Vertino, Kathleen A. DNP, RN, PMHNP-BC, CARN-AP
Communication failure is integrally linked to the incidence of preventable medical errors, the 6th leading cause of death according to the Institute of Medicine,1 and is a primary factor in almost 80% of adverse events and close calls in the Veterans Health Administration (VHA).2 Poor team communication directly affects nurses, contributing to decreased job satisfaction and low morale, and increased nurse turnover.3 Poor team communication can thus lead to conflict and hostility among team members and greater risk for bullying, harassment, and horizontal violence.3 The resulting negative nursing work culture places patients at greater risk for safety and quality errors.4 In contrast, nurse retention and satisfaction are increased when nurses feel empowered and autonomous, along with good working relationships with both management and peers.5
Highly functioning teams make fewer errors, indicating that teamwork is a critical component of a patient safety culture.6-8 Recognizing that patient safety and work environment are integrally related, the Agency for Healthcare Quality and Research and the Department of Defense conjointly developed a team-training initiative to promote and enhance improved patient safety, Team Strategies and Tools to Enhance Performance and Patient Safety (TeamSTEPPS©). TeamSTEPPS initiatives, and other team training programs, have been implemented in a variety of healthcare settings including emergency departments,9 intensive care units,10 operating rooms11 and labor and delivery rooms,7 and military hospital medical-surgical units.12 Emerging research supports the efficacy of team training initiatives in improving patient, staff, and organizational outcomes.6,9,11-16 However, despite well-documented success of TeamSTEPPS training, no studies utilizing this model in the VHA were found in this review.
This article reports on the implementation and evaluation of a TeamSTEPPS quality improvement (QI) initiative with nursing staff including RNs, licensed practical nurses (LPNs), and nursing assistants (NAs) on a single 25-bed acute inpatient unit within a 140-bed acute care VHA hospital in an urban setting. The primary purpose of this study was to determine if a structured team training initiative provided to inpatient nursing staff would improve staff attitudes toward teamwork. The study posed 2 questions:
1. Does TeamSTEPPS implementation with nursing staff on an inpatient unit result in improved attitudes toward teamwork?
2. Are there differences between occupational group (RNs, LPNs, NAs) and years of clinical experience on attitudes toward teamwork?
A pre-experimental pretest/posttest repeated-measures design was utilized.17 The sample was composed of full- and part-time staff (RNs, LPNs, and NAs) employed on a designated inpatient unit in a VHA hospital. All full- and part-time staff on the designated unit were asked to attend the TeamSTEPPS program. Intermittent or agency staff who were not regular full- or part-time staff on the designated unit were not eligible to participate. Twenty-six full- and part-time nursing staff were employed on the unit at the time of the project.
The project was deemed exempt by the State University of New York at Buffalo and Department of Veterans Affairs local institutional review boards. As a QI initiative, informed consent for participants was not required. To maintain the privacy and confidentiality of responses, few optional demographic questions were asked in the pretest (age, gender, occupational\category [RN, LPN, or NA], educational level, and years of nursing experience [NE]).
Following Kotter’s18 Change Model, researchers “established a sense of urgency” and “created a guiding coalition.”18(p21) The project director and executive-level nursing management identified nursing staff team communication as the most pressing quality and safety need. Specific concerns leading to this conclusion included problems with interpersonal conflicts among staff members, discontent on the unit as evidenced by a reported lack of communication and cooperation among coworkers, and a high staff turnover, including turnover in unit leadership.
