Background
Children’s Advocacy Centers (CACs) use multidisciplinary teams to provide coordinated interagency responses to allegations of sexual abuse and other severe maltreatment [1–Reference Herbert and Bromfield3]. There are >950 CACs in the United States, half in rural areas, and they serve hundreds of thousands of children annually [1]. Teams include CAC staff and professionals employed by independent organizations (e.g., child welfare, law enforcement, victim advocacy, mental health) [2–Reference Ballard, McGlone and Ramgolam4]. Teams are fluid, with members participating to varying degrees at varying times depending on their role and involvement in specific cases [Reference Driskell, Funke and Tolston5,Reference van Heteren, Raaphorst and Groeneveld6]. Effective teamwork is required for these fluid, cross-boundary teams to investigate allegations and connect families with needed services [2].
Teamwork is a broad term referring to both team functioning (i.e., how teams think, feel, and act) and performance. Team functioning includes affective, behavioral, and cognitive processes and states (e.g., trust, communication, shared goals) [Reference Kozlowski, Bell, Schmitt and Highhouse7–Reference McGuier, Kolko and Aarons9], while performance reflects team output, including work quality (e.g., task completion, efficiency, productivity) [Reference Kozlowski, Bell, Schmitt and Highhouse7,Reference Mathieu, Hollenbeck and van Knippenberg8]. More than a century of research across diverse work settings has established that team functioning influences performance [Reference Mathieu, Hollenbeck and van Knippenberg8]. In healthcare and child welfare settings, better team functioning and performance are associated with better outcomes for those served by the team [Reference Dinh, Traylor and Kilcullen10,Reference Kim, Trahan and Bellamy11]. In CAC teams specifically, team functioning is positively associated with team member-rated performance [Reference McGuier, Rothenberger and Campbell12] and case outcomes (e.g., service receipt) [Reference Ballard, McGlone and Ramgolam4]. Better team functioning is also associated with lower burnout and turnover intentions [Reference Ballard, McGlone and Ramgolam4,Reference Galletta, Portoghese and Carta13,Reference Martinussen, Adolfsen and Lauritzen14], making it important for the well-being of the child abuse workforce.
Team development interventions can improve team functioning, team performance, and workforce outcomes [Reference McEwan, Ruissen and Eys15–Reference Shuffler, Diazgranados and Maynard18]. Team training is a specific type of intervention that targets team functioning by teaching teamwork knowledge, skills, and abilities [Reference Hughes, Gregory and Joseph16–Reference Shuffler, Diazgranados and Maynard18]. On average, it has large positive effects on team functioning and performance [Reference McEwan, Ruissen and Eys15,Reference Salas, DiazGranados and Klein17]. In healthcare settings, a meta-analysis found that team training results in increased learning, increased skill use, and improved organizational and patient outcomes [Reference Hughes, Gregory and Joseph16]. However, most team training studies have focused on traditional “fixed” teams with clearly bounded stable membership, and less is known about team training in the context of fluid cross-boundary teams like CAC teams [Reference Driskell, Funke and Tolston5,Reference van Heteren, Raaphorst and Groeneveld6]. Fluid teams rely more on informal communication, require more intentional coordination, and face greater challenges in negotiating overlapping roles and responsibilities [Reference van Heteren, Raaphorst and Groeneveld6]. Team interventions designed for fluid cross-boundary teams are a promising approach to enhancing CAC team functioning, improving team performance, and strengthening the child abuse workforce.
Implementation of team interventions can be challenging, with common barriers including insufficient organizational support, the time and training needed for facilitators, time constraints for participants, and challenges in providing ongoing support and reinforcement of skill use [Reference Chen, Yau and Revere19,Reference Salas, Almeida and Salisbury20]. Many team training efforts have been conducted in resource-rich environments (e.g., business, military) and led by researchers and/or individuals with expertise in quality improvement. Because research and quality improvement roles rarely exist in low-resource settings, extending team training to other settings, like rural CACs, requires developing implementation strategies that rely less on this type of expertise and are appropriate for settings with limited resources and quality improvement infrastructure. Strategies to support the use of team training in low-resource settings can extend the reach and impact of team training efforts.
