Interventions to improve team effectiveness within health care: a systematic review of the past decade

The search strategy was developed with the assistance of a research librarian from a medical library who specializes in designing systematic reviews. The search combined keywords from four areas: (1) team (e.g. team, teamwork), (2) health care (e.g. health care, nurse, medical, doctor, paramedic), (3) interventions (e.g. programme, intervention, training, tool, checklist, team building), (4) improving team functioning (e.g. outcome, performance, function) OR a specific performance outcome (e.g. communication, competence, skill, efficiency, productivity, effectiveness, innovation, satisfaction, well-being, knowledge, attitude). This is similar to the search terms in the initial systematic review [8]. The search was conducted in the following databases: EMBASE, MEDLINE Ovid, Web of Science, Cochrane Library, PsycINFO, CINAHL EBSCO, and Google Scholar. The EMBASE version of the detailed strategy was used as the basis for the other search strategies and is provided as additional material (see Additional file 1). The searches were restricted to articles published in English in peer-reviewed journals between 2008 and July 2018. This resulted in 5763 articles. In addition, 262 articles were identified through the systematic reviews published in the last decade [3, 4, 7, 9‐28]. In total, 6025 articles were screened.

This systematic review aims to capture the full spectrum of studies that empirically demonstrate how healthcare organizations could improve team effectiveness. Therefore, the following studies were excluded:

In addition, how teams were defined was not a selection criterion. Given the variety of teams in the healthcare field, we found it acceptable if studies claim that the setting consists of healthcare teams.

Figure 1 summarizes the search and screening process according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) format. A four-stage process was followed to select potential articles. We started with 6025 articles. First, each title and abstract was subjected to elimination based on the aforementioned inclusion and exclusion criteria. Two reviewers reviewed the title/abstracts independently. Disagreement between the reviewers was settled by a third reviewer. In case of doubt, it was referred to the next stage. The first stage reduced the number of hits to 639. Second, the full text articles were assessed for eligibility according to the same set of elimination criteria. After the full texts were read by two reviewers, 343 articles were excluded. In total, 297 articles were included in this review. Fourth, the included articles are summarized in Table 1. Each article is described using the following structure:

Studies can also be upgraded or downgraded based on additional criteria. For example, a study is downgraded by one category in the event there are important inconsistencies. Detailed information is provided as additional material (see Additional file 2).

The categorization of our final set of 297 articles is the result of three iterations. First, 50 summarized articles were categorized using the initial categorization: team training (subcategories: CRM-based training, simulation training, interprofessional training, and team training), tools, and organizational intervention [8]. Based on this first iteration, the main three categories (i.e. training, tools, and organizational interventions) remained unchanged but the subcategorization was further developed. Training, related to the subcategory “CRM-based training”, “TeamSTEPPS” was added as a subcategory. The other subcategories (i.e. simulation training, interprofessional training, and team training) remained the same. Tools, the first draft of subcategories, entailed Situation, Background, Assessment, and Recommendation (SBAR), checklists, (de)briefing, and task tools. Two subcategories of organizational intervention (i.e. programme and (re)design) were created, which was also in line with the content of this category in the original literature review. Second, 50 additional articles were categorized to test and refine the subcategories. Based on this second iteration, the subcategories were clustered, restructured and renamed, but the initial three main categorizations remained unaffected. The five subcategories of training were clustered into principle-based training, method-based training, and general team training. The tools subcategories were clustered into structuring, facilitating, and triggering tools, which also required two new subcategories: rounds and technology. Third, the remaining 197 articles were categorized to test the refined categorization. In addition, the latter categorization was peer reviewed. The third iteration resulted in three alterations. First, we created two main categories based on the two subcategories “organizational (re)design” and “programme” (of the third main categorization). Consequently, we rephrased “programme-based training” into “principle-based training”. Second, the subcategories “educational intervention” and “general team training” were merged into “general team training”. Consequently, we rephrased “simulation training” into “simulation-based training”. Third, we repositioned the subcategories “(de)briefing” and “rounds” as structuring tools instead of facilitating tools. Consequently, we merged the subcategories “(de)briefing” and “checklists” into “(de)briefing checklists”. Thereby, the subcategory “technology” became redundant.

