Article Text

Building collaborative teams in neonatal intensive care
  1. Dara Brodsky,
  2. Munish Gupta,
  3. Mary Quinn,
  4. Jane Smallcomb,
  5. Wenyang Mao,
  6. Nina Koyama,
  7. Virginia May,
  8. Karen Waldo,
  9. Susan Young,
  10. DeWayne M Pursley
  1. Department of Neonatology, Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA
  1. Correspondence to Dr Dara Brodsky, Department of Neonatology, Beth Israel Deaconess Medical Center, Rose 3, 330 Brookline Avenue, Boston, MA 02215, USA; dbrodsky{at}


The complex multidisciplinary nature of neonatal intensive care combined with the numerous hand-offs occurring in this shift-based environment, requires efficient and clear communication and collaboration among staff to provide optimal care. However, the skills required to function as a team are not typically assessed, discussed, or even taught on a regular basis among neonatal personnel. We developed a multidisciplinary, small group, interactive workshop based on Team STEPPS to provide staff with formal teamwork skills, and to introduce new team-based practices; 129 (95%) of the eligible 136 staff were trained. We then compared the results of the pretraining survey (completed by 114 (84%) of staff) with the post-training survey (completed by 104 (81%) of participants) 2 years later. We found an improvement in the overall teamwork score from 7.37 to 8.08 (p=<0.0001) based on a range of poor (1) to excellent (9). Respondents provided higher ratings in 9 out of 15 team-based categories after the training. Specifically, staff found improvements in communication (p=0.037), placed greater importance on situation awareness (p=<0.00010), and reported that they supported each other more (p=<0.0001). Staff satisfaction was rated higher post-training, with responses showing that staff had greater job fulfilment (p=<0.0001), believed that their abilities were being utilised properly (p=0.003), and felt more respected (p=0.0037). 90% of staff found the new practice of team meetings to help increase awareness of unit acuity, and 77% of staff noted that they had asked for help or offered assistance because of information shared during these meetings. In addition to summarising the results of our training programme, this paper also provides practical tools that may be of use in developing team training programmes in other neonatal units.

  • Team training
  • Attitudes
  • Paediatrics

Statistics from


There is a strong emphasis on team training in many industries where human safety is at risk, such as aviation, nuclear power, military and chemical manufacturing.1 These fields recognise that optimal management of dynamic, high-risk situations requires both expertise in technical skills and mastery of team management, leadership, communication and resource allocation.2 Similarly, a team-based approach is central to clinical practice in the neonatal intensive care unit (NICU). The complex multidisciplinary nature of neonatal intensive care combined with the numerous hand-offs occurring in this shift-based clinical work environment, requires efficient and clear communication and collaboration among personnel to provide optimal care.

Despite a team-based focus in neonatal care, the skills required to function as a team are not typically assessed, discussed or even taught on a regular basis among NICU personnel. As in other medical fields, there is an assumption that neonatal clinicians will acquire teamwork skills through observation and experience. Formal teamwork training of clinicians has been shown to lead to fewer errors and better clinical outcomes in several clinical areas, including the emergency room department,3–5 operating room,4 ,6–11 antepartum obstetrics unit,12 ,13 delivery room14–19 and intensive care units.20 These findings have prompted numerous organisations including The Joint Commission and the Institute of Medicine21–23 to recommend team training as an effective strategy for curtailing medical errors. A potential added benefit of team training is improved staff satisfaction, which has been associated with increased patient satisfaction,24 enhanced patient loyalty,25 and higher perception of quality of care by patients26 and staff.27

In light of this, we sought to explore staff attitudes about teamwork in our level IIIB NICU at Beth Israel Deaconess Medical Center (BIDMC), and to identify teamwork components that worked well and areas that required improvement. We then developed a multidisciplinary, small group, interactive workshop to provide staff with formal opportunities to develop and practice teamwork skills, and to introduce new team-based practices. Following these, we determined the long-term effectiveness of this training and the impact of these new practices by comparing staff perceptions of teamwork pretraining and 2-years post-training.