The specific key actions that were used to apply Kotter’s model are outlined in Table, Supplemental Digital Content 1, http://links.lww.com/JONA/A288. The sense of urgency (step 1) was present and became clear during the needs assessment, providing a springboard for the project. Once the need was identified, the guiding coalition (step 2) was formed. As the vision for the project became clear, action steps were taken to begin a strategy for implementation (steps 3 and 4). Following the training, the project director worked with all staff to gather staff input via site visits, mentoring, coaching, and periodic e-mail communications (steps 4, 5, and 6), all elements of empowerment techniques supported in the literature as effective in initiating change.5,18
Stakeholder meetings were scheduled with the nurse managers (NMs) and the associate chief nurse to explain the project’s aim, purpose, and proposed time frame. Staffing constraints made scheduling the training a formidable task requiring considerable effort and coordination on the part of nursing leadership. The project director remained available for questions throughout the study. Executive nursing leadership and the unit NMs were kept apprised of all developments, progress, and barriers as the initiative was implemented and were provided with individualized education to ensure their knowledge of the TeamSTEPPS fundamentals.
An informational e-mail was sent to all unit staff approximately 4 weeks in advance of training. The e-mail introduced TeamSTEPPS and provided a link to a Web site, which identified dates for the training. Staff members were informed that the training would be 4 hours in length. The training would be provided during a regularly scheduled shift, however; when not possible, staff were informed that they would be paid overtime to attend. A follow-up e-mail asked staff to register for the training with the unit NM. The NM also introduced the training to the staff during staff meetings and requested they sign up for the training.
The project director, a board-certified psychiatric nurse practitioner, a VHA fellow mentor, and a TeamSTEPPS MasterTrainer, provided 5 four-hour training sessions to staff in July 2012. Training consisted of a customized version of the TeamSTEPPS involving formal presentation, discussion, and role-play exercises embodying clinical scenarios relevant and applicable to this specialty unit and nursing staff, a strategy successfully employed by other trainers.19 Stories of real-life patient situations including videos were interspersed in the sessions.
Measures and Data Collection
The 30-item TeamSTEPPS-Teamwork Attitude Questionnaire (T-TAQ) (available at http://teamstepps.ahrq.gov/taq_index.htm) was utilized for the pre/posttest assessment. T-TAQ assesses individual attitudes toward the 5 targeted components of teamwork: team structure, leadership, mutual support, situation monitoring, and communication. Cronbach’s α coefficient of internal consistency reliability for the teamwork constructs are reported as .70, .81, .83, .70, and .74, respectively (http://teamstepps.ahrq.gov/taq_index.htm). Responses were scored on a Likert scale (1 = strongly disagree, 5 = strongly agree). The T-TAQ and a demographic survey were administered to 26 participants (100%) at the beginning of training.
Kotter’s Change Model (steps 5 and 6) emphasizes the importance of staff empowerment to promote change and that transformational leaders are found in high reliability organizations. For 6 weeks following the training, the investigator made periodic site visits to collect field notes, in addition to providing intermittent mentoring, coaching, and support to staff via follow-up e-mails and phone calls in support of this concept. This hands-on approach models transformational leadership characteristics and supports the goal of empowering staff to advocate for their own development.20
At the end of the study period, participating staff were requested to confidentially retake the T-TAQ plus 2 additional yes/no exit questions: (1) Do you feel teamwork has improved on your unit as a result of TeamSTEPPS training? (2) Do you feel there has been a change in the culture on your unit as a result of TeamSTEPPS training? The project director visited the unit at various days and times in order to collect as many posttests as possible.
Data Management and Analysis
Data were analyzed using Systat Software, version 11 (San Jose, California). Analysis was completed for staff participants (78.3%; n = 18) who completed both the pretest and posttest T-TAQ. Data from 8 participants were excluded from the analysis as follows: 5 staff chose not to complete the posttest, and 3 staff were deemed ineligible due to not meeting inclusion criteria (full- or part-time assigned to the designated unit throughout the study period); therefore, 18 (78.3%) of the original 26 participants completed the posttest. The raw data were right-skewed, indicating that respondents expressed more than normally expected agreement with survey items. Consequently, the reflect-and-log procedure was used to decrease the impact of the nonnormal distribution.21 In addition to descriptive statistics, a repeated-measures analysis of covariance (ANCOVA) with 2 within (pre/post) and 1 between (occupational group) factor with 3 levels (RN, LPN, NA) was used. The covariates were years of experience and baseline (pretest T-TAQ) scores. A χ2 test was utilized to examine results of the 2 exit questions.
