Reporting Complex Multidisciplinary Healthcare Team Training Interventions: Development and Validation of a Checklist

Background Multidisciplinary healthcare teamwork training is regarded as a means to improve patient safety and quality of care. However, it is increasingly characterized by highly complex and contextualized interventions, challenging scholars and practitioners in evaluating reports of such interventions effectively. Con�icting or weak evidence as to whether these complex interventions ultimately affect patient outcomes suggests the need for more comparability, reproducibility, and transparency in scienti�c reporting. In the absence of an existing reporting guideline, we developed a new checklist for effective reports on the planning, content and implementation design and outcomes of complex teamwork training programs. Methods The checklist was developed through pragmatically consulting existing relevant frameworks for evaluating complex interventional studies as well as related key literature on teamwork training. We followed a stepped approach in data retrieval, applying a con�gurative synthesis. Subsequently, the checklist’s validity and usability were tested based on a literature review of peer-reviewed journal articles describing the implementation of the widely used comprehensive TeamSTEPPS™ curriculum. We used our checklist to evaluate the reporting in the included articles based on speci�c selection criteria. Results The Reporting Complex Multidisciplinary Healthcare Teamwork Training (ReCoMuTe) checklist presented in this paper exists of four distinct categories, each containing components and various explicating elements: Context and preparation; Description; Execution and delivery; Mechanisms of impact. Despite its wide use across various healthcare institutions and settings and the increasing number of publications on its implementation, we found that the reporting in papers on TeamSTEPPS implementation is often inadequate. Conclusion More standardized scienti�c reporting on the design and content of complex teamwork training interventions, their contextualized implementation processes and outcomes, can bolster the understanding of effective intervention design, implementation, and evaluation, and can enable well-informed decision making on investments. The ReCoMuTe checklist provides an instrument for evaluating complex interventions, such as studies on TeamSTEPPS implementation, and can assist authors, reviewers, editors, and others in further contemporary research on multidisciplinary healthcare teamwork.


Background
Despite the increasing use of team training (TT) approaches, there is still a dearth of well-de ned and comprehensive schemes supported by a wide consensus (Salas et al., 2019).Since around 2000, the number of published academic studies on TT in healthcare has increased dramatically (Fig. 1).
Evidence from the last two decades indicates positive associations among elements and effects of TT interventions, even at the level of patient mortality (Buljac-Smaradzic et al., 2010; Weaver et al., 2014;Hughes et al., 2016;Rosen et al., 2018;Salas et al., 2018).However, meta-analytic studies cannot be performed satisfactorily due to the great variability in the 'what,' 'how,' 'when,' and 'who' in the reports on TT, nor can the studied interventions be easily replicated (Cook et al., 2011a;Golub & Fontanarosa, 2015).Notwithstanding the gradual opening up of TT's 'black box,' disclosing various underpinnings of effective and safe teamwork, such as psychological safety and effective team leadership, much of the context of described TT interventions tends to remain unclear.This is because the scholarly publication reports are often inadequate regarding the content, settings, and implementation processes of TT interventions (Salas et al., 2018).Facing the challenge of conveying the particulars of a given TT suitably, we embarked on designing a checklist for such reporting.

Teamwork and Training
Since the publication of 'To Err is Human,' teamwork in healthcare has increasingly gained attention as a serious potential threat to patient safety (Kohn et al., 2000).Over the last 20 years, valuable knowledge and experience has increased in this area, most of which has been derived from aviation and other high-risk industries (Aron & Headrick, 2002;Helmreich et al., 1999;Gross et al., 2019;Rosenbaum, 2019).During this same period, TT has gradually become a standard in quality and safety schemes in healthcare (WHO, 2011; Weaver et al., 2014).
The aim of TT is to mitigate error risks resulting from "natural limitations of human performance and the functioning of complex systems" (Helmreich et al., 1999, p. 29).Trainability of essential healthcare teamwork competencies (i.e., attitudes, behaviors, and cognitions) has been studied extensively (Salas et al., 2005;Suter et al., 2009), and recent meta-analytic reviews provide evidence of the ability of TT to facilitate learning that leads to healthcare professionals' behavioral changes and improved patient and organizational outcomes (Weaver et al., 2014;Hughes et al., 2016;Rosen et al., 2018).Healthcare professionals often participate in various teams, engaging with others under continuously changing circumstances; yet, their individual professional roles mostly remain stable (Andreatta, 2010).Since patient-care pathways occur across many team as well as professional boundaries, effective inter-team and inter-professional communication and collaboration is essential for patient safety.Healthcare settings often involve complex social professional networks with varying backgrounds, educations, viewpoints, work locations, and other characteristics that moderate the effectiveness of their communication and teamwork (Leonard et al., 2004;Marshall et al., 2009).Furthermore, teamwork in healthcare is affected by several characteristics that are distinct from other industries, including low temporal stability and short timespans, a deeply entrenched professional culture (e.g., tacit, complex hierarchical structures), rotating or uctuating leadership, and protocols that impart high human interdependency (Wildman et al., 2012;Hollenbeck et al., 2012;Reay et al., 2016).
TT in healthcare fosters the improvement of interprofessional collaboration and communication as indispensable elements for (continuous improvement of) the quality and safety of care.TT is characterized by a plurality of teamwork-related topics, skill areas, training settings, instruments, and assessment procedures (Salas et al., 2008;Hughes et al., 2016;Gross et al., 2019).Scholars and practitioners have distinctive streams of knowledge and practices that provide extensive resources, such as crew resource management (CRM), Lean/Six Sigma, and simulation (Koning et al., 2006;Salas et al., 2007;Arora et al., 2015;Gross et al., 2019).Similarly, recent evidence-based training methods and validated measures provide practical instruments for addressing a wide variety of elements that comprise such teamwork (Salas et al., 2013;Marlow et al., 2017).One example is the TeamSTEPPS (i.e., Team Strategies and Tools to Enhance Performance and Patient Safety) curriculum, which was developed as a direct response to the aforementioned 'To Err is Human,' and was developed by the United States' Agency for Healthcare Research and Quality (AHRQ) and Department of Defense (Kohn et al., 2000;Clancy & Tornberg, 2007;King et al., 2008;Gross et al., 2019).
Numerous moderating factors can facilitate or impede team performance or TT's implementation and sustainability, particularly contextual factors such as organizational conditions and team climates/cultures (Sexton et al., 2000; Clay-Williams & Braithwaite, 2015; West et al., 2015).TT is increasingly couched in the context of other, contemporaneous quality improvement activities geared at large-scale organizational transformation programs (Robichaud et al., 2012;Gross et al., 2019).Hence, TT becomes, through this intertwinement with quality improvement efforts, subject to a wide array of social behavioral, systemic, and context-related factors (Foy et al., 2011).Healthcare TT comprises, in effect, a complex area in which scholars and practitioners alike acknowledge the consequences of re-institutionalization and re-professionalization processes (Reay et al., 2016).Relatedly, recent meta-analytical work revealed the essence of the learning processes of those participating in TT (Hughes et al., 2016).Optimally, healthcare professionals' learning processes follow tactics such as the 'spacing' of series of learning opportunities in near-real life settings (Keijser, 2019).In contrast to 'one-stop-shop' or 'single bullet' training sessions, more longitudinal type-of programs provide openings for such practice-based learning, often combining TT as part of organization quality improvement programs.
Moreover, TT sessions that alternate between uni-and multi-professional composition provide learning opportunities in which effective feedback can be shared in psychologically safe settings among peers (Fluger & DeNisi, 1998;Hughes et al., 2016).

