Search results

Our initial search, after removal of duplicates, yielded 7181 results. The second search (August 2020-March 2022) captured an additional 1341 results. The cited reference search yielded 1961 results. A total of 10,489 results were included in title/abstract review. Figure 2 provides a PRISMA flow diagram for included studies. After full-text review, 58 articles from 55 studies were included in analyses [40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97].

Fig. 2

PRISMA flow diagram of included articles. From: Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71. https://doi.org/10.1136/bmj.n71. For more information, visit: http://www.prisma-statement.org/

As shown in Fig. 3, publications on teamwork and implementation have increased substantially since 2000. Three articles on this topic (5%) were published between 2000 and 2007, 14 (24%) between 2008 and 2015, and 41 (71%) between 2016 and early 2023.

Fig. 3

Included articles by year of publication

Study characteristics

Interrater agreement was good for assessment of study quality (81% agreement on MMAT questions) and ratings of relevance (88% agreement). There were 20 high quality articles, 23 moderate quality articles, and 15 low quality articles. Fourteen articles were rated as high relevance, 22 as moderate, and 22 as low relevance. Only 4 were rated as both high quality and high relevance. We report study characteristics for all 58 eligible articles. Our narrative synthesis includes 32 articles categorized as moderate/high quality and moderate/high relevance; it excludes 26 articles categorized as low quality and/or relevance.

Studies were conducted in inpatient healthcare (n = 22), outpatient/ambulatory healthcare (n = 21), mental health settings (n = 9), and other settings (e.g., residential facilities, multiple settings; n = 6). There were 33 qualitative, 15 quantitative, and 10 mixed methods studies. All quantitative studies were descriptive observational studies.

Most studies examined team processes/states (n = 53); fewer examined team inputs (n = 27). Only two studies examined a team effectiveness outcome. The most common implementation outcomes were fidelity (n = 16) and other specified implementation outcomes (e.g., “extent of use,” “implementation success”) (n = 15). Less frequently identified implementation outcomes included adoption (n = 5), sustainment (n = 4), reach (n = 4), and perceptions of the innovation (e.g., acceptability, appropriateness, feasibility; n = 3). Approximately one-third of studies (n = 21) did not report specific implementation outcomes but described implementation determinants (i.e., barriers and facilitators).

Synthesis: team inputs & implementation outcomes

Team inputs examined in studies included team stability/instability and staffing shortages, aspects of team structure and composition, interdependence, and hierarchy and professional roles. Quantitative findings are presented in Table 1. A CERQual Summary of Qualitative Findings related to team inputs is shown in Table 2. A CERQual Evidence Profile is provided in Additional File 2 (Table A1).

Table 1 Summary of quantitative studies: team inputs & implementation outcomesTable 2 Summary of qualitative findings: team inputs & implementation outcomesTeam stability/instability and staffing shortages

Team stability/instability (i.e., consistency in membership over time) was examined in one mixed methods study [48, 49] and three qualitative studies [70, 81, 94]. A study of surgical teams found variations in membership stability but no association between stability and “implementation success” (i.e., composite measure based on number of uses of new technique, proportion of uses, and changes in use) [48, 49]. The authors suggested that stability facilitates the development of team coordination but that selecting small and exclusive teams may limit the spread of innovations within the organization. Another study found that a dedicated and stable team in which members were selected and trained together in the use of a new surgical technique led to quicker uptake and better integration into practice, theorizing that dedicated and stable teams increased trust, motivation, and collaborative problem-solving [81]. However, dedicated teams were difficult to sustain, and some sites instead used rotating team members from a larger pool of trained staff. In rural primary care, stability of team members facilitated sustainment of memory care clinics [70]. Lastly, another study in primary care found mixed perceptions of stable vs. rotating staff when adding a new team role (i.e., health coach); some team members liked rotating through different roles while others wanted more stability [94]. Across studies, we found that dedicated and stable team members facilitate implementation while instability in team membership is a barrier to implementation (moderate confidence).

Qualitative studies identified staffing shortages and turnover on teams as barriers to implementation [50, 67, 75, 78, 92]. In Veterans Health Administration (VA) clinics, “inadequate staffing posed an insurmountable barrier,” hindering communication and delivery of optimal care during the implementation of the patient-centered medical home (PCMH) model [92]. Similarly, staff shortages, turnover, and high workloads hindered guideline implementation in Kenyan hospitals [75]. Two studies found negative impacts of staffing shortages and turnover on sustainment. Staff turnover contributed to discontinuity in Dialectical Behavior Therapy (DBT) team members [78], and appropriate staffing (i.e., ensuring manageable workloads) and blocking time for team members were identified as critical to sustainment of a team-based model in the VA [67]. We found that staffing shortages and turnover hinder implementation (high confidence).