TeamTRACS: A team training adapted for CAC teams
In a prior study, we adapted TeamSTEPPS, an evidence-based team training intervention, to fit the needs, resources, and context of CAC multidisciplinary teams [Reference McGuier, Feldman and Bay21]. TeamSTEPPS is a widely used team training for healthcare settings developed by the Agency for Healthcare Research and Quality [22]. It targets teamwork knowledge, skills, and attitudes, and it has been shown to improve team functioning and patient outcomes [Reference Capella, Smith and Philp23–Reference Sawyer, Laubach and Hudak25]. Our adapted training, TeamTRACS (Team Training in Roles, Awareness, Communication, and Support), targets knowledge, skills, and attitudes related to team member roles, shared awareness, communication, mutual support, and goal-setting. Adaptations included changes to the setting and population, how the training was delivered and evaluated, and training content (e.g., adding information about roles on CAC teams). The adaptation process and modifications are described elsewhere [Reference McGuier, Feldman and Bay21].
A mixed methods pilot study of TeamTRACS with one CAC team found that TeamTRACS was acceptable, appropriate, and feasible, and participants rated the training content positively [Reference McGuier, Feldman and Bay21]. Participants recommended adding additional structured support before and after the workshop to support the team in taking actions to maximize the training’s impact [Reference McGuier, Feldman and Bay21].
Current study
We conducted a cluster-randomized controlled hybrid effectiveness-implementation trial of TeamTRACS using simultaneous mixed methods. The study had two goals:
Aim 1: Test the effectiveness of TeamTRACS.
Aim 2: Evaluate the acceptability, appropriateness, and feasibility of a structured, self-guided approach to implementation.
For our first aim (QUAN + qual), we evaluated reactions, learning, transfer, and results, aligning with Kirkpatrick’s four-level evaluation framework [Reference Kirkpatrick and Kirkpatrick26]. We tested differences in teamwork knowledge (learning) and skill use (transfer). We also explored the effect of TeamTRACS on team functioning, team performance, and individuals’ work-related outcomes (results). TeamTRACS was hypothesized to improve teamwork knowledge, use of teamwork skills, and team functioning, therefore improving team performance, workforce outcomes, and ultimately, the quality of existing services and outcomes for children who experience maltreatment (Figure 1). For our second aim, we evaluated the acceptability, feasibility, and appropriateness of a self-guided implementation approach. We also described implementation progress and identified facilitators and barriers to self-guided implementation. This trial was registered on clinicaltrials.gov (NCT05370157) and reported in accordance with reporting guidelines (see supplementary material 1).
Figure 1. Hypothesized outcomes and impact of TeamTRACS.
Methods
Community engagement
We partnered with the National Children’s Advocacy Center (NCAC) to refine our implementation guide, recruit CACs to participate in the trial, and deliver TeamTRACS. NCAC provides training and technical assistance and is a trusted resource for CACs. In addition, a community advisory committee of individuals with expertise in CACs provided feedback on research questions, study methods, interpretation of results, and dissemination plans [Reference McGuier, Feldman and Bay21].
Site recruitment
We recruited six rural CACs from the southern United States. The sample size was determined by budgetary constraints. The NCAC advertised the study through emails to CACs, and the research team and NCAC co-hosted an online information session. Interested CACs were directed to discuss their study with their team and then submit a brief survey with basic information about their team (e.g., location, team size) and why they were interested in the study. Participating CACs received $250 to recognize the administrative time needed to support participation.
Participants
Participants were multidisciplinary team members at participating CACs. Only current team members were eligible to participate at each timepoint (see Supplemental Figure 1 in supplementary material 2). CAC staff identified team members before each assessment.
Procedures
Randomization
CACs were randomized to TeamTRACS (n = 4) or a waitlist comparison (n = 2) after baseline data collection. We aimed to balance team size across conditions by creating three pairs of CACs that were as equal in number of participants as possible and then randomizing pairs to condition using a random number generator.
Assessment procedures
Quantitative data were collected through online surveys at baseline (prior to randomization) and follow-up (7 months after baseline; 2–3 months after the TeamTRACS workshop) and brief post-training surveys. Baseline and follow-up surveys were constructed in REDCap, a secure web-based platform, and individual survey invitations were emailed to all team members. Participants received $20 for each completed survey. Brief post-workshop surveys (<15 minutes) were completed on paper for in-person training workshops and through REDCap for virtual workshops.
