. Ceravolo, MS, RN
E-mail: dceravolo@chsbuffalo.org

Diane Ceravolo, MS, RN is the director of nursing practice for the Catholic Health system. As a Master TeamSTEPPS® trainer, she has worked extensively at the corporate hospital level to lead teamwork training, communication, and interdisciplinary collegiality initiatives. Her extensive career in nursing management and leadership includes leadership positions from the unit to the corporate level, with a focus on nurse recruitment, retention, and education.

Kelly Foltz-Ramos, MS, RN
E-mail: foltz@buffalo.edu

Kelly Foltz Ramos MS, RN, FNP-BC, RHIA is the Clinical Lab Coordinator at the University at Buffalo’s School of Nursing. Ms. Foltz-Ramos spent four years in staff development at a regional hospital system. In her role in staff development, she was involved in the transition of new nurses to practice through a nurse residency program and a corporate orientation initiative which included providing TeamSTEPPS® training. Ms. Foltz-Ramos’ current area of research involves teamwork and interprofessional education.

Yow-Wu Bill Wu, PhD
E-mail: nurwu@buffalo.edu

Yow-Wu (Bill) Wu, PhD, MA, MS is an associate professor at the University at Buffalo’s School of Nursing. Dr. Wu’s expertise includes research methods for the practice professions, psychosocial measurement and questionnaire construction, and advanced statistics. Dr. Wu teaches statistics for graduate nursing students and provides statistics support to nurses through the Center for Nursing Research at the University at Buffalo.

© 2013 OJIN: The Online Journal of Issues in Nursing
Article published May 31, 2013