In general, this sample exhibited normal distributions across age, education, and years of experience (Table 1). The mean age for the sample was 44.7 years. Mean years of experience in nursing was 12.0 with a very similar distribution across groups. The pretest sample was composed of 17 female (65.4%) and 9 male (34.6%) participants. More RNs participated in both pretest (time 0) and posttest (time 1) at 46.2% and 50%, respectively, as compared with LPN (26.9% at time 0, 28% at time 1) and NA participants (26.9% at time 0 and 22% at time 1). Thus, it is not surprising that 38.4% of the total sample of 26 had a BSN as compared with 30.8% with an ADN and 30.8% with a high school education.
Results of mean scores for each occupational group (RNs, LPNs, NAs) including unadjusted mean and SD scores at time O and time 1 for all groups are displayed in the Figure 1. Total unadjusted mean T-TAQ scores at time 0 and time 1 were 4.20 and 4.64, respectively, with a total change score of 0.44. Unadjusted mean scores at time 0 were 4.29, 3.77, and 4.38 for RNs, LPNs, and NAs, respectively. Mean change scores for RNs, LPNs, and NAs were 0.40, 0.95, and 0.12, respectively, indicating the LPNs had the largest change scores followed by RN and lastly NA change scores. Total unadjusted mean T-TAQ scores across all TeamSTEPPS teamwork constructs are displayed in Table 2.
Effect of Occupational Group and Years of Experience on T-TAQ Scores
Results for these analyses are displayed in Table 3, which describes values for both untransformed and transformed data. As noted, we transformed the data utilizing reflect and log to account for skew to the right and address the statistical assumption of normality for linear models. ANCOVA revealed significant increases in total T-TAQ scores (F1,13 =106, P ≤ .001) for untransformed data as well as transformed data T-TAQ scores (F1,13 = 74.6, P ≤ .001), indicating significant increases from pretest to posttest T-TAQ scores. Results for the 5 team constructs (time variable) with transformed data were as follows: team structure (F1,13 = 90.3, P ≤ .001), leadership (F1,13 = 79.0, P ≤ .001), situation monitoring (F1,13 = 36.7, P ≤ .001), mutual support (F1,13 = 54.2, P ≤ .001), and communication (F1,13 = 35.2, P ≤ .001), demonstrating statistical significance at P ≤ .001 across all 5 teamwork constructs. Neither occupational group nor years of experience moderated any pretest to posttest changes in total T-TAQ or in any of the subscales (all nonsignificant), indicating that there was no effect on the scores.
Staff Members’ Feelings Regarding Change
Of the 18 staff who completed the posttest, 17 (94.4%) answered the 2 exit questions regarding their feelings with respect to change in teamwork and culture on their unit. Eighty-two percent felt that teamwork had improved on their unit as a result of the TeamSTEPPS training (χ21 = 7.118, P = .0.008). However, a culture change on the unit was not endorsed by as many staff; only 61.1% responded yes to that question (χ21 = 0.529, P = .467, not statistically significant).
The results indicate that TeamSTEPPS training did result in improved attitudes toward teamwork. The ANCOVA (Table 3), controlling for years of experience (NE) and occupational group (Ocg), revealed that the posttest T-TAQ scores were significantly improved. Furthermore, the training resulted in improved attitudes on the 5 teamwork constructs measured by the T-TAQ regardless of an individual’s Ocg or years of experience in nursing. It should be noted that the 5 teamwork construct measures were significant regardless of the reflect-and-log transformations, suggesting that conservative analysis continued to produce statistically significant outcomes.
Posttest exit questions demonstrated that participants perceived a statistically significant improvement in the teamwork on their unit after the training. However, the notion of improved unit culture was not supported. It is possible that the question about “unit culture” did not translate as easily to perceptions of changes in the unit climate as did the notion of improved teamwork. However, the consistent improvement in all of the dimensions of the T-TAQ instrument does reflect positive attitudinal improvements in team structure, leadership, situation monitoring, mutual support, and communication. Most important to concerns about quality of care and patient safety is the notion of improved staff unit communication, which is linked to improved patient outcomes and staff satisfaction.
