Improving psychological safety and
engagement in team learning
the multiprofessional TeamSTEPPS™ intervention
Master Thesis MSc BA - Change ManagementFaculty of Economics and Business, University of Groningen
August 2013 ESTHER DE JAGER Student number: s2031388 Hoornweg 55, 9363 EE Marum +31(6)54788378 / jager.de.esther@gmail.com Supervisor / university dr. M.A.G. van Offenbeek Supervisors / field of study
Abstract
This research evaluates the TeamSTEPPS™ intervention at the department of cardiothoracic surgery of one of the largest academic hospitals in the Netherlands. It examines the influence of teamwork training on psychological safety and engagement in team learning and the
relationships between these variables. Also, it assesses the influence of professional hierarchy on psychological safety and the effect of teamwork training on this. This was done by a
longitudinal embedded single-case study based on quantitative data from questionnaires and qualitative data from interviews. It shows that teamwork training did improve psychological safety, which fully mediated the increase of engagement in team learning. This study found non physicians to have higher feelings of psychological safety than physicians. Both groups were affected similarly by teamwork training.
Keywords: Teamwork training, psychological safety, engagement in team learning, professional
hierarchy, physician-nurse collaboration, TeamSTEPPS™.
Acknowledgment:
I thank my university supervisor Dr. M.A.G. van Offenbeek for providing me with quick and critical feedback and being very flexible. Secondly, I would like to thank my supervisors at the hospital for giving me the opportunity to write my thesis there and providing me with the necessary information and connections to gather data, and I thank all the participants of my research for their time.
Content
Abstract...1
1. Introduction ...4
2. Theoretical framework ...6
2.1 Teamwork development ...6
2.2 Psychological safety and team learning ...8
2.3 Engagement in team learning ...10
2.4 Professional differences ...11
3. Method ...13
3.1 Research design ...13
3.2 Case context: TeamSTEPPS™ ...13
3.3 Data collection ...14 3.4 Measures ...15 3.5 Data analysis ...16 4. Results ...19 4.1 Quantitative analysis...19 4.2 Qualitative analysis ...24 4.3 Conclusions ...27 5. Discussion ...29 5.1 Discussion ...29
5.2 Strengths and limitations ...31
5.3 Academic and practical contributions ...32
5.3 Conclusions and recommendations for future research ...33
Appendices ...40
Appendix A - Teamwork according to TeamSTEPPS™ ...40
Appendix B - Interview protocol ...42
Appendix C - Selected questions from the TAV-questionnaire ...44
Appendix D - Factor loadings ...47
Appendix E - Coding scheme ...50
Appendix F - Coding table ...51
Appendix G - Normality test ...60
Appendix H - Regression and mediation analysis ...70
Appendix I - T-test output: before and after training ...73
Appendix J - T-test output: comparing professions on psychological safety ...74
Appendix K - Two-way ANOVA output ...76
1. Introduction
Continuous improvement and working towards a learning organization are closely connected (Bessant et al., 1994). Over the last years, many organizations have tried to adopt a culture of continuous improvement, in order to streamline their processes, whether this is in manufacturing or in service. The effective improvement of all kinds of processes in an organization can help save a lot of time, money and other resources in dealing with reliability, availability,
maintainability and performance issues (Moubray, 1997; Singh and Singh, 2013).
In health care, people are always investigating ways to improve existing cures or treatments, and fight illnesses and diseases in order to save patients’ lives. This drive makes one expect that an excellent level of multidisciplinary teamwork in order to learn and thus improve would be part of the culture in hospitals. Nevertheless, in January 2013 the general Dutch Safety Board released a report about conclusions that could be drawn after eight years of studying all kinds of accidents in The Netherlands (Dutch Safety Board, 2013). One of the spearheads of this report are the frequencies of fatal and non-fatal avoidable accidents in health care. As one of the things that could be a cause of accidents like these, the autonomous working of physicians was named. This report stresses the importance of interprofessional collaboration and
communication for patient safety, but is not the first to do that. In 1999, a report was issued in the United States that already called attention to the subject (Kohn et al, 1999). In reaction to this report, the ‘Team Strategies and Tools to Enhance Performance and Patient Safety’ (TeamSTEPPS™) program was developed jointly by the Agency for Healthcare Research and Quality and the US Department of Defense Patient Safety Program, which started being implemented in the US from 2007 forward. This evidence-based program focuses on
communication and teamwork in health care, in order to ultimately improve performance and patient safety (Clancy, 2009). Since 2009, a Dutch version has been available, which is currently being tested in various health care institutions.
Various studies have shown that psychological safety is needed to engage in learning (Kahn, 1990; Edmondson, 1999; Nembhard and Edmondson, 2006; Garvin et al., 2008). Psychological safety entails the confidence that the team will not reject, or punish someone for speaking up (Nembhard and Edmondson, 2006). Psychological safety is needed for open communication, which indicates that it is necessary in teams. With teamwork, closed-loop communication and mutual trust are coordinating mechanisms that make it happen (Baker et al., 2005; Salas et al., 2005). Psychological safety can be influenced by differences in hierarchy or status. Within multiprofessional teams, different professionals are on different hierarchical levels, so within the team, there is a power imbalance (Lockhart-Wood, 2000). How do different professionals, working in one team, cope with the openness of communication that teamwork requires? In light of the power imbalance due to professional hierarchy, it can be expected that physicians and non physicians differ in feelings of psychological safety and react differently to teamwork training (Shekelle, 2002; Dekker, 2009; Sirriyeh et al., 2012).
Since improving patient safety is a topical issue (Kohn et al, 1999; Salas et al., 2005; Clancy, 2009; Dutch Safety Board, 2013), which indicates the need for engaging in team learning, and we know that the latter is positively influenced by psychological safety, which is also important for collaboration in teams, it is important to know if psychological safety and engagement in team learning can be improved by teamwork training. Also, possible differences between professions of different hierarchical stance can influence training outcomes. Therefore, the research question for this study is the following: How does professional hierarchy influence
psychological safety and the effect of teamwork training on psychological safety within (multi-)professional teams, and how do teamwork training and psychological safety affect the engagement in team learning of non physicians and physicians?
This research question was investigated by a longitudinal embedded single-case study, since an organizational context provides the best possibility to measure the variables and
relationships in an internally valid way. By using quantitative as well as qualitative data, a combination of analyses can lead to optimal results.
2. Theoretical framework
In this chapter, the theoretical framework concerning the constructs from the research question is developed, researching the existing literature. Following this framework, five hypotheses are specified, which are combined in a conceptual model (figure 1).