For teams randomized to TeamTRACS, we conducted periodic reflections on the implementation process with a key informant in each CAC [Reference Finley, Huynh and Farmer27]. Reflections were conducted via videoconference approximately once a month. Participants received $30 for each reflection. The University of Pittsburgh Institutional Review Board approved all procedures. All participants provided informed consent.
Intervention conditions
TeamTRACS Intervention and Implementation Guide
CACs randomized to the intervention condition were offered TeamTRACS facilitated by an experienced team trainer from NCAC who used standard presentation slides and a detailed instructor guide. Each module includes didactic instruction, discussion, and interactive activities (e.g., identifying team strengths/weaknesses, role plays of new skills). Training materials are open source and available from the first author. The workshop could be delivered in-person or virtually; teams chose the delivery method and length (4–8 hours) as part of the implementation process. Intervention CACs also received a companion implementation guide to support their team in implementing and maximizing the effectiveness of TeamTRACS. The guide was designed to facilitate a self-guided implementation approach that requires few resources and supports teams in taking actions to improve teamwork. The implementation guide was adapted from TeamSTEPPS implementation materials and structured in four phases based on the EPIS (Exploration, Preparation, Implementation, and Sustainment) framework [Reference Aarons, Hurlburt and Horwitz28]. It incorporates principles of effective team training [Reference Gregory, Feitosa, Driskell, Salas, Tannenbaum and Cohen29] as well as feedback from our pilot study [Reference McGuier, Feldman and Bay21], our community advisory committee, and NCAC staff.
The 10-step guide integrates multiple discrete implementation strategies [Reference Powell, Waltz and Chinman30], including organize implementation teams (Step 1), assess for readiness and identify barriers and facilitators (Steps 1–3), develop an implementation blueprint (Step 4), and develop and implement tools for quality monitoring (Steps 6–8). Each specific step includes key actions, tools and resources, tips for success, and questions to answer before moving to the next step. The guide also includes suggestions for managing common challenges (e.g., preparing for turnover). Activities were designed to be completed by “change team” members at each CAC.
Waitlist
CACs randomized to the waitlist condition were offered teamTRACS after completion of follow-up data collection. The same trainer facilitated all training workshops.
Measures: Evaluation of TeamTRACS
Background information
As part of the baseline and follow-up surveys, participants reported their discipline, years working with their team, gender, race, ethnicity, and age. CAC staff provided data on attendance at monthly team case review meetings and changes in team members during the study. The trainer recorded attendance at the training workshops.
Reactions to training
Immediately after completing each TeamTRACS module, participants were asked to complete a brief survey. They rated the module’s relevance and usefulness on a 5-point Likert scale (1 “completely disagree” to 5 “completely agree”) and the likelihood of using the skills and strategies from the module on a 5-point scale (1 “not at all likely” to 5 “extremely likely”). Participants also provided open-ended comments on what skills they plan to use, what they found most useful, what they found most challenging, and suggestions for improvement. After completing all modules, participants rated overall relevance, usefulness, and likelihood of using skills/strategies and provided additional open-ended comments. They also rated the acceptability, appropriateness, and feasibility of the overall training; each outcome was assessed with two items rated on a 5-point scale from 1 “completely disagree” to 5 “completely agree” [Reference Weiner, Lewis and Stanick31].
Teamwork knowledge
We developed 16 multiple choice questions (3–4 per module) to assess knowledge taught in the TeamTRACS workshop (e.g., “What is a check-back?”). Item development was informed by TeamSTEPPS knowledge assessments [32,33]. Knowledge was assessed at baseline and follow-up for all participants. Workshop participants also completed knowledge questions as part of the brief survey following each module. Scores were calculated as a percentage of correct responses (possible range 0–100%).
Use of teamwork skills
We developed 41 items reflecting specific teamwork skills taught in TeamTRACS (see Table 1). Most items were adapted from existing scales [Reference Archibald, Trumpower and MacDonald34–Reference Schippers, Den Hartog and Koopman37]. Team members rated how frequently they used each skill. Internal consistency was high (α = 0.96 at both timepoints). Mean total scores were used in analyses.