Teamwork Training and Complexity
TT should also be placed in the context of complexity thinking that mirrors contemporary healthcare's daily reality in practice and academia, as re ected by the two following considerations.First, healthcare teams function within a context characterized by an ever-increasing fragmentation of members, processes, and technologies, embedded in a myriad of professional systems and structures.This fragmentation may indicate the signi cance of TT in providing professionals with adequate knowledge, skills, and attitudes for optimal functioning in their fast-paced and uid network-based interactions and collaborations (Zwarenstein et al., 2009).Second, delivering TT into professionals' relatively chaotic daily (clinical, organizational, and educational) activities also conveys a scholarly challenge: to apply a crosspollination of scienti c views to the eld of TT (Salas et al., 2018).
Practically, these considerations concur with the tendency of TT-based initiatives to progressively develop into complex, longitudinal programs, often focusing on (continuous) organizational change or practice improvement (Dixon-Woods et al., 2011; Romijn et al., 2019).Scienti cally, modern studies on the development of teams (and the multi-team and -unit settings they often occur in), need input from a larger variety of scholarly traditions and viewpoints, also introducing mixed methodologies that can enrich medicine's historical bio-medical approach (Salas et al., 2018).This implies that those reporting TT-based studies should strive for comprehensive (qualitative) descriptions of valuable experiences that are often concealed after (mere quantitative-type) investigations.A coming together of various scholarly streams could provide a better understanding of the efforts, implementations, and wise decisionmaking regarding TT and the consequent allocation of resources.Contemporary TT-based publications often neglect the importance of implementation contexts by conceptualizing healthcare TT as 'just training.'This disregard might be due to a relative dominance of biomedical over more social scienceoriented methodologies in the clinical domain (Greenhalgh, 2012;Pfadenhauer et al., 2017).The present study positions TT as a 'complex TT intervention' (CTTI) (May et al., 2007;Craig et al., 2008).
A complex intervention (CI) refers to implementation or use of various intertwining elements, that are deployed in a complex or 'adaptive' setting or system, and interact by following often unpredictable, non-linear causal processes (Booth et al., 2019;Petticrew et al., 2019).CIs consist "of multiple behavioral, technological, and organizational components" (May et al., 2007, p. 2) and comprises the governing of several interacting components that require distinct expertise and skill from those delivering or receiving the CI.Furthermore, CIs are regularly targeted at several groups, professionals, or organizational levels, and are often tailorable to the context (Craig et al., 2008;Hawe et al., 2009).Settings in which CIs are implemented often represent multi-layered environments comprising various social, cultural, political, procedural, and less-tacit dimensions, jointly structuring the daily reality of the inhabitants (Dixon-Woods et al., 2011).In other words, implementing CIs results in a multifaceted interplay between an intervention's aims, the context in which it is implemented and the various actors that are involved (Petticrew et al., 2019).The causal relationships between a CI and its intended or unexpected outcomes, or the CI's 'active ingredient,' are hard to de ne, which complicates evaluation.A CI co-exists with its context through an interplay whereby the intervention often requires adaptation to the context, including to changes in the context that occur over time (Øvretveit, 2011).
Various multiprofessional care pathways that encompass mutual decision making and re-organization of care are being designed and studied as CIs (Vanhaecht et al., 2010;Möhler et al., 2015;Moore et al., 2015).In fact, a wide range of quality improvement approaches, including plan-do-study-act cycles, are designed as series of interdependent phases and activities that align with the CI principles; they often comprise changes in social structures (Berwick, 1998;Taylor et al., 2014).Change resistance can play a signi cant role in such CI implementation; many healthcare professionals can be downright reluctant to participate in improvement initiatives.This holds particularly for professionals with a conservative, biomedical attitude toward adopting new ways of working, such as physicians (Audet et al., 2005;Wilkinson et al., 2011;Keijser et al., 2019).Nevertheless, CIs in which various professionals participate can promote facets of a CTTI (Deneckere et al., 2012).However, researchers should, advisably, explicitly consider whether or not to view an intervention as complex (Booth et al., 2019).Based on the above de nitions and the extant use of CIs in healthcare, we contend that most contemporary TT programs can, or should, be epitomized as CTTIs.