Figure 1: Conceptual model
2.1 Teamwork development
Research has shown that teams outperform individuals in a complex environment because they have more potential to offer for adaptability, productivity and creativity than any one individual can offer (Salas et al., 2005), which is necessary to cope with complex environments (Bigley and Roberts, 2001). According to Helmreich et al. (2001) teams make fewer mistakes than do individuals because of cross-checking and in-team assertiveness. This is why teamwork is more and more encouraged in health care environments. However, teamwork does not simply occur when people are put together (Alonso et al., 2006; Salas et a., 2005). ‘Rather than being
characterized by consensus and unity, teamwork can be as much defined by conflict, difference and struggles between competing sets of interests brought together within the group work situation’ (Finn, 2008, p. 105). It requires balancing opposing elements of integration and specialization both at the same time. It has been supported that one of the reasons for teams to fail is internal team processes (e.g. communication) (Salas et al., 2005; Gladstein, 1984;
In a perfect world, ‘teamwork in health care is characterized by consensus, cooperation and interdependency, as a function of complementary professional roles’ (Finn, 2008, p. 108). However, in practice health care teams are struggling to balance the fundamental contradiction of ‘an increasingly fragmented, specialized, professional division of labor’ and ‘mutual
interdependency that makes collaboration essential to achieving outcomes’ (Finn, 2008, p. 108). One important issue in this respect is that the professions of physician and nurse have
historically developed on a basis of professional hierarchy with different purposes, which leads to fundamentally different professional interests (Stein et al., 1990; Finn, 2008). This may make it difficult for these professions to operate together as part of a cohesive team in a
multiprofessional setting.
Teamwork research has shown that critical aspects of teamwork, like mutual performance monitoring, are linked together by specific coordinating mechanisms like closed-loop
communication and mutual trust (Baker et al., 2005; Salas et al., 2005). Alonso et al. (2006) developed an instructional model to train teamwork on the basis of these factors and found that mutual performance monitoring and communication are team skills that can be trained. These and other skills form the basis of the development of the TeamSTEPPS™ program (appendix A). The ultimate objective of the TeamSTEPPS™ program is to increase patient safety and improve performance ‘by reducing medical error through the development of a culture of safety and teamwork’ (Alonso et al., 2006, p. 408). Training outcomes are described by Alonso et al. (2006), when they explain that ‘attitudinal outcomes include mutual trust and team orientation’ (p. 408), knowledge outcomes include shared mental models, and performance outcomes ‘are characterized by changes in behavior and a change in results’ (p. 408). After thorough
assessment of a situation, an intervention can be created within the TeamSTEPPS™ curriculum which is tailored to the specific case.
performance monitoring is to create a learning process and that communication plays a substantial role in this.
This study builds on the idea that mutual trust is needed in communication and learning, and the changes in behavior that can be expected in cases of distrust, which will be explained in the next section.
2.2 Psychological safety and team learning
A learning process is explained as ‘a cycle of action and reflection – that is, doing and thinking, performing and conversing’ (Carroll and Edmondson, 2002, p.51; based on Argyris and Schön , 1996; Kolb, 1984). Senge (1990) explains team learning as mastering the practices of dialogue and discussion. This indicates the need for interaction between team members in order to be able to learn and improve. In the literature learning behaviors like seeking feedback, discussing and correcting errors are mentioned (Argyris and Schön, 1978, Edmondson 1999).
In a health care environment specifically, the need for collaborative learning is more and more prominent in multiprofessional teams. Nembhard and Edmondson (2006) agree with Finn (2008) when they explain that this is because of three trends: increasing knowledge, ongoing
specialization and increasing interdependence. But the health care environment poses specific barriers to improvement and learning. The accountability for patients’ lives and strong
hierarchical structure within and between professions make communication difficult, especially across professional boundaries (Nembhard and Edmondson, 2006; Institute of Medicine, 2003; Finn, 2008). It is acknowledged that physicians tend to operate autonomously as ‘captain of the ship’ (Dutch Safety Board, 2013; Shekelle, 2002). Also, training and education of physicians and non physicians in healthcare is often being done separately and not multiprofessionally, so collaboration is not being addressed as an essential element of their work. Therefore, collaborative learning requires a change of attitude of all team members involved.
354). This confidence has its roots in mutual respect and trust between team members. According to Edmondson (1999, p. 354), ‘team psychological safety involves but goes beyond interpersonal trust; it describes a team climate characterized by interpersonal trust and mutual respect in which people are comfortable being themselves’.
Zand (1972) shows in his experimental research with a high-trust and low-trust group that trust fosters open communication styles, constructive conflict resolution, and free information flow. The mechanism that trust enables knowledge sharing can be explained with social exchange theory, where trust is one of the central constructs (Luo, 2002). This theory suggests that ‘people participate in exchange behavior because they think that their reward will justify their cost’ (Liao, 2008, p. 1882). Social exchange is based on intangible costs and benefits, and can also be related to knowledge and learning. Edmondson (1999) explains that psychological safety at team level facilitates team learning because a team member does not have to fear rejection for making a mistake or bringing up errors. Garvin et al. (2008) also identify
psychological safety as one of the characteristics of the supportive learning environment. The link between interpersonal inferences in groups and factors like learning are substantiated by other authors (Argyris and Schön, 1978;Tyler and Lind, 1992). The connection between psychological safety and improvement was also established by Nembhard and Edmondson (2006) when they found a positive influence in their research of survey data from 23 neonatal intensive care units of psychological safety on engagement in quality improvement, since this requires being ‘emotionally open to giving and receiving feedback’ (Nembhard and Edmondson, 2006, p. 948) which can be interpersonally risky. Organizations should create a culture that embeds a climate of openness and commitment to education and development (Cabrera and Cabrera, 2002). On the one hand, cultural diverse teams, like multiprofessional teams, might be confronted with initial distrust, e.g. in case of prejudices that undermine the team members’ perceived trustworthiness (Hoegl and Muethel, 2007). On the other hand, social cognitive theory explains that personal, environmental and behavioral determinants are interconnected
(Bandura, 2012). Because environmental determinants are only one part of the equation, we assume that the psychological safety in the multidisciplinary teams is not that different from the feelings of psychological safety in the monodisciplinary teams.
which show that after a multiprofessional intervention, better communication and collaboration was found among participants. This leads to the following hypothesis for this study.
Hypothesis 1 (H1): Teamwork training will improve psychological safety.
2.3 Engagement in team learning
Kahn (1990) is the founding father of the construct of engagement. He explains that personal engagement is the employment and expression of the self in task behaviors that promote connections to work and others, being emotionally involved. Nembhard and Edmondson (2006, p. 948) defined engagement in the process of quality improvement as ‘being physically,
cognitively and/or emotionally connected to the improvement work’. They explain that
engagement is needed to overcome barriers to quality improvement. This is certainly the case with health care professionals in surgical teams where team members are working together on project basis, temporarily getting together for a specific surgical intervention on a individual patient, subsequently moving on to the next patient or other obligations (Tucker and
Edmondson, 2003). Time is precious in such a context and to invest it in an improvement effort must be an explicit choice. Engagement in team learning is the construct that covers this effort. In this study we define engagement as being cognitively, emotionally and physically involved. Team learning is defined as making a connection with other team members with the intention to learn from each other (like speaking up, giving and receiving feedback).