Table 1. Rating scales and sample items for study measures
Team functioning and performance
We used several measures to assess domains of team functioning targeted by TeamTRACS. Outcomes were understanding and respect of team roles [Reference Cronin, Bezrukova and Weingart38,Reference Mattessich, Murray-Close and Monsey39], shared awareness [Reference Battles and King35,Reference Eby, Meade and Parisi36], information exchange [Reference Kearney, Gebert and Voelpel40], conflict management [Reference Mathieu, Luciano and D’Innocenzo41], psychological safety [Reference Edmondson42], supportive behavior [Reference Aubé and Rousseau43], reflexivity [Reference Schippers, Den Hartog and Koopman37], learning behavior [Reference Edmondson42], and shared goals [Reference Edmondson42]. Subjective team performance was assessed with a validated scale [Reference Edmondson42]. Table 1 shows sample items and how these measures align with TeamTRACS modules. All measures had good internal consistency (α’s = 0.79–0.94). See Supplemental Table 1 in supplementary material 2 for psychometric information.
Team member outcomes
We assessed three individual-level work-related outcomes: emotional exhaustion, work engagement, and turnover intentions. Emotional exhaustion was assessed with a subscale of the Maslach Burnout Inventory [Reference Maslach, Jackson and Leiter44]. Work engagement was assessed with six items [Reference Rich, Lepine and Crawford45]. Lastly, participants rated their intentions to leave their current job [Reference Seashore, Lawler and Mirvis46]. See Table 1 for sample items and Supplemental Table 1 in supplementary material 2 for psychometric information.
Measures: Evaluation of self-guided implementation
Reactions to self-guided implementation
A key informant in each CAC rated the acceptability, appropriateness, and feasibility of self-guided implementation at the beginning, middle, and end of the study. Each outcome was assessed with two items [Reference Weiner, Lewis and Stanick31]. Items were highly intercorrelated (r = 0.58–1.00) and combined by averaging ratings at each timepoint. Internal consistency was excellent at all timepoints (α’s = 0.94–0.96).
Periodic reflections and document collection
Reflections with key informants provided detailed information on the implementation process as it occurred, including steps taken, challenges, and future plans [Reference Finley, Huynh and Farmer27]. We tracked the completion of implementation activities to assess fidelity to the implementation guide. Informants also provided feedback on the implementation guide and suggestions for improvement. We collected copies of relevant documents (e.g., completed worksheets) throughout the implementation process.
Data analyses
Preparatory analyses
For quantitative data, we calculated survey response rates, looked for patterns of missing data, and examined descriptive statistics. We examined within-team agreement on measures of team functioning and performance using the Average Deviation index (ADmd). Lower scores indicate better agreement, and scores below recommended upper limits indicate sufficient agreement within teams to justify aggregation [Reference Burke and Dunlap47].
Our measures of psychological safety, learning behavior, and overall performance exhibited poor team-level agreement in this sample, with ADmd values above the recommended upper limit for one or more teams at each timepoint (Supplemental Table 1 in supplementary material 2). Accordingly, we analyzed these measures as individual-level outcomes instead of creating aggregate team scores. For measures with adequate agreement, we created aggregate scores for each team reflecting the mean of team member scores.
Effectiveness of TeamTRACS (QUAN + qual)
We examined descriptive statistics for acceptability, appropriateness, and feasibility ratings of TeamTRACS. We also examined descriptive statistics for post-workshop ratings of relevance, usefulness, and likelihood of using skills and reviewed open-ended comments about the training. To evaluate the effects of TeamTRACS on team members’ knowledge and skill use, we conducted intent-to-treat analyses testing changes in knowledge and skill use from baseline to follow-up using linear mixed models to account for repeated measures and clustering within teams. We also conducted analyses comparing participants who attended the training to those who did not attend it. Analyses were conducted in R using the lme4 package and restricted maximum likelihood estimation (REML). Models included fixed effects of time (baseline vs. follow-up), intervention condition, and their interaction as well as random effects for person and team. We used the clubSandwich package to calculate cluster robust standard errors.