Our Aim: Improved Scienti c Reporting on CTTIs
The main objective of CI evaluation studies is often to determine whether the goal of the intervention was met.However, how and why an intervention works are equally important questions (Pettigrew et al., 2013;Brewster et al., 2015).Meta-analytical evaluation can contribute signi cantly to identifying the active ingredients of a CI, as well as the factors impeding its implementation in context.An indicative criterium for appropriate reporting of CTTI evaluation studies could be to judge publications on "whether an expert pro cient in [CTTI] research and practice could comprehend and reproduce the intervention based on the information given and the references provided" (Gross, 2019, p. 5).However, Weaver and colleagues reported in their literature review on healthcare TT that "nearly all reviewed studies failed to specify important content" (2010, p. 1756).Failure to describe an intervention's context, design, and outcomes in scienti c publications is also reported in systematic reviews scrutinizing healthcare simulation training, multidisciplinary team and leadership training, and CRM programs (Cook et al., 2011b;Husebo & Akerjordet, 2016;Gross et al., 2019).
Incomplete reporting has been noted in various domains of healthcare, including medical education (Cook et al., 2011a;Glasziou et al., 2014).Yet, "su cient information must be available for the judgement of the intervention's clinical bene ts, for replication, or for adaption of an intervention to different settings or countries" (Möhler et al., 2013, p. 2).Hence, it is imperative to understand the interplay between intervention and context to facilitate well-informed decisions for choosing, resourcing, delivering, and evaluating a CI adequately (Booth et al., 2019).Compared to conventional controlled trials, evaluating CIs in healthcare is particularly di cult due to the unknown factors and interactions between 'components.'Although these components may differ according to the setting, an accurate description of the interventions and their implementation processes as well as the impeding and facilitating factors can contribute to effective reporting (Foy et al., 2011).Currently, much remains unknown about the impact mechanisms and dynamics related to the context and conditions that foster effective application of CTTIs in healthcare (Gross et al., 2019;Salas et al., 2018).Expectedly, multidisciplinary approaches, involving clinicians, scientists, experts, and others at the table, can further this eld through a more detailed study of clinical and organizational behavior, the selection, development, and implementation of TT interventions, and their underlying mechanisms of change that accompany the aimed-for outcomes (or lack thereof) (Foy et al., 2011).Theory-driven evaluations can enhance the generalizability and help build a cumulative understanding of the nature of these changes.We believe that framing TT conceptually as a complex intervention accommodates a wider variety of invokable theories and analytic approaches that could govern further scienti c work on contemporary TT.After consulting the wealth of extant guidelines and frameworks, we did not nd a comprehensive reporting aid, suggesting a need to be addressed.Therefore, we developed a reporting checklist whilst attempting to answer this study's main research question: What elements are crucial for adequate reporting of CTTI implementation?

Methods
Our work was done in two phases: checklist development and face validity testing.

Phase 1: Checklist Development
When developing the checklist, we followed a con gurative synthesis of data (Gough et al., 2009;Mohler et al., 2010; Fig. 2).First, we pragmatically inventoried frameworks, guidelines, standards, and approaches relevant to our study's domain, using online database searches and repositories (e.g., EQUATOR network) (Simera et al., 2010;Bragge et al., 2017).We applied a forward snowballing approach to search for publications using, revising, extending or referring to existing frameworks, and adopted a backward snowballing approach to identify key publications that could provide relevant evidence on the quality of the reporting in the published research (Möhler et al., 2010;Wohlin, 2014).Next, we composed the initial main structure of our envisioned checklist, building on our prior search results.Iteratively, after reviewing retrieved sources and discussions between both researchers, we reached a consensus on the main structure.Then, using the categories framing this initial structure, two authors (WK, RW), individually, identi ed and recorded relevant sub-items describing elements of reporting and implementing CTTIs.The results were rst discussed for relevance and eligibility, and were included after reaching a consensus as elements to the checklist.We continued to search for relevant elements until saturation was reached (i.e., no new aspects were found in the publications).During this development process, we iteratively checked for overlap and redundancy, and also used these feedback loops to ensure clarity in the checklist's wording.During the entire process, the third author was available for (consensus) discussions; both researchers kept notes and convened at various points to compare ndings and discuss their checklist development and to include adequate descriptions of each element.

Phase 2: Face Validity
Testing of the face validity of the 1.0 checklist was based on a literature review of publications reporting the TeamSTEPPS (TS) curriculum.TS is a multifaceted, complex, evidence-based quality improvement intervention, which emphasizes enhancing healthcare team performance through tailorable training designed to enhance team knowledge, skills, and attitudes (Clancy & Tornberg, 2008;Ward et al., 2017).The TS curriculum provides a comprehensive set of connected, but independently utilizable, instruments and approaches that facilitate an understanding and training of key actions for a culture of continuous change based on effective teamwork.TS focuses on such elements as team leadership, mutual performance monitoring, situational awareness, backup behaviors, re exibility, adaptability, team/collective orientation, shared mental models, mutual trust, and closed-loop communication and feedback (Clancy & Tornberg, 2007 (Jones, 2013a); and primary care facilities (Treadwell et al., 2015).Reportedly, TS implementation contributes to improved staff knowledge, skills, and attitudes (Jones et al., 2013b;Sawyer et al., 2013), as well as clinical outcomes (Capella et al., 2010;Spiva et al., 2014).
Our rational for choosing TS as a feasibility 'test case' was twofold.First, it enabled us to review a relatively new intervention, thus deferring a resourceintensive review of a relatively great number of publications featuring a wide variety of approaches.Second, TS is often viewed as an archetypical format for complex multidisciplinary team development interventions (West & Lyubovnikova, 2013;Rosen et al., 2018;Chen et al., 2019;Gross et al., 2019;Hughes et al., 2019), and the variability in its implementation has been suggested to be relevant to understanding its effectiveness (Ward et al., 2017).