As explained in the previous section, psychological safety is needed in the learning process. Nembhard and Edmondson (2006) already show in their study that psychological safety is positively related to engagement in quality improvement. Also Kahn (1990) already mentions this relationship when he names safety one of the psychological conditions to be engaged. This study will try to support earlier findings by analyzing this hypothesis:
Hypothesis 2 (H2): Psychological safety is positively related to engagement in team learning.
In the section about psychological safety, it was explained that after a multiprofessional
mandatory, and effort is expected to be made, we expect that training also influences engagement in team learning directly.
Hypothesis 3 (H3): Teamwork training will increase engagement in team learning.
2.4 Professional differences
As already mentioned before, the physician and nursing professions have historically developed differently. Class difference and gender issues have separated them in the past (Stein et al., 1990; Hall, 2005, Sirriyeh et al., 2012). Since the industrial revolution, nursing shifted from untrained and mostly unpaid work of women to a profession discipline based on legislation of protected licensing. Education in the profession of physicians originates from the 9th century in Italy. The medical profession was led by male professionals that had received education at universities and passed examinations. This was expensive and universities were only
accessible to men. Women were encouraged to go into nursing, as this would suit the virtues of womanhood (Stuart, 1993). The collaboration of physicians and nurses has been subject of many articles and discussions over decades, leading to legislation about medical activities (Roodbol, 2005). Nowadays, more women become physicians, and more men become nurses. Gender and class issues have diminished due to the emancipation of the nursing profession, but are still the source of different subcultures that can hinder collaboration (Hall, 2005; Sirriyeh et al. 2012). Physicians are trained to take the lead in diagnosis and therapy and assume responsibility for clinical decisions (Hall, 2005). Also, a power imbalance exists between physicians and non physicians; the physicians being the more powerful ones (Lockhart-Wood, 2000). They can influence the career of non physicians, being higher up the hierarchical chain. These issues have an impact on psychological safety (Nembhard and Edmondson, 2006). Since physicians have more power than non physicians, we expect that in general, the psychological safety of physicians may always be higher than that of non physicians.
Hypothesis 4 (H4): The psychological safety of physicians is higher than that of non physicians.
supporting quality improvement programs. They receive these interventions as an opportunity to blame them as ‘captain of the ship’ if something goes wrong with a patient, and feel that this is just another time absorbing element that takes them away from their other responsibilities (Shekelle, 2002). It is difficult to find a balance between accountability and learning (Dekker, 2009; Sirriyeh et al., 2012). Physicians are not only held accountable for errors by team
members, but also by their superiors, and even by people from outside the hospital. This is why psychological safety for physicians not only depends on factors within the team, but is also influenced by external factors, which is less so for non physicians. Teamwork training is only aimed at internal factors and, we expect, will therefore have a lesser effect on physicians that on non physicians. This leads to assume that professional hierarchy will influence the effect of teamwork training on psychological safety. We expect that non physicians’ psychological safety will increase more than that of physicians.
Hypothesis 5 (H5): The effect of teamwork training on psychological safety is higher for non physicians than for physicians.
3. Method
After introducing the theoretical framework of this study in the previous chapter, this chapter focuses on the research methods used in this study. First, the research design is described and the context of the embedded single-case. Then, quantitative and qualitative data collection and measures are clarified, and finally the analysis of the data is explained.
3.1 Research design
For this study, a longitudinal embedded single-case research design was developed (Yin, 2009), studying physicians and non physicians in the same context. An organizational context provides the best possibility to measure the influence of teamwork training on psychological safety and engagement in team learning, because this cannot be genuinely simulated in an artificial environment. Also, the differences in feelings of psychological safety between non physicians and physicians and the effect of teamwork training on these feelings, possibly moderated by professional hierarchy, are difficult to reproduce in an artificial setting. The actual effect of the teamwork training on patient safety will not be assessed.
3.2 Case context: TeamSTEPPS™
This study took place in one of the largest academic hospitals in the Netherlands, which has three core businesses; health care, education and research. It employs over 10.000 employees, of which 84 are dedicated to the department of cardiothoracic surgery, which received the TeamSTEPPS™ training.
In the Netherlands there is a lot of attention for quality and patient safety in health care (Dutch Safety Board, 2013). This academic hospital also focuses on these issues. In the perioperative pathway, the hospital introduced instruments to improve quality and safety, and it is presently searching for further appropriate ways to create a culture of continuous improvement. Because of this, a training program to improve collaboration, mutual performance monitoring and
Time
safety issues before the patient leaves the operating room. Furthermore, if one of the team members deems it necessary, a debriefing can be held by the whole team after the patient is transported to another care unit. After the training these procedures were implemented.
Before introducing the TeamSTEPPS™ intervention and the Sign-Out and debriefing procedures throughout the hospital, it was decided to do a pilot training and introduction of procedures with the department of cardiothoracic surgery. This pilot is now being evaluated. This multidisciplinary team exists of 84 employees from five different disciplines; surgeons, anesthesiologists, OR-nurses, anesthesiology assistants, and perfusionists. The first two disciplines can be categorized as physicians, the latter three categorize as non physicians. Each discipline had their own team interventions, focusing on interdisciplinary teamwork and team climate, aiming to enhance a learning climate between peers. The monodisciplinary interventions were tailored per team based on a team climate assessment. All team members were also expected to attend two multidisciplinary sessions about teamwork, communication and the introduction of the Sign-Out and debriefing procedure. These sessions were set up so that employees were mixed together in multidisciplinary groups, representing all five disciplines involved. Besides that, physicians had private coaching sessions on leadership development. After the last training sessions, the Sign-Out and debriefing procedures were officially
implemented. The timeline of the whole of the intervention is depicted in figure 2.
Figure 2: Timeline of the TeamSTEPPS™ intervention
3.3 Data collection
A questionnaire was issued to each of the five monodisciplinary teams, to all team members (35 physicians and 49 non physicians), in order to get quantitative data on psychological safety and engagement in team learning at two different times (before and after teamwork training), and test the hypotheses that were presented in the previous chapter. To back up the quantitative analyses and gain more insight into how these factors are connected, twelve additional
semi-February - June 2012: TAV-questionnaire September / October 2012: TeamSTEPPS Training April/May 2013: TAV-questionnaire and interviews November 2012:
structured interviews were held. This is how in depth information can be gathered about the mechanisms causing certain changes, keeping the focus on research factors, but giving the informant enough space to provide insights and elaborate. Also, by measuring in two different ways, methodological triangulation improves the reliability of the results. This study is based on data that was recovered from both TAV-questionnaires and the interviews, using a concurrent triangulation strategy (Creswell, 2003).