In this paper, we seek to describe the impact of team training in our NICU, but also to share the tools developed for our programme. It is our hope that these resources will be of use to other NICUs interested in adopting similar programmes for their staff.


We established a nine-member multidisciplinary committee of NICU staff consisting of the Chief of the Department of Neonatology, two neonatologists, a neonatal nurse practitioner, a clinical nurse specialist, a nurse manager, a respiratory therapist and two clinical nurses. Committee members attended a 2-day ‘Train the Trainer Program’ developed by the BIDMC Department of Obstetrics and Gynecology's Patient Safety Team Training Group. This programme was based on modifications of Team STEPPS, which was co-developed by the American Institutes for Research, the Department of Defense, and the Agency for Healthcare Research and Quality (AHRQ),28 and was shown to improve obstetrical care.29 ,30

To assess staff opinions about teamwork components that worked well within the NICU and areas that required improvement, the NICU committee then created a survey (table 1) modified from teamwork assessment surveys developed by AHRQ.31 Because a validated team training assessment instrument did not exist at the time of our initial survey, we created our abridged survey by establishing face validity. To develop our survey, committee members used a consensus-building approach to identify specific items from the AHRQ survey that were relevant to a neonatal intensive care environment.

Table 1

Content of pretraining and post-training survey*

The survey was given to all staff providing direct clinical care in the NICU, including nurses, respiratory therapists, nurse practitioners and physicians. The survey was anonymous except for identification of the evaluator's clinical position. We asked staff to rate each statement based on a 5-point Likert scale with options of ‘strongly disagree’, ‘disagree’, ‘neither agree nor disagree’, ‘agree’, or ‘strongly agree’. Neonatal staff also responded to open-ended questions about how the unit could work better as a team. We also asked staff to rate how well the NICU functions as a team using a 9-point Likert scale from 1 to 9, corresponding to ‘extremely poor’ and ‘excellent’, respectively.

Based on the results of this survey, the committee created a team training curriculum for NICU staff. While we applied components from the BIDMC Obstetrics Department's training programme, we inserted realistic neonatal scenarios to tailor the programme to neonatal clinicians. The workshop content focused on providing clinical staff with formal tools to support the four primary teamwork skills: communication, leadership, peer support and situation monitoring (table 2).

Table 2

Content of training sessions

We created two realistic neonatology-specific videotapes to encourage discussion about ineffective and effective teamwork in practice (see online supplementary appendix 1a: Ineffective Teamwork; online supplementary appendix 1b: Effective Teamwork). To address deficient areas identified in the initial survey, we also created three new team-based applications (table 3): (1) team meeting (brief, scheduled staff gathering once per shift); (2) team update (spontaneous meeting in anticipation of future events); (3) team debriefing (small group spontaneous meeting in response to a specific event).

Table 3

Teamwork applications

Staff training was performed during 3-h small group workshops (see online supplementary appendix 2: Workshop PowerPoint Presentation). All clinical staff in the NICU were required to attend a workshop. Eighteen of the 20 workshops were facilitated by two committee members of different disciplines; the remaining two workshops were facilitated by two nursing staff. At the beginning of each workshop, participants viewed the ‘Ineffective Teamwork’ videotape, and facilitators encouraged group discussion about clinicians’ actions that could have been improved. Following these, facilitators presented a 5 min didactic discussing the impact of medical errors and the benefit of teamwork to minimise errors. Then, the four teamwork skills were introduced with an emphasis on encouraging participation and group interaction. Audience members were asked to work through specific realistic examples to apply tools that they learned. After watching the ‘Effective Teamwork’ videotape, participants reiterated the effective teamwork skills that they had observed. The three new team-based skills were then discussed by the facilitators, and examples of these activities were provided. Attendees were asked to complete an evaluation form immediately following the workshop.