Inclusion/exclusion
PubMed, Web of Science, Scopus, Embase, and PsycInfo were consulted to identify reports on TS, using the keywords "TeamSTEPPS" and "Team STEPPS."Searches were ltered by peer-reviewed journal articles or reviews, published in English, in the period starting from TS's launch (2007) until November 2017.Eligible publications described TS implementation in multidisciplinary healthcare teams with the goal of improving teamwork and/or patient safety.Studies reporting on TT in interprofessional educational settings, or only describing interventions that were derived from the original TS curriculum (including those only using a single instrument or tactic, such as huddles or debrie ng) were excluded, as were reports of monodisciplinary TS use (e.g., nurses), or quality improvement programs using TS concepts or describing simulation interventions without accounting for the implementation process.To compare implementation methods, we chose studies conducted in Western countries: in Europe, North America, South America, or Oceania.The included literature reviews on TS implementation underwent snowball sampling, to nd eligible papers, according to the procedures described by Wohlin (2014).However, no new records were found during this process.
After screening and selecting the abstracts and titles based on our inclusion criteria, full-text versions of the articles were retrieved and reviewed.We excluded reports that a) failed to describe implementation phases or skills that were taught, as well as those that b) did not indicate facilitation by a TS 'master trainer' (Clancy & Tornberg, 2008).Since facilitation is imperative to team training (Salas et al., 2002), we regarded the mentioning of a dedicated TS master trainer role as suggestive of well-designed and well-implemented TS programs.A TS master trainer is typi ed as a person who is knowledgeable and capable of guiding TS implementation, including performing site assessments to determine performance gaps, and coordinating, preparing, training, and providing process consultations at the unit, departmental, and organizational levels (AHRQ, 2014).

Data extraction
Full texts of included articles were reviewed in terms of publication and study characteristics (e.g., country, organizational type, setting, study design, and type of data collection).The selected publications were assessed using Atlas TI™ (version 8.0.33.0;Scienti c Software Development GmbH, Berlin, Germany) for text coding based on the identi ed elements in our reporting checklist.To identify codes and explicatory or exemplary quotations, we initially coded a batch of records thematically (n = 12).Then, we compared and discussed the coded quotes and rated the level of reporting of each publication.Subsequently, both researchers separately coded and rated the remaining article in batches, using the selected quotes from the initial step as a reference.The codes were discussed iteratively, and any disagreements were resolved.
Rating classi cation of our checklist's components was done using a 4-point scale: "0" = nothing reported; "1" = minimum reporting: general description, without details; "2" = moderate level of reporting: moderate description or speci ed report on one or more, but not all of the checklist's components, and "3" = optimal reporting: speci ed and detailed reporting.The four scores were also categorized into 'low' or 'high': "0" or "1" representing a "low" and "2" or "3" representing a "high" score.With a similar method all included publications were also coded for impeding and facilitating factors.Finally, four levels of training evaluation were used to assess each record on how the TT effects were measured (Kirkpatrick, 1994;Jones et al., 2013b).

Checklist Development
In the initial developmental phase, existing evaluation frameworks and reporting checklists, focusing on less linear processes than traditional randomized controlled trials (RCTs), offered valuable input.Our "iterative phased approach that harnesses qualitative and quantitative methods" (Campbell et al., 2000, p. 696) helped to develop our envisioned checklist with its purpose to "assess delity and quality of implementation, clarify causal mechanisms and identify contextual factors associated with variation in outcomes" (Craig et al., 2008, p. 34) (Medical Research Council, 2000) and to evaluate "more complex, less advantageous settings" (Glasgow et al., 1999(Glasgow et al., , p. 1322)).Eventually, the existing knowledge was imperative for de ning our nal checklist's 'backbone', comprising four categories: context and preparation; description of intervention; execution and delivery; and mechanisms of impact (Table 1).
Apart from frameworks for the evaluation of worksite-based programs, including those that distinguish facilitating and impeding factors (Fleuren et al., 2004;Wierenga et al., 2016), we sought input from team training interventions targeting healthcare providers (Weaver et al., 2010;Zhu et al., 2015;Marlow et al., 2017).We consulted several frameworks and guidelines describing how to report interventions best, for example the CReDECI 2 guidelines (Criteria for Reporting the Development and Evaluation of Complex Interventions in healthcare), which build on the aforementioned Medical Research Council framework (Campbell et al., 2000;Möhler et al., 2015).During our investigation we also searched for resources focusing on speci c study designs, since our main objective was the reporting of interventions and their implementation.In addition to the aforementioned sources, we reviewed a broad range of reporting guidelines and frameworks (and their more recent extensions) to nd missing components or elements, such as: TIDieR for reporting interventions (Hoffmann et (Möhler et al., 2013;Möhler et al., 2015).These efforts mainly contributed to textual adaptations and alterations, as well as some changes to the composition and ordering of the elements, but did not provide signi cant additional topics for our checklist.
During the nal developmental stages, we formulated components by combining the terminology used in the studied frameworks.Several papers described the proli c growth of terminology in implementation evaluation science and presented guidelines or 'meta-frameworks' for them (Damschroder et al., 2009;Bragge, 2017;Pfadenhauer et al., 2017).Therefore, we attempted to use the de nitions and terminology provided by these guidelines as often as possible in the Reporting Complex Multidisciplinary Healthcare Teamwork Training (ReCoMuTe) checklist.However, more speci c domain-related terminology (i.e., CTTIs) was also included in our nal version (Table 1 Our initial search resulted in 1,100 potential abstracts (Figure 2).After removing duplicates (n = 347), the remaining abstracts (n = 753) were screened independently by two researchers (RW and WK), after which they discussed any disagreements.When criteria were not described fully in the title and abstract, or in cases where the researchers doubted the nature of the quality of the improvement, the article was included for full-text analysis.One article (Motley & Dolansky, 2015) in this literature review could not be retrieved.Snowballing reviews did not result in new records.The full text of the papers (n = 135) were then assessed by two researchers (RW and WK) for eligibility, and any disagreements were discussed.A third reviewer (CW) was available for the latter process but was not needed.The entire literature-selection process produced a total of 27 articles for this review (Table 3).2).Other studies were often conducted in acute or surgical clinical settings (n = 18).Six studies described TS implementation across a hospital system or healthcare region.Additionally, about half of the included studies were conducted in university hospital settings (n = 14), while the remaining studies were performed in mental health, rural hospital, military, or general hospital settings.Two publications did not report the studied TS setting.
Fifteen studies used a pre-post intervention study design.Other study designs used longitudinal (n = 6); mixed methods; cohort study; cluster design experimental study; or cross-sectional comparisons with a pre-post interventional design.One study did not report its research design.Surveys were the most used data collection method (n = 22).Only four studies used qualitative data collection methods, such as interviews, focus groups, or anecdotal evidence.Observational data were used in nine of the 27 studies.
Five studies applied all of Kirkpatrick's four levels of team training assessment as evaluation methods to measure intervention effects (Kirkpatrick, 1994; Table 2).Three articles measured three levels, nine measured two levels, and eight measured one level.Two papers did not report any level of training evaluation.