As shown in the project timeline (figure 2), before the training started the TAV-questionnaire was issued digitally to the 84 employees of the department of cardiothoracic surgery, divided by discipline. Six months after the introduction of the Sign-Out and debriefing, the
TAV-questionnaire was issued again and interviews were held.
Previous to forming the final interview protocol (appendix B), two pilot interviews were held to assess the quality of the protocol. This did not lead to changing the questions, but to the awareness of the interviewer that it was necessary to be flexible in stating the questions and to go along with the interviewees when they made a sidestep from the Sign-Out subject.To get a thorough view from all viewpoints, the interviewees were selected using stratified convenience sampling: subgroups were formed on the basis of discipline, and within the subgroups the final selection was made on basis of availability, because of irregular working hours due to surgery schedules. From all disciplines two or three employees were interviewed, twelve in total. These constitute five interviews with physicians and seven interviews with non physicians (see table 1).
Physicians Non physicians
surgeons 2 OR-nurses 3
anesthesiologists 3 anesthesiology assistants 2
perfusionists 2
Total 5 Total 7
Table 1: Overview of interviews held
3.4 Measures
Psychological safety was measured in the TAV-questionnaire with 8 items on a 5-point Likert-scale to cover the construct on feelings of respect, trust, support and feeling secure to speak up in the monodisciplinary team. These items entailed questions like ‘Team members pay attention to the needs and feelings of other team members’, and ‘Team members trust each other’. In the interview protocol, questions to assess the extent to which respondents feel safe to speak up about issues or ideas regarding teamwork in the multiprofessional team were included.
Engagement in team learning was measured in the TAV-questionnaire with 6 items on a 5-point Likert-scale to cover the construct on how employees are physically, cognitively and/or emotionally involved in team learning in their monodisciplinary team. These items entailed questions like ‘The team is cooperating in a good way’ and ‘The team shows a continuous improvement curve’. The extent to which team members are engaging in team learning in the multiprofessional team will also be measured through certain questions during the interviews about the execution of the Sign-Out and debriefing surgery evaluation tools, representing team learning. This will show how physically, cognitively and/or emotionally connected to the learning process they are in doing this.
The division in professional hierarchy was made by gathering the TAV-questionnaires by discipline. Previous to starting the interviews, also the concerning discipline was noted. This information will be used to analyze the difference in psychological safety between physicians and non physicians, in which physicians are assumed to have a higher position than non physicians in the professional hierarchy (Lipworth et al., 2013).
3.5 Data analysis
TAV-questionnaires
The quantitative data from the TAV-questionnaires was analyzed using SPSS software (v. 20). Descriptive statistics, such as frequencies, means and standard deviations were used to
Data reduction
The internal validity was tested by executing a factor analysis of both constructs (Hair et al., 1998). Factor loadings need to meet certain criteria. According to Hair et al. (1998), factor loadings higher than .5 have practical significance. Since this research has a relatively small sample size, it accepted loadings of .6 or higher. The goal of the factor analysis was to obtain theoretically meaningful constructs that are likely to be correlated, so an oblique rotation was used: oblimin (Hair et al., 1998). The questions that were pulled from the questionnaire were first analyzed together, but since psychological safety and engagement in team learning are closely correlated, the two constructs did not show in the exploratory factor analysis (appendix D). All questions loaded onto one factor. A confirmatory factor analysis with two factors also did not provide satisfactory results. Therefore, as a last resort, both constructs were analyzed separately, which did provide satisfactory factor loadings (appendix D). Also, the Bartlett’s test of sphericity, which indicates the significance of the hypothesis of items not belonging together, and KMO measure of sampling adequacy, which needs to be higher than .6, were calculated to assess the strength of the relationship among items within a construct.
The reliability of the constructs was tested by calculating the Cronbach’s alpha coefficients, assessing the internal consistency of the group of factors per construct from the factor analysis. In order to have acceptable reliability, the Cronbach alpha needs to be at least .7 (Hair et al., 1998). A value close to 1.0 could indicate too much overlap of the questions, so ideally a value between .8 and .9 should be calculated.
Hypothesis testing
Hypothesis testing was done by performing independent samples T-tests to measure
Interviews
Validity was examined by performing two pilot interviews. As mentioned before, this did not lead to changing the questions, but to the awareness of the interviewer that it was necessary to be flexible in stating the questions and to go along with the interviewees when they made a sidestep in answering the questions. The interviews were held and transcribed in Dutch and read by the interviewee to see if this was an accurate representation of what was meant. If necessary, alterations were made. A coding scheme was made that specified the constructs of psychological safety and engagement in team learning and their subcategories, and indicated when to label a remark with a certain code (appendix E). Following the steps of Creswell (2003), the interviews were then read again to get a general impression about the information that they could deliver. After that, the transcriptions were coded manually and outcomes were put in a table, divided by subcategory (appendix F). These outcomes were then examined and interpreted to find meanings and patterns that would explain the research question.
4. Results
This chapter shows the results of the various analyses that were performed in order to be able to accept or reject the hypotheses. It starts with the descriptive statistics of the data followed by the results of the analyses in order to answer the research question of this study.
4.1 Quantitative analysis
Data reduction
The factor analyses were done for both constructs separately, leading to good factor loadings and KMO and Bartlett scores. None of the items had to be deleted. The factor loadings are reported in appendix D, the KMO and Bartlett scores are depicted in table 2.
Psychological safety Engagement in team learning
Number of items 8 6
KMO measure of sampling adequacy .933 .861
Bartlett’s test of sphericity .000 .000
Table 2: Strength of the relationship among items
After determining the significance of the Shapiro-Wilks test, not all data showed to be normally distributed (appendix G). This did not lead to an alteration of the plan for hypothesis testing, since skewness and kurtosis remained lower than 1 at all times and the sample size is not small. Lumley et al. (2002) showed that the t-test and regression test are robust toward non normal data. The descriptive statistics showing correlations, sample sizes, response rates, means, standard deviations and Cronbach alphas are shown in tables 3 and 4.
Descriptive statistics
Table 3 and 4 show the descriptive statistics of the data and correlations between variables.
Descriptive statistics and correlations
mean SD 1 2 3 Cronbach α
1. Training 1.42 .50
2. Psychological safety 3.69 .80 .27** .928
3. Engagement in team learning 3.83 .66 .26** .89** .879
4. Professional hierarchy 1.27 .45 -.11 -.27** -.20*
N=122
Professional hierarchy: physicians = 2, non physicians = 1 *. Correlation is significant at the 0.05 level (2-tailed). **. Correlation is significant at the 0.01 level (2-tailed).