During the first 2 weeks after implementing the team meetings, at least one of the committee members joined the team meetings to ensure that the format was being followed, and to offer improvements to the structure of the meeting. The most common recommendation offered by observers was to decrease the amount of information provided during the nurse review of patients. Committee members have continued to randomly participate in team meetings, updates and debriefings (ie, during their clinical shifts), offering suggestions to the format, as needed. All new clinicians were required to participate in a modified team training workshop during their orientation period.

To determine whether this training led to any long-term positive changes, the teamwork survey was distributed 2 years later. This second survey also assessed staff opinions about the new practices (ie, team meetings, updates, debriefings) that were instituted post-training, and asked if communication among NICU team members had improved since the training. To assess the frequency of team meetings, the survey also contained questions about the incidence of team meetings, and a new paging system enabled us to calculate the number of team meeting pages during a 3-month period more than 5 years after the team training.

Differences between the pretraining and post-training surveys were assessed using χ2 statistics for categorical data and the Kruskal–Wallis test for the Likert scale measures. All analyses were performed using SAS V.9.2 and a p value of less than 0.05 was considered significant. When corrected for multiple comparisons using the Bonferonni correction, the p value of significance was 0.0033.

The BIDMC institutional review board examined our proposed activities and provided verification that analysis of the effectiveness of our training does not constitute human subject research and did not require further institutional review.


At the time of this project, 136 staff members were providing direct clinical care in our NICU. Of these, 114 (84%) completed the pretraining survey. The distribution of disciplines of the survey responders reflected the overall staff: nurses 74%, physicians/nurse practitioners 8% and respiratory therapists 8%. Table 4 summarises the pretraining scores of each of the 15 team-based categories with responses based on a 5-point Likert scale with 1=strongly disagree, 2=disagree, 3=neither disagree or agree, 4=agree, and 5=strongly agree. Overall, results of the pretraining showed fairly high staff satisfaction with team work, with average ratings above 4 in 11 of the 15 categories. The lowest scores were observed in those questions related to situational awareness.

Table 4

Comparison of pretraining and post-training survey results†

Of the 136 eligible staff, 129 (95%) attended a workshop and 118 of the participants (91%) completed the postworkshop evaluation. On these evaluations, all participants rated the workshop as valuable to their professional development, with 98 (83%), 17 (14%), and 3 (2.5%) providing a rating of ‘strongly agree’, ‘really agree’, or ‘agree’, respectively. Many attendees valued the opportunity to hold multidisciplinary discussions resulting in increased awareness of the perspective of those in other disciplines. Staff were encouraged to provide suggestions to improve teamwork in the NICU. The committee reviewed this feedback and incorporated into practice those suggestions that were prevalent, feasible, practical and of high potential impact. Examples of these practice changes are provided in table 5.

Table 5

Examples of practice-based changes resulting from workshop discussions

Two years later, 81% (n=104) of the 129 workshop attendees completed the post-training survey. The professional breakdown of those who took the post-training survey (nurses 75%, physicians/nurse practitioners 12%, and respiratory therapists 10%) mirrored that of the pretraining survey. Table 4 compares the results from the post-training survey with those from the pretraining survey. One of the most significant differences was in the overall teamwork score, which improved from 7.37 to 8.08 (p=<0.0001), based on a range of poor (1) to excellent (9). In the teamwork-related categories, 9 out of 15 responses received higher ratings in the post-training survey.

Specifically, staff noted in the post-training survey that there was greater encouragement of open and honest communication (#3, p=0.037), and believed that they were better informed (#4, p=0.0002). After the training, staff placed greater importance on being aware of the status of other staff member's patients (#9, p=0.0032), and of events occurring in the NICU (#10, p=<0.0001). They were also more willing to hear constructive feedback (#5, p=0.022). Furthermore, a greater number of clinicians knew whom they should approach when they needed help (#11, p=<0.0001). One of the most striking findings was that staff satisfaction was rated higher post-training, with responses showing that staff had greater job fulfilment (#13, p=<0.0001), believed that their talents and abilities were being utilised properly (#14=0.003), and felt more respected (#15, p=0.0037).