Quality of reporting
An assessment of the included publications' levels of applying the eleven elements of the ReCoMuTe checklist revealed that a majority of the studies scored low on most of them (Table 3).The most extensively reported were from the checklist's second category, description of intervention, namely the elements 'objectives, content, planning, and participants' (D1) whereby more than half (52%) of the studies had moderate or high scores (Table 3).All the other components were, on average, reported insu ciently, scoring nil or minimum.
Applying the four-level ratings to the E2 ('Learning & Transfer') and E3 ('Faculty') elements was di cult due to the absence or meager reporting details.
Therefore, we decided to rate these elements based on a two-tier scoring system: 'reported' or 'not reported' (Table 3).Consequently, the summed ratings for these two elements are less comparable than the others.
The Thomas and Galla (2013) study was optimal on almost all the checklist's components, followed by four others (Stead et

Final version
During the checklist's development as well as validity evaluation, while assuring no element was left out and the meaning of the descriptions remained intact, we searched for an overlap between the elements and for well-balanced appropriateness, clarity, and conciseness in wording.While validity testing helped to augment the elements' description conciseness, it did not reveal new categories, components, or elements for our checklist, suggesting completeness in our design for assessing reports on the implementation of TS or similar CTTIs.

I -Context & preparation (C1-4)
The rst category of the checklist (see, Table 1) describes the context and preparatory activities before actual deployment of an intervention.The environment or setting in which the intervention is implemented and deployed is referred to as 'context' (Pfadenhauer et al., 2017;Booth et al., 2019).Further, the checklist sets out to address four questions: Did the implementers asses the needs and requirements?(C1) Is there support from leadership as well as from participants?(C2) How is the training intervention contextualized?(C3) How and by whom is the intervention organized?(C4) The local infrastructure, availability of resources, knowledge, and experience can vary signi cantly between healthcare organizations, subsequently affecting implementation potential and execution (Kitson et al., 2008).Moreover, the signi cance of barrier assessments to inform anticipatory sustainability efforts and the involvement of decision-makers and management have been discussed in previous studies (Kemper et

II -Description of intervention (D1-3)
The second category provides an overview of the elements pertaining to the CI characteristics, including its participants, facilitation, and sustainability efforts.
The following questions are typically addressed: What are the intervention's objectives, content, and planning?(D1) Which team(s) and participants are involved and what characterizes them?How is facilitation planned and who will facilitate?(D2) What are the sustainability strategies and activities?(D3).This category also comprises planned and anticipated support outside the training sessions and sustainability activities.
Detailed descriptions of the interventions' objectives, practicalities (e.g., content, planning, resourcing), as well as their theoretical rationale and related strategies and tactics are imperative for scienti c meta-analysis and the replicability of studies

III -Execution and deployment (E1-3)
The third category provides components and elements that facilitate reporting effectively and objectively on what occurred during the intervention implementation, including related educational and pedagogical perspectives pertaining to training activities.We thus address the following questions: What and how much was done-planned and unplanned?What were the deviations from the planned implementation (and why)?(E1) How was it done and received?(E2) Who delivered the intervention?(E3) Our di culties with rating the included studies' reports of elements E2 and E3 might have been due to the fact that we selected studies in which a TS master trainer was present.TS master trainers follow a (minimum of) two-and-a-half-day standardized training session, and their roles and tasks are well described in the TS documentation, which possibly kept the researchers from providing extensive reports on this issue.Additionally, we realized that the checklist's E3 ('Faculty') component overlapped somewhat with the D2 elements ('Facilitation').While comprehensive multidisciplinary team training curricula, such as TS, are on the rise, healthcare organizations are increasingly applying customized 'self-made' approaches, using a variety of trainers, coaches, and other faculty members (Gross et al., 2019).We suggest that, due to the essential role of these individuals in teaching, coaching, and assisting others, the reports should describe their roles in detail, using both elements.