Although this study focused on only 1 unit in a large VHA facility, the positive effects of the TeamSTEPPS program on this small sample of staff suggest that this initiative could be evaluated for broader application within other units and settings. Additional posttraining mentoring and coaching were implemented at the conclusion of this project to reinforce initial attempts to strengthen communication and teamwork. There is no way to filter out the positive effects of this additional intervention or to determine possible interaction effects between the training and mentoring activities, but these results do suggest the need for future studies that could compare outcomes related to TeamSTEPPS training with and without posttraining coaching and mentoring activities.
Field notes collected during site visits provide anecdotal evidence of improved unit morale. For example, 2 staff members shared success stories with respect to improved coworker relationships that they felt were a result of improved communication. One staff member, who had previously decided to leave the unit reconsidered that decision after the study training. Another individual shared ideas regarding the development of a small-scale improvement project that would measure a patient safety outcome on 1 of the unit shifts. Finally, once all data were collected and analyzed, another staff member sent an e-mail to apologize for not participating in the posttest because of workload responsibilities. This same person also reported the implementation and success of a new shift report procedure, based on modification of a TeamSTEPPS tool.
The study used a small convenience sample and required attendance at the training sessions, thus overt and public administrative support for the project may have created bias. Furthermore, the training sessions themselves, initially planned to be provided all the same week, were spread out over a 2-week period in order to provide training to the maximum number of staff. Since the training was provided in mid to late July and many staff attended the 4-hour training while being paid overtime to attend, either prior to or following their regular shift, fatigue could have played a role in their ability to fully absorb the material. Another possible limitation was staff turnover that occurred during the study period; specifically, new nurses were hired, and it is impossible to assess whether or how their addition affected teamwork or the unit culture. Finally, we did not identify whether any of the staff participants had previously or currently served in the military and as such may have received team training in the past.
Results of this project suggest that TeamSTEPPS, when tailored to individual unit needs and combined with mentoring and coaching, can promote a positive change in staff attitudes toward team structure, leadership, situation monitoring, mutual support, and communication. In addition, this project demonstrated the feasibility and adaptability of implementing a 6-week unit-based project, with the goal of improving staff attitudes toward teamwork, with staff on a specialty unit caring for patients with complex care needs. Feedback from course evaluations was positive, indicating that staff enjoyed the training and found it useful, practical, and applicable to their particular unit.
The project offers a number of possible implications for nursing practice and for future directions. The project supports previous studies that document the effectiveness of TeamSTEPPS training with frontline staff in improving attitudes toward teamwork, improving patient safety, reducing turnover, enhancing job satisfaction, promoting higher levels of patient and nurse satisfaction, and achieving better clinical outcomes.3-8
TeamSTEPPS can be customized to meet the needs of any small- or large-scale study and is readily accessible from the Agency for Healthcare Research and Quality (AHRQ) Web site (http://teamstepps.ahrq.gov/). The 5 teamwork constructs contained in the T-TAQ instrument can be utilized separately to assess specific aspects of teamwork, on a small or large scale, and then a customized educational program can be developed to target a particular area of teamwork. Many hospitals have the resources in existing educational departments to pursue this type of research and educational program.
Although it was beyond the scope of this project to evaluate the TeamSTEPPS initiative in relationship to organizational, patient safety and quality outcomes, this project provides a foundation for future projects exploring these important and possibly related outcomes.
The author thanks Dr Nancy Campbell-Heider, University at Buffalo, School of Nursing, chair and advisor, for her tutelage, mentorship, and editorial assistance. She has inspired the author’s development as a researcher and nurse scholar, by being a role model. Rocco Paluch, statistician and data analyst, committed time and attention to ensure the methodology was sound. Lizabeth Weiss, MS, RN, nurse executive, served as the author’s clinical preceptor throughout the project and assisted in navigating the VA system. Dr Marthe Moseley, VHA Office of Nursing Service, willingly shared ideas, reviewed drafts, and provided valuable feedback.