We also explored the effects of TeamTRACS on team functioning, team performance, and individuals’ work-related outcomes. These measures included both individual-level outcomes (i.e., psychological safety, learning behavior, team performance, work engagement, emotional exhaustion, turnover intentions) and team-level outcomes (i.e., team roles and respect, shared awareness, information exchange, conflict management, supportive behavior, reflexivity, clear direction). Changes in individual-level outcomes were tested using linear mixed models as described above. Because this study was underpowered for team-level outcomes, we plotted data for team-level measures of team functioning and performance to visually examine team-level scores at baseline and follow-up and did not conduct hypothesis testing.
Evaluation of self-guided implementation (quan + QUAL)
The second goal of this study was to evaluate a self-guided approach to implementing TeamTRACS. We examined descriptive statistics for acceptability, appropriateness, and feasibility ratings of self-guided implementation. We assessed the proportion of implementation activities completed in each intervention CAC as a measure of fidelity to the implementation guide.
We used rapid analysis [Reference Gale, Wu and Erhardt48] to create summaries of periodic reflections. Interviewers reviewed automatically generated interview transcripts and created summaries using a spreadsheet template. The template listed multiple domains (e.g., implementation progress, positive feedback, suggestions for change) and provided space for key points and exemplar quotes in each domain. At the end of the study, we integrated summaries from individual interviews into an overarching narrative of implementation for each team. Then, we compared and contrasted narratives across teams to identify themes, including facilitators and barriers to self-guided implementation.
Results
CAC characteristics
Participating CACs were in five states in the southern region of the United States. Twelve CACs expressed interest, seven CACs were invited to participate, and six CACs agreed. Of the CACs that were not invited to participate, two were in urban areas, two were participating in other team improvement projects, and one served multiple urban and rural counties with combined case review meetings. All participating teams reported serving rural populations; county populations ranged from approximately 36,000 to 56,000.
Participation rates
Table 2 shows the number of survey responses and participation rates by team and overall. As expected, team members changed during the study; there were 172 unique participants across timepoints (Supplemental Figure 1 in supplementary material 2). Key informants were the team coordinator at each CAC; some also had another role (e.g., advocate, forensic interviewer). For 3 of 4 teams, one key informant completed all periodic reflections and ratings of the self-guided implementation process. In the remaining team, the key informant changed partway through the study because the initial informant went on leave.
Table 2. Study participation and training workshop format
Participant characteristics
Participant characteristics were assessed using baseline responses when available and follow-up responses for those without baseline responses. Participants identified as non-Hispanic white (73%), Hispanic/Latino of any race (7%), American Indian/Alaska Native (6%), Black/African American (5%), and Asian (1%); the remainder (8%) were missing or selected “prefer not to answer.” Participants identified as female (66%) and male (27%); the remainder (7%) selected “prefer not to answer.” Supplemental Figure 2 in supplementary material 2 shows participants’ professional disciplines. Most participants had worked with their team for more than 1 year (13% <1 year; 33% 1–3 years; 23% 4–6 years; 32% 7 years or more).
Training delivery
Table 2 shows the workshop format and length chosen by each participating team. In-person training workshops were conducted in a single location chosen by the team, and the virtual training was delivered through Zoom with participants joining individually from locations of their choosing. Because teams determined who to invite, the proportion of invited members who participated is unknown. All workshops were facilitated by the same experienced team trainer using the same materials (i.e., training slides, instructor guide). Although workshop timing and pace varied, all content was delivered during each workshop. Some teams provided food or other incentives for team members to attend; the research team did not provide any incentives.
Reactions to training
Most training attendees completed ratings of the workshop content (n = 41; 77%). Overall, training attendees rated the content highly relevant (M = 4.83) and useful (M = 4.83) and reported that they were very to extremely likely to use the skills taught (M = 4.59). Ratings for individual modules were similarly high. Attendees agreed that the training was acceptable (M = 4.63, range 4–5), appropriate (M = 4.70, range 4–5), and feasible (M = 4.54, range 3–5). Open-ended comments were positive, with some attendees suggesting that the training should be longer.
Descriptive data
Supplemental Table 2 in supplementary material 2 provides descriptive statistics and intercorrelations for baseline measures of teamwork knowledge and skill use, team functioning, team performance, and team member outcomes. Most measures were correlated with one another, except for teamwork knowledge scores which were not significantly correlated with any other measures.