IV -Mechanisms of impact (M1)
The fourth category of the ReCoMuTe checklist provides a set of elements to assist in detailed reporting on impact evaluation.These help to assess possible causal mechanisms and relationships between actual training and facilitatory activities vis-à-vis effects, or lack thereof.Such evaluations of interventions and detailed analyses of either facilitating or impeding mediators, unexpected pathways, and consequences help to answer: Why and how did the change (not) happen the way it happened?
Explaining the intervention's mechanisms of impact can encompass the dynamics that interrelate or overlap across the checklist's categories and their components.Reports that observed or narrated facilitating and impeding factors as part of a process evaluation provide essential information for, for example, further successful replication.Additionally, an initial inventory of determinants serves authors with a practicable instrument for more complex assessments of what happened.Explicating in detail the 'why' of the observed and measured effects, as well as the often unexpected and tacit dynamics imparted by the implementation efforts, requires authors' reports to be based on a convening of viewpoints, including change management and implementation science.The table resulting from our thematic analysis of the 27 included publications on TS implementation provides an exemplary multistudy overview of determinants (Table 4).

Moderators
Facilitating and impeding factors are moderators that intentionally or unintentionally regulate or determine an implementation process and/or its outcomes (Fleuren et al., 2004).Such factors can be described by their characteristics, which can be classi ed into ve categories: ( • Support by leadership [1,2,4,5,6,7,8,12,13] • Change team recognized a need for change [2] • Change ownership was taken up and driven by local change team [8] • Di cult to convince leadership [13] Resources • Plenty of time [10] • Infrastructure was already present [4,27] • Overtime training hours were budgeted [12] • Lack of resources (su cient nances, training time) [4,8,10,23,26] Turnover • Low staff turnover [4,12] • High staff turnover [10] Job position of the implementer • Trainers from different disciplines [6, 8, • Physicians trained by other physicians [13] Collaboration/interaction • Allowing participants to contribute their own ideas/opinions [7,8,14] • Creation of video vignettes and scenarios achieved a high level of staff resonance and buy-in [6] 3. of rewards • Encourage use of learned skills by handing out small aids [5,27] Preparation • Participating in questionnaires about patient safety [10] • Providing information before training [5]

Discussion
In the absence of a preexisting standard, we embarked on developing a checklist for reporting complex interventions, comprising multidisciplinary healthcare TT.Our effort was informed by current frameworks for scienti c reports on the design, development, implementation, and evaluation of CTs in healthcare (Möhler et al., 2013;Möhler et al., 2015;Moore et al., 2015), as well as approaches to evaluating team training implementation (Weaver et al., 2010;Zhu et al., 2015) and worksite health promotion programs (Wierenga et al., 2014).Moreover, by consulting extant key literature on related topics and studying other frameworks and guidelines relevant to the domain or quality improvement and teamwork in healthcare, we synthesized the retrieved data following an iterative method that resulted in the new ReCoMuTe checklist.
The nal checklist does not directly provide structures derived from general theories, such as implementation (Damschroder et al., 2009) or teamwork (e.g., the input-mediator-output framework; Ilgen et al., 2005).Additionally, we do not include distinct levels for training outcome assessment (Kirkpatrick, 1994).
Although only a few of the publications in our review used this approach for assessing the effectiveness of their interventions, we contend that such a comprehensive method for training evaluation should be regarded as a standard procedure (Gross et al., 2019).
During the checklist's development, we did not nd applicable approaches for categorizing team training into sub-types or domains.Despite some attempts, this eld is presumably too young for such distinctions (Husebo & Akerjordet, 2016).However, the need for improved classi cations of teamwork development interventions was increased by recent reports discussing the wide variety of interpretations and jargon, which further impedes optimal reporting in publications (Gross et al., 2019).Therefore, we concur with West and Lyubovnikova (2013) that the 'team' concept must be viewed with some caution, since teams in healthcare can vary widely in, for example, skills, culture, task-focus, and temporal stability or ' uidity'.
However, scienti c research publications primarily focus on aligning with the author guidelines presented by the journals.The ReCoMuTe checklist provides a structured overview of the relevant elements when reporting CTTIs and their implementation.Therefore, while developing our checklist we chose not to include standard items and sections, such as those provided by the IMRaD structure (i.e., introduction, methods, results, and discussion) (Moss & Thompson, 1999).
Furthermore, sections of the publications containing, for example, a study's rationale, ethical considerations, methodological approaches, and statistical results, were viewed as outside our checklist's scope; more appropriate (reporting) frameworks exist for them.We chose to develop a new checklist over alternatives such as amendments to or extensions of existing frameworks (e.g., Golub & Fontanarosa, 2015;Cheng et al., 2016).The rationale for this lies in our primary focus on reporting quality (for research accumulation purposes).In our view, extant frameworks provide excellent guidance on other essential scienti c publication elements.Hence, we would encourage the combined use of ReCoMuTe and other frameworks and guidelines.

Validity
Using the ReCoMuTe checklist on selected reports on the implementation of a standardized TS curriculum a rmed our checklist's applicability and face validity as an overview of relevant elements when reporting.Assessing the 27 included studies with the checklist did not reveal new items; however, reviewing the included studies provided some additional insights regarding their published reports.
While substantiating previous accounts (Husebo & Akerjordet, 2016), we found, not surprisingly, that the majority of studies originated from the U.S., where TS was originally developed.Furthermore, acute and academic settings seem to be popular settings for TS implementation.Possibly, this is due to the natural discourse of healthcare team training science and practice.In the last two decades, an important knowledge base has been sprouting from high-risk industries (e.g., aviation) to similarly risk-prone healthcare settings, in particular ORs and ERs (Gross et al., 2019; Table 2).Aptly, team training practices in other industries show an evolution that comprises several generations of experimenting and developing before reaching a wide consensus on best practices (Helmreich et al., 1999).Moreover, the initial TS version (v1.0) was primarily aimed at such settings.TS was designed to provide a tailored-to-context set of approaches and building blocks for various settings and, recently, AHRQ has provided new versions adapted for non-OR settings (e.g., long-term care or o cebased clinics).While our results corroborate previous references to TS interventions as being primarily used in university hospitals (Ward et al., 2015), recent research on adverse incidents suggests that TS and similar interventions are relevant to regional hospitals as well (Langelaan et al., 2017).
Rating the sample of publications in our review using the ReCoMuTe checklist results in a low reporting score.Apparently, despite TS's highly standardized