1. Institute of Medicine. To Err Is Human: Building a Safer Healthcare System. Washington, DC: National Academy Press; 1999.
2. Veterans Health Administration. National Patient Safety Improvement Handbook (1050.01). March 4, 2011. www.patientsafety.gov
. Accessed November 10, 2013.
3. Vessey J, DeMarco R, DiFazio R. Bullying, harassment and horizontal violence in the nursing workforce. Annu Rev Nurs Res. 2010; 28 (1): 133–157.
4. Hayes L, O’Brien-Pallas L, Duffield C, et al. Nurse turnover: a literature review [review]. Int J Nurs Stud. 2006; 43 (2): 237–263.
5. Rathert C, May DR. Health care work environments, employee satisfaction, and patient safety: care provider perspectives [research support, non-US Gov’t]. Health Care Manage Rev. 2007; 32 (1): 2–11.
6. Baker DP, Day R, Salas E. Teamwork as an essential component of high-reliability organizations. Health Serv Res. 2006; (4 pt 2): 1576–1598.
7. Mann S, Marcus R, Sachs B. Lessons from the cockpit: how team training can reduce errors on L & D (grand rounds). Contemp Ob/Gyn. 2006; 51 (1): 34–38.
8. Salas E, DiazGranados D, Weaver SJ, et al. Does team training work? Principles for health care. Acad Emerg Med. 2008; 15 (11): 1002–1009.
9. Morey JC, Simon R, Jay GD, et al. Error reduction and performance improvement in the emergency department through formal teamwork training: evaluation results of the MedTeams project [multicenter study research support, US Gov’t, non-PHS]. Health Serv Res. 2002; 37 (6): 1553–1581.
10. Pronovost P, Berenholtz S, Dorman T, et al. Improving communication in the ICU using daily goals [research support, US Gov’t, PHS]. J Crit Care. 2003; 18 (2): 71–75.
11. Awad SS, Fagan SP, Bellows C, et al. Bridging the communication gap in the operating room with medical team training [comparative study]. Am J Surg. 2005; 190 (5): 770–774.
12. Ferguson SL. TeamSTEPPS: integrating teamwork principles into adult health/medical-surgical practice. Medsurg Nurs. 2008; 17 (2): 122–125.
13. Haig KM, Sutton S, Whittington J. SBAR: a shared mental model for improving communication between clinicians. Jt Comm J Qual Pt Saf. 2006; 32 (3): 167–175.
14. Leonard M, Graham S, Bonacum D. The human factor: the critical importance of effective teamwork and communication in providing safe care. Qual Saf Health Care. 2004; 13 (Suppl 1): i85–i90.
15. Capella J, Smith S, Philp A, et al. Teamwork training improves the clinical care of trauma patients [comparative study]. J Surg Ed. 2010; 67 (6): 439–443.
16. Clark PR. Teamwork: building healthier workplaces and providing safer patient care. Crit Care Nurs Q. 2009; 32 (3): 221–231.
17. Campbell D, Stanley J. Experimental and Quasi-experimental Designs for Research. Chicago, IL: R. McNally; 1966.
18. Kotter JP. Leading Change. Boston, MA: Harvard Business School Press; 1999: 1–186.
19. Rabinowitz M, Johnson LE, Mazzapica D, et al. Storytelling effectively translates TeamSTEPPS skills into practice. J Contin Educ Nurs. 2010; 41 (11): 486–487.
20. Curtis EA, Sheerin FK, de Vries J. Developing Leadership in nursing: the impact of education and training. Br J Nurs. 2011; 20 (6): 344, 352.
21. Tabachnick B, Fidell L. Using Multivariate Statistics. 3rd ed. New York: Harper Collins; 1996.