Changes in teamwork knowledge
In an intent-to-treat mixed model (N = 165), there was a significant interaction of condition and time; participants in teams randomized to TeamTRACS exhibited higher knowledge scores at follow-up than participants in waitlist teams (Table 3, Model 1; Figure 2a). The same interaction was found when comparing participants who attended a TeamTRACS workshop to those that did not attend (Table 3, Model 2; Figure 2b). Knowledge scores for those who attended a TeamTRACS workshop were high immediately after the workshop (average score = 90%). These results indicate that TeamTRACS was effective in increasing teamwork knowledge.
Table 3. Changes in teamwork knowledge and skill use by intervention condition and training workshop attendance
* p < 0.05.
Figure 2. Teamwork knowledge and skills: Marginal effects of intervention condition & timepoint.
Note: (a) Teamwork knowledge: Intent-to-treat model. (b) Teamwork knowledge: Workshop attendance model. (c) Teamwork skills: Intent-to-treat model.
Changes in teamwork skill use
Use of teamwork skills was high at both timepoints (baseline M = 4.17, follow-up M = 4.19). Both intent-to-treat and workshop attendance mixed models found no significant effects of condition, time, or their interaction on skill use (Table 3; Figure 2c).
Exploratory analyses: Changes in team functioning and performance
Individuals’ perceptions of team outcomes
Intent-to-treat linear mixed models found no significant effects of condition, time, or their interaction on individuals’ perceptions of psychological safety, learning behavior, or team performance. See Supplemental Figures 3–5 in supplementary material 2 for plots of marginal effects.
Team-level outcomes
Plots of team-level outcomes by condition and timepoint are presented in Supplemental Figures 6–12 in supplementary material 2. Most teams exhibited small and inconsistent changes over time. One intervention team (TeamTRACS Team 3) exhibited increases greater than 0.35 on 6 of 7 measures, indicating substantial improvement in several outcomes. The largest increases were in reflexivity (mean difference = 0.85) and clear direction (mean difference = 1.12). This team had the lowest baseline scores on 6 of 7 measures, indicating poorer baseline functioning than the other participating teams. No other teams exhibited consistent patterns of change.
Exploratory analyses: Changes in workforce outcomes
Intent-to-treat linear mixed models found no significant effects of condition, time, or their interaction on individuals’ work engagement, emotional exhaustion, or turnover intentions. Plots of marginal effects are provided in Supplemental Figures 13–15 in supplementary material 2. Individuals in the comparison condition reported lower emotional exhaustion at baseline that increased over time (predicted means = 1.92 baseline and 2.21 follow-up), while individuals in the intervention group were relatively stable (predicted means = 2.30 baseline and 2.21 follow-up); however, this interaction was not significant in mixed models (p = .12).
Evaluation of self-guided implementation
Reactions to self-guided implementation
Key informants provided ratings of the acceptability, appropriateness, and feasibility of self-guided implementation. Ratings were positive, with average scores greater than 4 (“agree”) at the beginning, middle, and end of the study (M = 4.17, M = 4.17, M = 4.29, respectively). Participants provided substantial feedback during reflections and appeared comfortable reflecting on their progress. Participants appreciated the structure of the guide and provision of worksheets (see Table 4).
Table 4. Qualitative findings: Facilitators and barriers to implementation of TeamTRACS
Quotes were lightly edited for clarity.
Progress through self-guided implementation
Only one team completed all steps in the implementation guide. Some teams continued to progress through the implementation guide after the end of the study and completion of the follow-up survey. See supplementary material 2 in supplementary material 2 for a table depicting the progress each team made through self-guided implementation and summaries of implementation progress in each team based on the findings from periodic reflections. Most teams did not complete later implementation steps intended to encourage teamwork skill use, which may have limited the impact of TeamTRACS on skill use and team outcomes.
Challenges to self-guided implementation
Most teams (3 of 4) struggled to form and maintain a change team. The only team to complete all implementation steps was the team that successfully created a change team, suggesting that early and consistent engagement of a change team facilitates implementation progress. Without a change team, responsibility for implementation typically fell to one CAC staff member (see Table 4).