Table 3: Descriptive statistics and correlations between variables
Additional descriptive statistics
Physicians Non physicians
n response rate mean SD n response rate mean SD 1. Psychological safety before training 22 62.8% 3.15 0.75 49 100% 3.68 0.86
2. Engagement in team learning
before training 22 62.8% 3.50 0.62 49 100% 3.76 0.70
1. Psychological safety
after training 11 31.4% 3.73 0.53 40 81.6% 4.00 0.65
2. Engagement in team learning
after training 11 31.4% 3.83 0.47 40 81.6% 4.08 0.59
1. Total psychological safety
(before and after training) 33 47.1% 3.34 0.73 89 90.8% 3.82 0.78 2. Total engagement in team
learning
(before and after training)
33 47.1% 3.61 0,59 89 90.8% 3,90 0.67
Hypothesis testing
H1: Teamwork training will improve psychological safety.
To test if the intervention did improve psychological safety, an independent samples T-test with data from all respondents was done, since the data was received anonymously and could not be paired. The output of this test is shown in appendix I, and the results are reported below.
The hypothesis is supported.
H2: Psychological safety is positively related to engagement in team learning
The effects of psychological safety on engagement in team learning were tested with a multiple regression analysis that included all measures, before and after the training. In the regression, both psychological safety and the training were entered as predictors. The output of the regression test is shown in appendix H (step 3), the results are reported below.
The hypothesis that psychological safety is positively associated with engagement in team learning is supported.
Psychological safety and training explained a significant proportion of variance in engagement scores (R2 = .80, F(2, 119) = 41.79, p < .0005). Psychological safety scores significantly predicted engagement in team learning scores with the following formula: Engagement in team learning = 1.07 + 0.73 x psychological safety (p < 0.0005).
H3: Teamwork training will increase engagement in team learning
Since the data was received anonymously and could not be paired, an independent samples T-test with data from all respondents was done to T-test if engagement in team learning before and after the training had improved. After that, a mediation analysis was done, to determine if the training influenced engagement in team learning directly. Because the testing of hypothesis 1 showed the increase of psychological safety, and testing of hypothesis 2 showed a relationship between psychological safety and engagement in team learning, it is possible that although engagement in team learning increased after teamwork training, this is (partly) mediated by psychological safety.
The output of these tests and steps of the mediation analysis are shown in appendices H and I, the results are reported below.
Mediation analysis
Engagement in team learning
Step 1 Step 2 Step 3
Training .35* .03
Psychological safety .43* .74*
R² .07* .07* .80*
Adjusted R² .06* .06* .79*
*. Regression is significant at the 0.01 level
Step 1 – multiple regression with dependent (engagement in team learning) and independent variable (training) Step 2 – multiple regression with mediating factor (psychological safety) and independent variable (training) Step 3 – multiple regression (training and engagement in team learning) including mediating factor (psychological safety
Table 5: Mediation analysis
Although the t-test shows a significant increase of engagement in team learning, the mediation analysis reported that this effect is fully mediated by psychological safety, since the influence of teamwork training is no longer significant when psychological safety is added in the regression (B-values, table 5). Therefore, the hypothesis is rejected.
H4: The psychological safety of physicians is higher than that of non physicians
To test if psychological safety was higher for physicians than for non physicians, independent samples T-tests were done, since the data was received anonymously and could not be paired. The output of the independent samples T-tests is shown in appendix J, the results are reported below.
The hypothesis is rejected, because both before and after the training, non physicians scored psychological safety higher than physicians did.
Before the training:
The mean on psychological safety of non physicians is 3.68, and that of physicians 3.15. The output of the independent samples T-test on the mean scores on
psychological safety between professions shows that before the training the difference is significant (t(69) = -2.512, p = .014).
After the training:
The mean on psychological safety of non physicians is 4.00, and that of physicians 3.73. The output of the independent samples T-test on the mean scores on
psychological safety between professions shows that after the training the difference is not significant (t(49) = -1.274, p = .209).
Both periods together:
The mean on psychological safety of non physicians is 3.82, and that of physicians 3.34. The output of the independent samples T-test on the mean scores on
H5: The effect of teamwork training on psychological safety is higher for non physicians than for physicians
The two-way ANOVA was done on all data, with professional hierarchy and teamwork training as between-subject factors. The output is shown in appendix K. The test determined that there was not a significant difference in the effect of teamwork training on psychological safety between physicians and non physicians (F(1, 118) = .654, p = .420). Physicians’ psychological safety did in fact increase more than that of non physicians, so the hypothesis is rejected.
In the next section, the interviews will be analyzed to be able to compare the results of the quantitative analyses with what was said by the employees.
4.2 Qualitative analysis
This section studies the coded texts from the interviews and tries to find out how the results from the quantitative analyses can be explained. The original Dutch quotes that were translated can be found in appendix L.
The effects of teamwork training on psychological safety
Non physicians
Employees differed in their opinion about how the TeamSTEPPS™ intervention changed psychological safety. On the one hand, it was said that ‘this way, it creates an extra dimension.
You are being taken seriously in your profession’ (#6), and on the other hand: ‘It’s mostly the same ones that communicate. I don’t think you can change people that quickly’ (#10). Several
persons said something about how the difference in professional hierarchy is getting smaller. Within the teams, people address each other by first name. ‘That difference in status that was
there before, that does change. People are becoming more equal, you can address anyone’ (#2). But still this person also claimed that ‘I can say something to one person, but not to another. I think I may get into trouble then. During the intervention it was said that it should be possible, but in practice it is difficult’ (#2).
Things that are mentioned by physicians are that positive things are being addressed now as well, and people cooperate better: ‘It went from ‘no’ and having to force a ‘yes’, to a positive
‘yes’. So if you approach people for favors, that’s much better now’ (#3). The influence of social
pressure was mentioned when someone said ‘You have to keep that balance: respecting each
other and having a good relationship, as friendly as possible, but still having the freedom to address one another when things go wrong’ (#9). Also it was mentioned that ‘It is still the case that some persons do not get along together’ (#4).
The effects of teamwork training on engagement in team learning
Non physicians
Non physicians agreed to physical involvement in (the actual execution of) the Sign-Out
procedure. It is seldom forgotten, only in very specific circumstances. The debriefings are rare, but executed when necessary. They could also see the effects of the Sign-Out procedure: ‘You
notice that it is checked before the next surgery’ (#7) and ‘I think that same day, the purchase order was made. That makes it worth to do it, not annoying, but a good thing’ (#1).
The feedback during the Sign-Outs mostly covers technical issues, not personal feelings: ‘I
never mentioned something during the Sign-Out like: I did not feel treated well today, or I couldn’t act the way I wanted today. Not because it didn’t happen, but because I didn’t think it would solve anything’ (#8).