CRM and TeamSTEPPS are well-known principle-based trainings that aim to improve teamwork and patient safety in a hospital setting. Both types of training are based on similar principles. CRM is often referred to as a training intervention that mainly covers non-technical skills such as situational awareness, decision making, teamwork, leadership, coping with stress, and managing fatigue. A typical CRM training consists of a combination of information-based methods (e.g. lectures), demonstration-based methods (e.g. videos), and practice-based methods (e.g. simulation, role playing) [9]. However, CRM has a management concept at its core that aims to maximize the use of all available resources (i.e. equipment, time, procedures, and people) [324]. CRM aims to prevent and manage errors through avoiding errors, trapping errors before they are committed, and mitigating the consequences of errors that are not trapped [325]. Approximately a third of CRM-based trainings include the development, redesign or implementation of learned CRM techniques/tools (e.g. briefing, debriefing, checklists) and could therefore also be categorized in this review under programme [39, 40, 42, 51, 56, 58, 59, 61, 62].

The studies show a high variety in the content of CRM training and in the results measured. The majority of the studies claim an improvement in a number of non-technical skills that were measured, but some also show that not all non-technical skills measured were improved [43, 47, 66]. Moreover, the skills that did or did not improve differed between the studies. A few studies also looked at outcome measures (e.g. clinical outcomes, error rates) and showed mixed results [49, 52, 53]. Notable is the increasing attention toward nursing CRM, which is an adaptation of CRM to nursing units [66, 67]. Most studies delivered a low to moderate quality level of evidence. Although most studies measured the effect of CRM over a longer period of time, most time periods were limited to one or two evaluations within a year. Savage et al. [58] and Ricci et al. [56] note the importance of using a longer time period.

Simulation-based training uses a specific method as its core, namely, simulation, which refers to “a technique to replace or amplify real-patient experiences with guided experiences, artificially contrived, that evokes or replicates substantial aspects of the real world in a fully interactive manner” [326]. The simulated scenarios that are used can have different forms (e.g. in situ simulation, in centre simulation, human actors, mannequin patients) and are built around a clinical scenario (e.g. resuscitation, bypass, trauma patients) aiming to improve technical and/or non-technical skills (e.g. interprofessional collaboration, communication). We only identified studies in a hospital setting, which were mostly focussed on an emergency setting. All studies reported improvements in some non-technical skills (e.g. teamwork behaviour, communication, shared mental model, clarity in roles and responsibilities). In addition, some studies report non-significant changes in non-technical skills [98, 137, 140, 155]. Some studies also looked at technical skills (e.g. time spend) and presented mixed results [63, 112, 152, 159]. Sixty-nine studies focused on simulation-based training, of which 16 studies delivered a moderate to high quality of evidence.

Tools are instruments that could be implemented relatively independently in order to structure, facilitate or trigger teamwork.

Teamwork can be structured by using the structured communication technique SBAR (Situation, Background, Assessment, and Recommendation), (de)briefing checklists, and rounds.

SBAR is often studied in combination with strategies to facilitate implementation, such as didactic sessions, training, information material, and modifying SBAR material (e.g. cards) [202, 204, 206‐208, 211]. In addition, this subcategory entails communication techniques similar or based on SBAR [203, 205, 209, 210, 212]. One study focused on nursing homes, while the remaining studies were performed in a hospital setting. Most studies found improvements in communication; however, a few found mixed results [208, 209]. Only (very) low-level evidence studies were identified.

Briefings and debriefings create an opportunity for professionals to systematically communicate and discuss (potential) issues before or after delivering care to a patient, based on a structured format of elements/topics or a checklist with open and/or closed-end questions. Studies on (de)briefing checklists often evaluate the implementation of the World Health Organization surgical safety checklist (SSC), a modified SSC, SSC-based checklist, or a safety checklist in addition to the SSC. The SSC consists of a set of questions with structured answers that should be asked and answered before induction of anaesthesia, before skin incision, and before the patient leaves the operating theatre. In addition, several studies presented checklists aiming to better manage critical events [221, 223, 233]. Only one study on SSC was conducted outside the surgery department/operating theatre (i.e. cardiac catheterization laboratory [222]). However, similar tools can also be effective in settings outside the hospital, as shown by two studies that focused on the long-term care setting [249, 260]. Overall, included studies show that (de)briefing checklists help improve a variety of non-technical skills (e.g. communication, teamwork, safety climate) and objective outcome measures (e.g. reduced complications, errors, unexpected delays, morbidity). At the same time, some studies show mixed results or are more critical of its (sustainable) effect [215, 222, 231, 242]. Whyte et al. [262] pointed out the complexity of this intervention by presenting five paradoxical findings: team briefings could mask knowledge gaps, disrupt positive communication, reinforce professional divisions, create tension, and perpetuate a problematic culture. The quality of evidence varied from high to very low (e.g. Whyte et al. [262]), and approximately one third presented a high or moderate quality of evidence. Debriefings can also be used as part of a training, aiming to provide feedback on trained skills. Consequently, some articles focused on the most suitable type of debriefing in a training setting (e.g. video-based, self-led, instructor-led) [245, 246, 253, 263] or debriefing as reflection method to enhance performance [258, 261].