All teams experienced frequent turnover and understaffing that affected their engagement in implementation as well as attendance at the training workshop. In 2 of 4 teams, the primary contact person and implementation lead was promoted during the study. In one team, this person continued to lead implementation efforts, allowing for continued progress, while in the second team responsibility was passed to another person and progress slowed. All teams reported that attendance at the training workshop was limited because of emergent demands on team members. For example, in one team, all child welfare workers were unable to attend the training workshop because they were required to provide coverage for a neighboring county with severe understaffing.
Participants also identified competing priorities that made implementation challenging, particularly within the short timeframe of the study. Key informants described TeamTRACS implementation and strengthening team functioning as an important part of their job but not part of their daily tasks (see Table 4). Informants also reported holiday-related demands (e.g., events, vacations) that slowed implementation progress after the training workshops were conducted. For some teams, qualitative data suggested continued progress and improvements after the follow-up survey was completed.
Discussion
This study tested the effectiveness of TeamTRACS, a team training intervention for CAC multidisciplinary teams, and evaluated its implementation. It contributes to existing scientific literature by examining the effectiveness of team training in fluid cross-boundary teams [Reference Driskell, Funke and Tolston5] and extending outside of hospital and healthcare settings where most research on interprofessional collaboration has been conducted. It also provides insight into how team training is implemented in real-world settings.
Our evaluation of TeamTRACS effectiveness was grounded in Kirkpatrick’s framework [Reference Hughes, Gregory and Joseph16,Reference Kirkpatrick and Kirkpatrick26]. We evaluated reactions to training, learning (i.e., knowledge change), transfer (i.e., change in skill use), and results (i.e., changes in team functioning and performance). Reactions to TeamTRACS were very positive, and there were significant increases in knowledge for participants in the TeamTRACS condition.
Self-reported skill use was high and did not differ over time or by condition. All participating teams sought out the training and described challenges in the domains targeted by the training, suggesting a need for improvement despite high ratings of skill use. Our measure of skill use was developed for this study, and because baseline scores were high, there was little room for improvement. Participants’ ratings were likely inflated by self-report biases, including social desirability bias [Reference Rosenman, Tennekoon and Hill49]. It is also possible that skill practice during the workshop increased participants’ understanding and awareness of their teamwork skills, leading participants to recalibrate their self-ratings downward following the training and resulting in an underestimate of training effectiveness [Reference Rosenman, Tennekoon and Hill49]. Relatively few studies have examined changes in skill use after interprofessional training, and skill use is typically self-reported [Reference Reeves, Goldman and Gilbert50]. Future research could use observational measures, standardized role plays, or ratings of other participants’ skill use to reduce self-report biases and better measure teamwork skill use.
Most changes in team functioning and performance were small and inconsistent, although one TeamTRACS team exhibited substantial and consistent improvements. Our findings suggest that team training can improve teamwork in fluid cross-boundary teams; however, incomplete implementation of TeamTRACS, as discussed below, may have limited its effectiveness. There were no significant effects of TeamTRACS on individuals’ work engagement, emotional exhaustion, or turnover intentions.
Users viewed the implementation guide and self-guided process favorably. However, only one team completed all steps in the implementation guide. Challenges to implementation included difficulty forming and maintaining change teams, turnover and understaffing, competing priorities, and the short timeframe of the study. Turnover is a common barrier to implementation in team-based settings [Reference McGuier, Kolko and Aarons9] and may be especially relevant for interventions targeting the team. Team training efforts that focus on changing team processes and structures (e.g., revising protocols, adjusting meeting agendas), rather than individual behaviors, may be more effective in settings with high turnover and fluidity. Future studies should also examine ways to assess the extent of team fluidity to better understand its effects.
In this study, training workshops were conducted at the end of the year, prior to holiday events and vacations, and most teams only had a month (January) to make changes before the follow-up survey was conducted. The short timeframe for follow-up may have limited our ability to find changes. Final reflections suggested that some teams continued to progress through the implementation guide after the follow-up survey. In a test of team training in a surgical ward, Aaberg and colleagues [Reference Aaberg, Ballangrud and Husebø51] were able to complete all implementation steps in a 12-month period. The researchers in that study were actively involved as part of the change team, which may have facilitated progress [Reference Aaberg, Ballangrud and Husebø51]. We found that periodic reflections likely served as a prompt for greater attention to TeamTRACS implementation as well as attention to team functioning. Future efforts to implement TeamTRACS should consider a longer implementation period and integrating recurring check-ins with someone external to the team to help teams maintain their focus.