Medical engagement and leadership
The importance of physician engagement in TS was noted in several studies in our sample.While the multidisciplinary character of TS is reported as an important determinant for implementation success in six studies (Mayer et  It has been acknowledged for some time that physician competence in effective engagement in quality and safety improvement requires a more integrated educational system (Aron & Headrick, 2002).Also, a "lack of training focus to address hierarchical differences and incivility" has been mentioned as an impeding factor for TS implementation (Clapper & Ng, 2013, p. 287).Soklaridis and colleagues (2007) found that a barrier to team collaboration was limited formalized education on interprofessional collaboration for physicians, and even fewer training opportunities for other healthcare professionals.As several studies reported the lack of physician involvement as a possible reason for underperforming implementation success (Jones et al., 2013a;Sheppard et al., 2013;Wong et al., 2016), it seems important to address this issue.
Clapper and Ng recommend a preparatory adjustment of existing team culture and improving physicians' leadership skills when implementing TS (Clapper and Ng, 2013).Interestingly, the recently emerged concept of medical leadership has not only gained signi cant attention inside the medical profession and beyond, but also directs physicians' education and training particularly toward embracing behaviors and non-technical (leadership) skills in teamwork (Dath et al., 2015;Keijser, 2019).Healthcare teamwork and medical leadership have been suggested to be inextricably connected as bilaterally scaffolding principles (Husebo & Akerjorder, 2016; Salas et al., 2018).It is well-reported that physicians and their historically sovereign professional status have a signi cant impact on psychological safety and, consequentially, on engagement in quality improvement work, such as CTTIs (Nemhard & Edmondson, 2006).Contrastingly, new insights into medical education suggest more collective practice-based learning, such as in CTTIs, as essential strategies in educating physicians (Cruess et al., 2018).However, because of insu cient reporting in publications describing multidisciplinary teamwork training, including physicians, it often remains unclear if and how the challenges regarding physicians' involvement and their leadership can be addressed.

Practical Implications
The ReCoMuTe checklist could serve to help upgrade the comprehensiveness of reporting intervention studies.These include investigations of the e cacy of TT as an intervention for team performance improvement and studies using TT as an investigative approach to scrutinize speci c topics, such as quality improvement, medical leadership development, interprofessional learning, or organizational transformation.Establishing and assessing research evidence supporting CTTIs might not always be the primary objective of eld implementors and their teams (Ward et al., 2017).Yet, we hope our checklist will also provide a practical instrument for these groups.
The ReCoMuTe checklist importantly aims to assist authors of papers involving TT and to also be instrumental for editors and reviewers when assessing the suitability of submitted work to enhance their readership's access to detailed accounts of experiences and learned lessons.Although we set out to facilitate the work of designers, implementers, and evaluators of complex, multidisciplinary TT interventions, we believe the current checklist also serves as a guiding instrument for less complex types of multidisciplinary TT programs.Arguably, within the longitudinal perspective of a few weeks to several years (i.e., typical range for implementing team training programs), any implementation effort comprising multidisciplinary and organizational dynamics can spiral into complexity at some point (Husebo & Akerjordet, 2016).Moreover, context as well as interventions tend to change during the phases of implementation and sustainability activities, which is typically the period that is covered in scienti c reports.Hence, there is a need to describe such evolutions and their related dynamics (e.g., unexpected changes impacting participants' engagement), which can be aided by the ReCoMuTe checklist.
Ultimately, well-constructed publications can feed into meta-analyses, substantiating the well-informed design and delivery of CTTIs.However, RCTs have only ).We hope that our checklist will help to motivate those engaging in meta-analytic endeavors of meticulous comparative studies and to embolden them in their demanding review processes to be rigorous in judging a report's eligibility, and maybe even disregard those that have 'stripped away' information on essential contextual matters (Booth et al., 2019;Greenhalgh, 2012).
Due to journals' wordcount limits, authors increasingly report details through an extra publication, (e.g., a study protocol) or a website (Möhler et al., 2015).
Furthermore, since almost no clinicians and only a few researchers have time to personally contact authors requesting additional information, editors should encourage their submitting authors to allocate locations beyond their primary publication (e.g., online supplementary documents) (Hoffmann et al., 2014).
We emphasize, though, that the ReCoMuTe checklist is not a 'recipe,' nor a comprehensive guideline.It is not a complete list of 'must-haves' to direct the design or implementation of CTTIs.Moreover, the scienti c relevance of some of the elements to a CTTI's effectiveness is neither established nor disputed (e.g., team composition, work environment, training strategy) (Hughes et al., 2016).Instead, it is a starting point for those endeavoring to research CTTIs and their implementation.The checklist comprises a structured overview of known elements, and only future research can reveal the causal relationship for most of them with impact mechanisms and effectiveness.Therefore, some caution in using the new ReCoMuTe checklist in further work is warranted, for which we provide the following considerations.