Teamwork can be facilitated through technology. Technology, such as telecommunication, facilitates teamwork as it creates the opportunity to involve and interact with professionals from a distance [271‐273]. Technology also creates opportunities to exchange information through information platforms [276, 277]. Most studies found positive results for teamwork. Studies were performed in a hospital setting and presented a level of evidence varying from moderate to very low.

Teamwork could be triggered by tools that monitor and visualize information, such as (score) cards and dashboards [278, 279, 281, 283, 284]. The gathered information does not echo team performance but creates incentives for reflecting on and improving teamwork. Team processes (e.g. trust, reflection) are also triggered by sharing experiences, such as clinical cases and stories, thoughts of the day [280, 282]. All seven studies showed improvements in non-technical skills and had a very low level of evidence.

In contrast with the previous two categories, organizational (re)design is about changing organizational structures. Interventions can be focused on several elements within a healthcare organization, such as the payment system [292] and the physical environment [299], but are most frequently aimed at standardization of processes in pathways [286, 288] and changing roles and responsibilities [287, 289, 298], sometimes by forming dedicated teams or localizing professionals to a certain unit or patient [290, 291, 295, 300]. Most studies found some improvements of non-technical skills; however, a few found mixed results. Only four studies had a moderate level of evidence, and the others had a (very) low level.

A programme most frequently consists of a so-called Human Resource Management bundle that combines learning and educational sessions (e.g. simulation training, congress, colloquium), often multiple tools (e.g. rounds, SBAR), and/or structural intervention (e.g. meetings, standardization). Moreover, a programme frequently takes the organizational context into account: developing an improvement plan and making choices tailored to the local situation. A specific example is the “Comprehensive Unit-Based Safety Program” (CUSP) that combines training (i.e. science of safety training educational curriculum, identify safety hazards, learn from defects) with the implementation of tools (e.g. team-based goal sheet), and structural intervention (i.e. senior executive partnership, including nurses on rounds, forming an interdisciplinary team) [309, 319, 322]. Another example is the medical team training (MTT) programme that consists of three stages: (1) preparation and follow-up, (2) learning session, (3) implementation and follow-up. MTT combines training, implementation of tools (briefings, debriefing, and other projects), and follow-up coaching [5, 304, 305, 316]. MMT programmes are typically based on CRM principles, but they distinguish themselves from the first category by extending their programme with other types of interventions. Most studies focus on the hospital setting, with the exception of the few studies performed in the primary care, mental health care, and healthcare system. Due to the wide range of programmes, the outcomes were diverse but mostly positive. The quality of evidence varied from high to very low.

The main limitation of this review is that we cannot claim that we have found every single study per subcategory. This would have required per subcategory an additional systematic review or an umbrella review, using additional keywords. As we identified a variety of literature reviews, future research should focus on umbrella reviews in addition to new systematic literature reviews. Note that we did find more studies per subcategory, but they did not meet our inclusion criteria. For example, we excluded multiple studies evaluating surgical checklists that did not measure its effect on team functioning but only on reported errors or morbidity. Although this review presents all relevant categories to improve team functioning in healthcare organizations, those categories are limited to team literature and are not based on related research fields such as integrated care and network medicine. Another limitation is that we excluded grey literature by only focusing on articles written in English that present empirical data and were published in peer-reviewed journals. Consequently, we might have excluded studies that present negative or non-significant effects of team interventions, and such an exclusion is also known as publication bias. In addition, the combination of the publication bias and the exclusion of grey literature has probably resulted in a main focus on standardized interventions and a limited range of alternative approaches, which does not necessarily reflect practice.

This review shows the major increase in the last decade in the number of studies on how to improve team functioning in healthcare organizations. At the same time, it shows that this research tends to focus around certain interventions, settings, and outcomes. This helped to provide more evidence but also left four major gaps in the current literature. First, less evidence is available about interventions to improve team functioning outside the hospital setting (e.g. primary care, youth care, mental health care, care for disabled people). With the worldwide trend to provide more care at home, this is an important gap. Thereby, team characteristics across healthcare settings vary significantly, which challenges the generalizability [327]. Second, little is known about the long-term effects of the implemented interventions. We call for more research that monitors the effects over a longer period of time and provides insights into factors that influence their sustainability. Third, studies often provide too little information about the context. To truly understand why a team intervention affects performance and to be able to replicate the effect (by researchers and practitioners), detailed information is required related to the implementation process of the intervention and the context. Fourth, the total picture of relevant outcomes is missing. We encourage research that includes less frequently used outcomes such as well-being of professionals and focuses on identifying possible deadly combinations between outcomes.