Team structure/composition

Studies examined multiple aspects of team structure and composition, specifically team size, workload, longevity (i.e., how long team members had worked together), history of change, and team member characteristics. Team size was examined in two studies of DBT. In a mixed methods study, team size was positively correlated with fidelity, and qualitative data suggested that team size may increase as a result of successful implementation [47]. In contrast, another study found that DBT team size was not associated with the number of DBT components adopted and was negatively associated with reach, suggesting reach may reflect high workloads [72]. In VA mental health clinics, team workload (i.e., number of patients seen) was negatively associated with sustainment of trauma-focused therapies [68]. In these studies, team longevity and history of change were not associated with implementation outcomes [47, 68]. Team member characteristics, specifically team member competency/expertise, experience, and commitment/engagement, were identified as facilitators of implementation in some qualitative studies [40, 70, 81, 84, 95].

Overall, few findings could be made from quantitative studies examining team structure and composition. Two studies of team size found mixed results, and workload, longevity, and history of change were examined in only one study each. Across qualitative studies, we found team member competency/expertise, experience, and commitment/engagement facilitate implementation (moderate confidence).

Team interdependence

One quantitative study examined team interdependence [65]. In multidisciplinary child abuse teams implementing a mental health screening/referral protocol, task interdependence (i.e., reliance on team members to share resources and coordinate workflows) was positively associated with reach but not time to adoption. Outcome interdependence (i.e., extent to which outcomes are evaluated at the team vs. individual level) was significantly negatively correlated with time to adoption but not reach. Neither task nor outcome interdependence were associated with team members’ perceptions of acceptability, appropriateness, or feasibility of the innovation [65]. Because only one study examined interdependence, no review findings were made.

Hierarchy & professional roles

Hierarchy, power distributions, and rigid roles were identified as barriers to implementation in several qualitative studies [50, 53, 74, 97]. Flatter hierarchies (i.e., more equal distribution of power and authority) supported guideline implementation in pediatric primary care; practices with low compliance to guidelines had more hierarchical relationships while practices with high compliance had more shared decision-making [97]. In a setting with hierarchy and rigid division of roles, nurses trained in an innovation reported concern that their decisions would be questioned by physicians without expertise in the innovation but more authority [74]. Similarly, in surgical teams, rigid professional roles and a hierarchical team culture constrained open discussion and created contention over how and when a “time-out” should be completed, resulting in inconsistent use and poor fidelity [50, 53]. Across studies, we found that in multidisciplinary settings, rigid professional roles, hierarchical relationships, and power differentials are barriers to implementation (moderate confidence).

Summary of team inputs & implementation outcomes

There was no overlap among team input variables and implementation outcomes examined in quantitative studies (Table 1). Accordingly, we were unable to generate estimates of effects or ratings of evidence quality. Qualitative review findings are shown in Table 2. We found: 1) Dedicated and stable team members facilitate implementation while instability in team membership is a barrier to implementation (moderate confidence); 2) Staffing shortages and turnover hinder implementation (high confidence); 3) Team member competency/expertise, experience, and commitment/engagement facilitate implementation (moderate confidence); and 4) In multidisciplinary settings, rigid professional roles, hierarchical relationships, and power differentials are barriers to implementation (moderate confidence).

Synthesis: team processes/states & implementation outcomes

Studies examined overall team functioning as well as specific affective states, behavioral processes, and cognitive states. Quantitative findings are presented in Table 3, and a GRADE Evidence Profile is provided in Additional File 2 (Table A2). A CERQual Summary of Qualitative Findings related to team processes and states is shown in Table 4. A CERQual Evidence Profile is provided in Additional File 2 (Table A3).

Table 3 Summary of quantitative studies of team processes/states & implementation outcomesTable 4 Summary of qualitative findings: team processes/states & implementation outcomesOverall team functioning

Nine studies examined quantitative associations between overall team functioning and implementation outcomes. Team functioning was positively associated with intervention fidelity in 2 of 3 studies. One study examined implementation of transition programs for adolescents with chronic health conditions in 29 teams. More positive team climate, measured by the Team Climate Inventory (i.e., shared vision, participative safety, task orientation, support for innovation), at study start was associated with greater improvements in quality of chronic care delivery one year later [45]. Additionally, improvements in team climate were associated with greater improvement in care delivery [45]. These findings were consistent across teams working with different patient populations, suggesting the influence of team climate generalizes across teams and settings [45]. Greater team climate for innovation was also associated with greater fidelity (i.e., implementation of more program elements) among DBT teams [47]. In contrast, no significant associations were found between team climate and fidelity to a multifaceted cardiovascular disease management intervention, with qualitative data suggesting variation in the influence of teamwork across practices [77]. There was no overlap in the metrics of association reported in these studies; therefore, we were unable to generate an estimate of the effect of team functioning on fidelity. The quality of the evidence for fidelity was rated very low because of serious methodological limitations, serious inconsistency, and very serious imprecision due to the small number of studies.

Three studies examined associations between teamwork and adoption, with no significant associations found. The first study found that teamwork climate (i.e., perceived quality of collaboration between personnel) was not significantly associated with adoption of a comprehensive safety program in intensive care units, although there were associations between adoption and organizational constructs (e.g., lower safety climate, more management support) [