All non physicians that were interviewed believed the Sign-Out to be a good thing and supported the introduction of the procedure. They said things like ‘[The Sign-Out] improves
mutual understanding, for each other and for the situation’ (#5) and ‘It has to improve safety, and I think you can capture most things this way’ (#7). They also supported the introduction of
the Sign-Out and debriefing procedures at other departments: ‘I think that wherever, you have to
want to do this, no matter what’(#6), and ‘I would recommend it’ (#7). They posted one concern,
that of too much bureaucracy and to many checklists. More is not always better, it could actually be the reverse. The way the procedures are designed now is well received.
Physicians
Physicians also agreed to the physical involvement in the Sign-Out procedure, incidental forgetting and rare execution of the debriefings. ‘From what I see, it has become routine.
of non-execution: ‘But he had seen enough, so he thought: I’m through and walking out of here’ (#11) and ‘Sometimes it is hard, because some procedures take 12-14 hours, so the people
involved at the start are others than the ones at the finish’ (#11).
They also agreed with the purpose of the Sign-Out and debriefings when they said, among other things, ‘The fact that others around you can give feedback immediately is an element of
safety’ (#9), ‘Not often, but every now and then we agree: we have to do this differently next time’, ‘In the end it is in the interest of patient safety’ (#11), and ‘It is mainly meant to prevent fuss in the hallway, and that sort of thing. To clear the air, because people are talking about that among each other’ (#12). Most of them were not really enthusiastic, but do it because it is the
right thing to do, as was explained: ‘I don’t really like it, but it needs to be done’ (#11). They also agreed on introducing it to other departments.
The effects of professional hierarchy on psychological safety and the training effects
Non physicians
Some mentioned that professional hierarchy could play a role, but not with every physician. ‘There is a different [physician] every time, most of the time you really team up, but there are
certain types of [physicians] that work more in a hierarchical way. Then there is more…the [physician] gives orders instead of engaging in a dialogue’ (#5). Also, the way the physicians
engage the nursing team members: ‘really showing interest, looking at you and asking: did
something special happen in your opinion?’ (#8). This was mentioned by some, but others said
that this did not influence them at all. They felt as good as everyone else, but acknowledged the different role they had in comparison to physicians during surgery. Everyone has their own responsibilities, but in the end, the physician decides what happens.
Physicians
Physicians acknowledged that sometimes it is hard for non physicians to speak up. ‘They
already had the right, but were sometimes afraid to use it’ (#3). Some expected this to be easier
now, because of the multiprofessional intervention emphasizing this right and indicating equality in this procedure. One physician said that ‘hierarchy, I think, has a lot to do with making people
feel that they can say what they want to’ (#12). Another physician pointed out that feedback by
physicians is desired, because ‘the higher you position is hierarchically, the harder it is to get
In the next section these findings will be combined with the findings from the quantitative analyses.
4.3 Conclusions
This section combines the findings from the quantitative and qualitative analyses and articulates the conclusions that can be drawn in order to answer the research question of this study.
Psychological safety
Psychological safety was improved after the teamwork training. Professional hierachy
differences are diminishing, but it is hard to forget the past and start over fresh. Also a person’s character has an influence on feelings of safety, and this was hardly changed by the training.
Engagement in team learning
Engagement in team learning has increased. Testing showed that this was due to the increase in psychological safety, and not so much a result of the teamwork training directly. This
connection between psychological safety and engagement in team learning was not mentioned often in the interviews. Even when asked directly, most team members answered in ways that indicated that they were not afraid of reprisals and would speak out anyway. Only one or two indicated to fear the reaction of some physicians, or pointed out the uselessness of speaking out, because it would not change anything. The intervention has succeeded with the
introduction of the Out and debriefing procedures. It is commonplace to execute the Sign-Outs now; the debriefings are rare. Most physicians and non physicians acknowledged the usefulness, which led to support and recommendation of the Sign-Out and debriefings to other departments.
Influence of professional hierarchy on psychological safety
non physicians scored higher on feelings of psychological safety and a significant difference in increase of psychological safety between the two professions after the training was not found.
5. Discussion
This chapter will answer the question of how teamwork training affects psychological safety and engagement in team learning of physicians and non physicians. Also, the way that professional hierarchy influences psychological safety and the effect of teamwork training on psychological safety within the (multi-)professional teams is assessed, and how psychological safety affects the engagement in team learning of physicians and non physicians. It will discuss the strengths and limitations of these conclusions and how these were handled. Furthermore it will address the academic and practical contributions of this study and make recommendations for further research.
5.1 Discussion
This section discusses every part of the research question separately, on a conceptual level, relating it back to the existing literature.
Effects of teamwork training on psychological safety
Since teamwork training aims to increase trust, it was expected that teamwork training would therefore increase psychological safety. This was supported by the findings of Vazirani et al. (2005), when they found that after a multiprofessional intervention, communication and
collaboration among participants improved. The data of this study supported the hypothesis and reported that the teamwork training improved psychological safety by creating a platform to give and receive feedback. As explained by social cognitive theory, personal, environmental and behavioral determinants are interconnected (Bandura, 2012). Personal determinants, like a person’s character, are also part of the equation and these are hard to change.
There will always be people that speak up more easily than others. This also shows from the data that came out of the interviews. It is mostly the same ones that speak up.
Effects of psychological safety on engagement in team learning
Nembhard and Edmondson (2006) already showed that psychological safety is positively related to engagement in quality improvement. It was also established in literature that
psychological safety improved engagement in team learning. The fact that this was not mentioned in the interviews could be because this would mean that people would have to concede to the weakness of not feeling confident enough to speak up, which is against the nature of their professional identity (Weaver, 2012).
Effects of teamwork training on engagement in team learning
Since after a multiprofessional intervention, communication and collaboration among participants improve (Vazirani et al., 2005), and in mandatory teamwork training the effort is made compulsory, we expected that teamwork training would also influence engagement in team learning directly. In this study, psychological safety fully mediated the increase of engagement in team learning. The data from the interviews showed acknowledgement of the usefulness of the new patient safety tools, which led to support and recommendation of the Sign-Out and debriefings to other departments. It could be that this was not due to the
intervention, but that this need was already felt before the training started. In that case, one of the pillars of engagement, feeling the need (Kahn, 1990), would not have been effected by the training. This would explain the extent of the mediating effect of psychological safety. All this shows physical, cognitive and emotional involvement. It is to be expected that if the training had been done without introducing official procedures as the Sign-Out and debriefing, the
engagement in team learning would decrease again after a while. From the data that came from the interviews, it can be derived that the Sign-Out and debriefings are mainly performed
because they are compulsory. If not, the improved situation would probably over time return to the former circumstances, abandoning the newly introduced procedures. In the literature, studies that assess if the long term effects of communication training could not be found at this time.