By contrast, comparison data showed that our interventions had no impact on the staff's perception of whether staff as a whole shared the same purpose and vision (#1). Furthermore, the staff's assessment of their ability to acknowledge and respond to conflicts did not change (#2, #6 and #7), and staff did not feel any more comfortable about approaching leaders if they had a concern (#8).

Finally, we reviewed the post-training survey results to determine the impact of team meetings, debriefings and updates that had been instituted after training. We found that staff held team meetings during 98% of the daytime shifts and 73% of the night-time shifts 2 years after these applications were introduced. This high frequency of team meetings persisted more than 5 years after the training; review of paging records during a 3-month period revealed that staff held team meetings during 94% of the daytime shifts and 76% of the night-time shifts; most of the missed opportunities were on a holiday or weekend shift (77%) or during a night shift (81%). In the post-training survey, staff noted that debriefings and updates occurred less frequently than team meetings (65% and 51%, respectively). Ninety percent of staff believed that team meetings were helpful at increasing awareness of the current and potential acuity in the unit. Indeed, 77% of staff noted that they had asked for help or offered assistance to others because of information shared during the team meetings. Similarly, improvement in communication was observed with 77 (84%) of the 91 responders believing that communication between NICU team members improved after the team training programme.


Studies have shown that medical errors typically occur because of system failures, rather than individual mistakes, and that improved teamwork is an essential factor to reduce the number of hospital errors.3 ,5 ,28 ,32–36 In the NICU, the potential for medical errors is magnified because there are team members of different disciplines, ranges of personnel experience, numerous hand-offs, high patient acuity, complicated clinical scenarios and numerous disruptions and distractions. Furthermore, even minor mishaps can have potentially serious consequences in the NICU. In this study, we outline an approach to formal team training of multidisciplinary clinicians in a NICU, report the staff's perceptions about the workshop training, and compare the staff's attitudes about teamwork, pretraining and 2 years post-training. An extensive review of healthcare training studies identified a lack of programme design included in the literature and imparted a critical need for future studies to include team training programme content.37 Thus, we also provide practical tools that may be of use in developing team training programmes in other neonatal units.

We found that staff rated the training extremely highly. We hypothesise that one of the reasons for the success of the workshop was our use of teaching principles specifically targeted to adult learners.38 By allowing participants to contribute their own ideas in a small group setting, and having learner-centred discussions that were facilitated rather than teacher-dominated, staff were actively engaged in their learning. In addition, by basing the content on relevant clinical experiences and linking new concepts to these existing frameworks, participants were more invested in learning new team-based approaches.

Our study found that even among staff who rated the BIDMC NICU highly for teamwork in the pretraining survey, there was still a substantial increase in the overall teamwork score as well as in the ratings for specific teamwork categories 2 years later. Specifically, staff found improvements in communication, placed greater importance on situation awareness and reported that they supported each other more. While difficult to measure, we believe that there were intangible factors that also led to these improved perceptions, including having overt ‘buy-in’ from recognised NICU leaders (eg, Chief of Neonatology, nurse manager and nurse educator); empowering team members by asking for their opinion in surveys and during workshops; having workshop facilitators and learners from distinct disciplines; and raising general awareness of the importance of teamwork.

An unexpected benefit of implementation of the team training workshops was the creation of venues for open discussion of current problems with NICU workflow and teamwork. Although not the primary intent of the workshops, numerous suggestions for improvement were provided by staff members after the workshops, and many of these suggestions were incorporated directly into practice.

We were pleased to see that our training, and the new applications, seemed to have also improved staff satisfaction. Because we had not focused our training on staff satisfaction, we had not anticipated this finding. It is difficult for us to assign the contributions of the training and the practice changes to these improvements. We hypothesise that higher staff satisfaction was, at least in part, related to the unique opportunity for staff to express their opinions to multidisciplinary colleagues during the workshops. Staff were also able to provide additional thoughts about NICU teamwork during their written evaluations. In addition, the committee made a constructive effort to promptly respond to staff suggestions for NICU improvements.