The limited impact of TeamTRACS on skill use and team-level outcomes may be at least partially because most teams did not complete later implementation steps intended to encourage skill use. Interestingly, the team that struggled the most to implement TeamTRACS was the only team with consistent improvements in team outcomes from baseline to follow-up. This may be because the team’s baseline team scores were low, and there was more room for improvement. Persistent attention to how the team was working and greater reflexivity may have led to improvements even without completion of all implementation steps. This result is encouraging, as it suggests that even incomplete implementation of team training can lead to improvements in fluid cross-boundary teams. However, these findings may simply reflect regression to the mean and should be interpreted cautiously given the overall null findings for team-level outcomes.
Limitations and future directions
Although this study found significant improvements in teamwork knowledge, it is unclear how meaningful these changes are, as teamwork knowledge was not correlated with skill use or other outcomes at baseline. Although our study is relatively large compared to other studies of team training, we did not have enough power to empirically test our full theory of change (Figure 1). Similarly, with only six teams, this study was not powered to detect changes in team-level measures of team functioning and performance. Larger studies are needed to test the effects of TeamTRACS on team-level outcomes. Survey participation rates varied by team and timepoint, and non-participants may have perceived the team differently. Periodic reflections were conducted with the same informant in most CACs. Although this allowed for continuity over time, other informants may have had different perspectives on implementation. In addition, there was no attention control in the comparison condition.
Our rigorous mixed methods study contributes to the limited literature on effective strategies to improve teamwork and interprofessional collaboration [Reference Reeves, Pelone and Harrison52] and extends this work to fluid teams operating outside of healthcare settings [Reference Driskell, Funke and Tolston5]. Team trainings such as TeamTRACS are flexible approaches to improving teamwork that can be integrated with other training approaches, such as simulation training, to further enhance their effectiveness. We found that our structured, self-guided implementation approach was feasible and identified facilitators and barriers to implementation progress. This study was conducted in real-world settings using existing training infrastructure, increasing the generalizability of our findings to other settings with limited resources to support teams.
Conclusions
We tested the effectiveness of TeamTRACS in fluid cross-boundary teams serving children experiencing maltreatment. TeamTRACS was positively received and increased teamwork knowledge, but there were no effects on teamwork skill use or individuals’ work-related outcomes, and team-level changes were inconsistent. We also evaluated a self-guided implementation approach, advancing the development of realistic implementation strategies for high-need, low-resource settings. Self-guided implementation was positively received but challenging for teams to complete within a short timeframe. Additional efforts are needed to build capacity for implementation and continuous improvement in low-resource settings. Interventions to improve teamwork in fluid teams have the potential to improve interprofessional collaboration across systems and ultimately improve service quality and outcomes.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/cts.2025.10203.
Acknowledgments
We greatly appreciate the valuable contributions of our community advisory committee members and the support of the National Children’s Advocacy Center. We would also like to thank the participating Children’s Advocacy Centers and every team member who took the time to participate in this study.
Author contributions
Elizabeth A. McGuier: Conceptualization, data curation, formal analysis, funding acquisition, investigation, methodology, project administration, supervision, visualization, writing – original draft, writing – review & editing; Jaely D. Wright: Data curation, formal analysis, project administration, visualization, writing – review & editing; Greg Flett: Conceptualization, investigation, project administration, writing-review & editing; Scott D. Rothenberger: Formal analysis, methodology, writing – review & editing; Eduardo Salas: Conceptualization, writing – review & editing; David J. Kolko: Conceptualization, funding acquisition, methodology, supervision, writing – review & editing.
Funding statement
This work was supported by the Clinical and Translational Sciences Institute at the University of Pittsburgh (UL1 TR001857) and the National Institute of Mental Health (MH123729). REDCap is supported by the University of Pittsburgh Clinical and Translational Sciences Institute (UL1 TR001857). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Competing interests
The authors report there are no competing interests to declare.
Open access