Effects of professional hierarchy on psychological safety
Differences in professional hierarchy and social status indicate power imbalance between the professions (Lockhart-Wood, 2000) and have an impact on psychological safety (Nembhard and Edmondson, 2006). This led to the proposition that in multiprofessional teams, the
the before and after training measurement. The fear of loss of status from physicians could be playing a role in scoring lower than non physicians (Shekelle, 2002; Dekker, 2009; Sirriyeh et al., 2012). Zwaan et al. (2013) studied adolescent behavior in relation to status and their results seemed ‘to indicate that high-status adolescents feel a greater need to use aggression to re-establish a clear status hierarchy when their position is more likely to be threatened’ (p. 215). This shows the protectiveness of high-status individuals of their status. It may thus be expected that high-status professionals will be more reserved towards discussing errors (Dekker, 2009), which is critical for experiential learning and improvement (Argyris and Schön, 1978).The higher increase of psychological safety among physicians could be because of the fear of physicians of consequences for their job they had when the teamwork training started. Some indicated there was an expectation that the outcome of the training would influence the assessment of their performance, and, therefore, threaten their job security. The basis of this was a IGZ-rapport that ordered the hospital to improve patient safety. After the training, it was clear that jobs were not threatened, so physician’s feelings of safety were increased more than those of non physicians, who had not experienced that fear.
5.2 Strengths and limitations
This study is subject to several strengths and limitations with respect to internal and external validity.
Internal validity
A longitudinal study can be influenced by significant events or other changes outside of the intervention that occur between the first and second measurement. At this time none of these issues are known to the researcher. Also, the interviews were held six to seven months after the training sessions of the intervention, which made it for some employees hard to recall what the initial effect was on team work. This may have influenced the results.
Control variables like age and gender were also not measured and could therefore not be taken into account. Gender could be of influence, since the origin of the professions had a strong gender division. Age could be of influence, because differences in hierarchical positions used to be more excepted and emphasized than it is now. Therefore, gender and age could be bases of different subcultures with its own characteristics. This could not be compensated for, and
therefore it could have influenced the outcomes of this research.
Also, the interviews were held and coded by one researcher. This could lead to bias and
influence the interviews and the outcomes of the coding procedure. This was limited by creating the interview protocol and coding scheme, making the procedure as transparent as possible.
External validity
This research was based on one case, and therefore, it is difficult to generalize the outcomes to the whole of the health care field and all teamwork training programs. Also, the second
measurement of the TAV-questionnaire had a very low response rate from physicians (11 out of 35). This could indicate a selection bias (if opposing persons were not responding), making the outcomes less representative for the whole of the department. Still, since a whole department was included in the research and interviews were held with representatives from all disciplines, the outcomes may have a predictive value for similar interventions under similar conditions in the future and provide insight in the mechanisms that are present in such multiprofessional team work. It indicates how these mechanisms can be influenced and also points out the limitations of team work training designed in this way.
5.3 Academic and practical contributions
This research makes a number of contributions to the academic field. First, it supports the earlier supported relationship between psychological safety and engagement in team learning (Nembhard and Edmondson, 2006). Second, this seems to be the first study to assess the effect of the TeamSTEPPS™ teamwork training on psychological safety and engagement in team learning. Third, in contrast of what was thought, it found that higher professional hierarchy does not automatically lead to higher feelings of psychological safety.
relatively comfortable in their position and do not desire the same hierarchical stance as physicians do. It shows that physicians have lower feelings of psychological safety and were fearing for their job security. They could therefore be as vulnerable, or even more so, than non physicians. This means that with future training or interventions like this, more attention should be given to what influence changes may have on the position of all team members, not only for the ones in the lower positions of professional hierarchy but also for the ones with the higher positions. In the hospital of this study, attention should be given to how the purpose of the intervention is communicated. Job security should not be an issue for employees while participating in teamwork training., since this does not support open communication and collaboration.
5.3 Conclusions and recommendations for future research
This research found that teamwork training did improve psychological safety, which is important to know, because it is the basis of successful learning (Nembhard and Edmondson, 2006). This statements was supported by the fact that psychological safety in turn fully mediated the
increase of engagement in team learning. Teamwork training did not affect engagement in team learning directly however. Non physicians showed to feel higher psychological safety than physicians did and teamwork training affected both groups in a similar way. This shows that it is not always the non physicians that need attention because of supposed insecurity due to being victim of hierarchical issues. A full assessment is needed to determine the strengths and weaknesses of all groups in a multiprofessional teamwork intervention.
References
Alonso A, Baker DP, Holtzman A, Day R, King H, Toomey L, Salas E (2006). Reducing medical error in the Military Health System: How can team training help? Human Resource Management
Review, Vol. 16, p 396-415.
Argyris C and Schön D (1978). Organizational learning: A theory of action perspective. Reading, MA: Addison-Wesley.
Baker DP, Salas E, King H, Battles J and Barach P (2005). The role of teamwork in the
professional education of physicians: Current status and assessment recommendations. Journal
on Quality and Patient Safety, Vol. 31, Issue 4, p 185-202.
Bandura A (2012). On the functional properties of perceived self-efficacy revisited. Journal of
Management, Vol. 38, issue 1,p 9-44.
Baron R and Kenny D (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of
Personality and Social Psychology, Vol. 51, p 1173–1182.
Bessant J, Caffyn S, Gilber J, Harding R and Webb S (1994). Rediscovering continuous improvement. Technovation, Vol. 14, issue 1, p 17-29.
Bigley GA and Roberts KH (2001). The incident command system: High-reliability organizing for complex and volatile task environments. Academy of Management Journal, Vol. 44, Issue 6, p 1281-1299.
Bouckenooghe D, Devos G, Van den Broeck H (2009). Organizational change Questionnaire - climate of change, processes, and readiness: development of a new instrument. The Journal of
Psychology, Vol. 143, Issue 6, p. 559-599.
Carroll JS and Edmondson AC (2002). Leading organizational learning in health care. Quality
and Safety In Health Care, Vol. 11, Issue 1, p 51-56.
Clancy CM (2009). Ten Years After To Err Is Human. American Journal of Medical Quality, Vol. 24, p. 525-528.
Creswell JW (2003). Research design. Qualitative, Quantitative and mixed methods approaches
(second edition). Thousand Oaks, CA, US: Sage Publications Inc.
Dekker, SWA (2009). Just culture: Who gets to draw the line? Cognition, Technology & Work, Vol. 11, Issue 3, p 177-185.
Dutch Safety Board (2013). Veiligheid in perspectief – acht jaar ongevalsonderzoek door de
Onderzoeksraad voor Veiligheid 2005-2012, Den Haag.
Edmondson AC (1999). Psychological safety and learning behavior in work teams.
Administrative Science Quarterly, Vol. 44, Issue 2, p 350-383.
Finn R (2008). The language of teamwork: reproducing professional divisions in the operating theatre. Human Relations, Vol. 61, Issue 1, p 103-130.
Freeman M, Miller C and Ross N (2000). The impact of individual philosophies of teamwork on multi-professional practice and the implications for education. Journal of Interprofessional Care, Vol. 14, Issue 3, p 237-47.