Interestingly, the survey content that did not rate higher after the study interventions had some similarities (#2, #6, #7 and #8). Each of these statements focused on either resolving conflicts or challenging senior clinicians. While it is difficult to guide adults about how to improve these actions, perhaps the introduction of role play into the training scenarios, might have helped staff develop these challenging skills.39

As for the new applications, there are two possible explanations for the greater prevalence of team meetings during non-holiday weekday shifts. During the daytime weekday shifts, there are more deliveries and admissions and a greater number of procedures and consultations. Weekday staff may thus feel a greater need for team meetings. Alternatively, the presence of clinical managers during non-holiday weekdays, may encourage staff to convene these meetings. The lower incidence of debriefings and updates was as expected, given that they were intended to occur only as needed.

By comparison with other team training studies, we found that our teamwork programme had unique characteristics. In previous studies, team training of neonatal staff targeted team-related issues in the delivery room, often with use of simulation training.14–16 ,18 ,19 In two studies that trained NICU clinicians, physicians did not participate.40 ,41 In our study, we intentionally focused on all disciplines and provided training about teamwork in the intensive care setting as well as the delivery room.

Most studies that examined the effectiveness of team training in other fields focused on shorter time periods with evaluation either immediately after the intervention4 ,14 ,42 or up to 8 months later.5 ,7 ,9 ,16 ,19 ,40 ,43 We identified very few studies that evaluated the attitudinal impact of staff as long as 1 year after training,6 ,10 ,20 ,44–46 with one study investigating outcomes 18 months later.17 In our study, we found that improved teamwork attitudes persisted 2 years after our interventions.

There are several limitations to this study. Because staff surveys were anonymous, and we tallied scores for staff as a whole, we did not measure changes in an individual staff member's attitudes. Neither of our surveys captured all the staff and, thus, there may have been some responders who completed either the pretaining or post-training survey, but not both. Lack of a validated survey design may have led to unintentional subjective survey statements and biased ratings. While there are several confounding variables in the NICU environment that might weaken our findings, we found similarities in patient acuity, daily census, leadership and staff turnover during the two survey periods. However, our study did not account for potential differences in staff morale at the time the surveys were distributed. Finally, we did not measure changes in teamwork knowledge or behaviours, nor impact on patient outcomes but rather, our survey focused on staff attitudes about these areas.

In the future, we are interested in determining if staff's optimism wanes over time, and if so, defining when this happens, so that additional training can be provided. The committee also hopes to determine if the staff's perceived progress in functioning as a team and greater satisfaction correlates with changes in staff behaviours, family satisfaction and error reduction.

To our knowledge, this is the first report of a comprehensive approach to team training in the NICU. The unique features of our programme that we believe were critical to its success are: (1) incorporating multidisciplinary involvement in all aspects of the programme, including the planning committee, workshop facilitators and the learners attending each workshop; (2) removing traditional medical hierarchies by empowering staff of all disciplines to serve as leaders, trainers and content experts; (3) extending the concept of team training to include process improvements, by responding to specific concerns expressed by staff during training sessions. We believe that the tools provided in this report can be readily adapted by other NICU providers to establish a successful team training programme, and that the lessons learned by our experience can contribute to team training efforts in other clinical arenas.


We would like to acknowledge Bruce Wahl and Tom Laws in the Department of Information Systems at Beth Israel Deaconess Medical Center (BIDMC), Boston, MA who produced the team training videotapes and Rick Gore in the BIDMC Clinical Engineering Department who obtained the team meeting pages. We would also like to thank Marie C McCormick, MD, ScD for her editorial insights.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:


  • Contributors All authors listed above meet the authorship requirements as defined by the ICJME. Specifically, all authors (DB, MG, MQ, JS, WM, NK, VM, KW, SY and DP) provided substantial contribution to conception and design, acquisition of data or analysis and interpretation of data. All authors contributed to the draft of the article or revising it critically, and have provided final approval of the submitted version of this manuscript.

  • Competing interests None.

  • Ethics approval Institutional Review Board.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement All additional unpublished data is available for sharing for anyone who is interested.  Please contact first author Dara Brodsky at

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.