Garvin DA, Edmondson AC and Gino F (2008). Is yours a learning organization? Harvard
Business Review, Vol. 86, Issue 3, p 109-116.
Gladstein DL (1984). Groups in context: A model of task group effectiveness. Administrative
Science Quarterly, Vol. 29, p 499-517.
Hair JF, Tatham RL and Anderson R (1998). Multivariate Data Analysis. Englewood Cliffs, NJ: Prentice Hall.
Hall P (2005). Interprofessional teamwork: professional cultures as barriers. Journal of
Interproffesional Care, Vol. 19, suppl. 1, p 188-196.
Helmreich RL, Wilhelm JA, Klinect JR and Merritt AC (2001). Culture, error, and Crew Resource Management. In Salas E, Bowers AC and Edens E (Eds.), Improving teamwork in organizations (pp. 305−331). Hillsdale, NJ: Erlbaum.
Hoegl M and Muethel M (2007). Shared leadership in dispersed innovation teams: mutual influence and proactive followership. Academy of Management Annual Meeting Proceedings, 2007, p 1-6.
Institute of Medicine (2003). Keeping patients safe: Transforming the work environment of
nurses. Washington DC: National Academy Press.
Johnson AM and Lederer AL (2005). The effect of communication frequency and channel richness on the convergence between chief executive and chief information officers. Journal of
Management Information Systems. Vol. 22, p 227-252.
Kahn WA (1990). Psychological conditions of personal engagement and disengagement at work. Academy of Management Journal, Vol. 33, 692–724.
Kohn LT, Corrigan JM, and Donaldson MS (1999). To err is human: building a safer health
system. Washington, DC: National Academy Press.
Liao L-F (2008). Knowledge sharing in R&D departments: a social power and social exchange theory perspective. The International Journal of Human Resource Management, Vol. 19, Issue 10, p 1881-1895.
Lockhart-Wood K (2000). Collaboration between nurses and doctors in clinical practice. British
Journal of Nursing. Vol. 9, Issue 5, p 276-280.
Lumley T, Diehr P, Emerson S, and Chen L (2002). The importance of the normality assumption in large public health data sets. Annual Review of Public Health, Vol. 23, p 151-169.
Luo X (2002). Trust production and privacy concerns on the internet: a framework based on relationship marketing and social exchange theory. Industrial Marketing Management, Vol. 31, p 11-118.
McIntyre RM and Salas E (1995). Emerging principles from complex environments. In Guzzo RE and Salas E(Eds.), Team effectiveness and decision making in organizations (p 9−45). San Francisco: Jossey-Bass.
Nembhard IM and Edmondson AC (2006). Making it safe: The effects of leader inclusiveness and professional status on psychological safety and improvement efforts in health care teams.
Journal of Organizational Behaviour, Vol. 27, p 941-966.
Roodbol PF (2005). Dwaallichten, struikeltochten, tolwegen en zangsporen: onderzoek naar taakherschikking tussen verpleging en artsen. Dissertation; Rijksuniversiteit Groningen, Faculty
of Economics and Business.
Salas E, Sims DE and Burke CS (2005). Is there a ''Big Five'' in Teamwork? Small Group
Research, Vol. 36, p 555.
Schaufeli WB, Salanova M, Gonzáles-Romá V and Bakker AB (2002). The measurement of engagement and burnout: A two sample confirmatory factor analytic approach. Journal of
Happiness Studies, Vol. 3, p 71–92.
Schein EH and Bennis W (1965). Personal and organizational change via group methods. New York: Wiley.
Shekelle PG (2002). Why don’t physicians enthusiastically support quality improvement programmes? Quality and Safety In Health Care, Vol. 11, Issue 1, p 6.
Sirriyeh R, Lawton R, Armitage G, Gardner P and Ferguson S (2012). Safety subcultures in health-care organizations and managing medical error. Health Services Management Research, Vol. 25, Issue 1, p 16-23.
Stein LI, Watts DT and Howell T (1990). The doctor-nurse game revisited. The New England
Journal Of Medicine, Vol. 322, Issue 8, p 546-549.
Stuart ME (1993). Nursing the endangered profession? The Canadian Nurse, Vol. 89, p 19-22.
TeamSTEPPS™ website: http://TeamSTEPPS.ahrq.gov, accessed 19th February 2013.
Tucker AL and Edmondson AC (2003). Why hospitals don’t learn from failures: Organizational and psychological dynamics that inhibit system change. California Management Review, 45, p 55-72.
Tyler TR and Lind EA (1992). A relational model of authority in groups. In: Advances in
experimental social psychology, Vol. 25. Zanna, Mark P. (Ed.); San Diego, CA, US: Academic
Press, p 115-191 [Chapter].
Van den Bossche P, Gijselaers WH, Segers M and Kirschner, PA (2006). Social and cognitive factors driving teamwork in collaborative learning environments. Small Group Research, Vol. 37, Issue 5, p 490-521.
Vazirani S, Hays RD, Shapiro MF and Cowan M (2005). Effect of a multidisciplinary intervention on communication and collaboration among physicians and nurses. American Journal Of Critical
Care, Vol. 14, Issue 1, p 71-77.
Weaver R (2012). Power, professional identity and the doctor who can heal all ailments. Medical
Yin RK (2009). Case study research - design and methods, 4th edition, Thousand Oaks, CA,
US: Sage Publications Inc.
Yukl G (1994). Leadership in organizations. Englewood Cliffs, NJ: Prentice Hall.
Zand DE (1972). Trust and managerial problem solving. Administrative Science Quarterly, Vol. 17, p 229–239.
Zwaan M, Dijkstra JK and Veenstra R (2013). Status hierarchy, attractiveness hierarchy and sex ratio: Three contextual factors explaining the status–aggression link among adolescents.
Appendices
Appendix A - Teamwork according to TeamSTEPPS™
TeamSTEPPS™ logo:Teamwork according to TeamSTEPPS™ Source: http://TeamSTEPPS.ahrq.gov
About TeamSTEPPS
TeamSTEPPS is a teamwork system designed for health care professionals that is:
A powerful solution to improving patient safety within your organization.
An evidence-based teamwork system to improve communication and teamwork skills among health care professionals.
A source for ready-to-use materials and a training curriculum to successfully integrate teamwork principles into all areas of your health care system.
Scientifically rooted in more than 20 years of research and lessons from the application of teamwork principles.
Developed by Department of Defense's Patient Safety Program in collaboration with the Agency for Healthcare Research and Quality.
TeamSTEPPS provides higher quality, safer patient care by:
Producing highly effective medical teams that optimize the use of information, people, and resources to achieve the best clinical outcomes for patients.
Increasing team awareness and clarifying team roles and responsibilities.
Resolving conflicts and improving information sharing.
Eliminating barriers to quality and safety.