Limitations
We have several considerations regarding the development and testing of the ReCoMuTe checklist.First, we used a non-exhaustive and pragmatic approach, which was not based on extensive consensus building (e.g., expert group consultation); it addressed our own need for adequate reporting, suitable for the purpose of knowledge accumulation, given the abundance of existing reporting standards ( Above all, we hope that we have stepped up the discipline used by authors and editors of journals for all the reporting criteria voiced in this paper. Second, using one distinct type of CTTI (i.e., TeamSTEPPS) for face validation may have instilled a bias.Conceivably, using a wider spectrum of CTTI approaches for validity testing could have revealed elements that were not part of this study.Although TS's unequivocal tailorable characteristics offer signi cant variability in its application in practice, using clear reporting criteria (i.e., the ReCoMuTe checklist) might help prevent publications on TS, or similar programs, from becoming a heterogeneous 'hodge-podge,' and thus from reducing their potential (Petticrew et al., 2013).
Third, the present study's focus on intervention, implementation, and process evaluation, included elements of the 'partial system perspective.'This enabled the capturing of information on the interplay between environmental and organizational aspects, such as processes, cultures, and values, and the interventions themselves.However, meso and macro system perspectives (e.g., economic, social, and political factors; regulatory matters; policies) must also be considered when consolidating results from various sources to produce transferable ndings (Booth et al., 2019).Healthcare systems' reform and transformation endeavors should, ultimately, be aligned with the level of the teams and their performance, since it is at this level that healthcare is created and delivered (Keijser & Martin, 2019).
We believe these considerations will encourage eventual further testing of the ReCoMuTe checklist.The efforts should preferably include other settings than academic and acute types, since the signi cance of improving patient safety and healthcare quality certainly extends beyond these settings as well.Further work could entail publication of an 'explanation and elaboration paper,' providing exemplary research questions, discussing methodologies, advising on data collection, and discussing exemplary publications.Parallel to this, based on our current work, it is our opinion that the TT domain needs a more established operational de nition of teamwork and training approaches, including a glossary of concepts, terminology, and instruments that would aid future standardization and subsequent reporting.
In sum, we believe the status quo of TT reporting is in its early stages of growing toward maturity.Suggested future work in this area should exceed the scope and depth of the present study and must contribute to a more comprehensive type of reporting guideline or standard.Until then, the (current version of the) ReCoMuTe checklist provides a structured set of elements that has been proven to be useful for analyzing reports on TS interventions, so that further studies and reports on those interventions could, eventually, save more lives.

Conclusions
This paper originates from the authors' impression of inadequate reporting of CTTIs, and an increasing call for standards to enhance completeness, We present in this paper the new ReCoMuTe checklist which can be used to step up the quality and completeness of reporting on multidisciplinary TT in complex intervention formats to enable, ultimately, better replication of these interventions, fostering a better understanding of the necessary (professional) behaviors and other (e.g., contextual) factors involved.The ReCoMuTe checklist can be used by: authors to structure their reports on the interventions; editors and reviewers to evaluate the completeness of the offered descriptions; and readers or practitioners interested in an effective design and implementation of TT.The checklist may also support the planning of purposeful research agendas, while also contributing to wise spending of resources in the current generation of varied CTTIs and related research activities (Hoffmann, 2013;Möhler et al., 2013).We tested the ReCoMuTe checklist using publications reporting on the implementation of TS.
We found that TS as described in studies we included has been primarily implemented in high-demand settings in U.S.-based university hospitals, and that training effects were mostly researched using pre-post training and longitudinal research designs.Also, these publications generally reported the importance of multidisciplinary participation and physician engagement.Not surprisingly, most of the TS publications included in our systematic analysis did not report a majority of the elements of the ReCoMuTe checklist.
Our study a rms that, given the extant reporting levels, more adequate reporting is needed in this area to foster more practical use and replication.We also note a 'tailored-standardized paradox'.The use of a comprehensive type of CTTI, such as TS, which is explicitly designed to be implemented in a tailored manner, based on local, or contextual, needs and demands, imparts a variability in implementation modes, which makes comparability between single implementation reports di cult.Considering these non-linear characteristics of implementing CCTI programs in healthcare, further work on standardization, such as for reporting, can help scholars and practitioners nd a balance between situational adaptability and standardization (Chen et al., 2019).We contend that, like the evolutionary development of healthcare quality improvement science and TT in other industries, the emerging science of complex and multidisciplinary TT interventions is gradually becoming 'institutionalized' but it is not su ciently 'professionalized' yet (Audet et al., 2005).However, TT and related (e.g., leadership) developments are becoming integral elements of audits, professionalization, and educational schemes in healthcare, which is a similar process to how TT evolved in the eld of aviation (EASA, 2018; Rosenbaum, 2019).
Our study also stresses physicians' crucial role and position in CTTIs, in line with their relevance to effective system reform and safety culture (Nembhard & Edmondson, 2006;Porter & Tesiberg, 2007).Physician involvement, or lack thereof, in CTTIs, has a signi cant in uence on the failure or success of implementation efforts.This indicates a need to ensure effective investments in assisting the medical profession and its members, for instance through enhancing their leadership competency development and a reciprocal adaptability between various professional identities ( Stepping back from the present study to a more macro-level perspective, we suggest decision makers to consider a strategic, synchronous deployment of CTTIs across their organizations' departments and teams.Since healthcare professionals tend to work within multiple clinical teams, more strategically

Figure 3 .
Figure 3. Flowchart of the literature-selection process entered the eld of healthcare team training recently (Panella et al., 2012; Husebo & Akerjordet, 2016; Strasser et al., 2018; Dinius et al., 2019).Since RCTs represent the 'Gold standard' in evidence-based practice, this possibly heralds a certain timeliness for the ReCoMuTe checklist (Craig et al., 2008; Bhattacharyya et al., 2009 reproducibility, and comparability of TT(Mohler et al., 2015;Gross et al., 2019).Incomplete descriptions of interventions in publications on CTTI hinders an effective use of resources invested in healthcare TT and the evaluation of its implementation.As expected, we found in this systematic literature-review study that the current quality of reporting CTTIs is remarkably poor.Hence, better-quality reporting on the 'what, how, who, when, and why' of CTTIs is urgently needed.Due to this inadequate reporting, we cannot point out causal relationships and dynamics relating to the failure or success of TT implementation.The complexity of the phenomenon under study, as well as the multidisciplinary perspectives taken, fuel the need for adequate evaluations or reports on TT interventions.

Table 1 .
). Categories, components and elements of the 'Reporting Complex Multidisciplinary Healthcare Teamwork Training' (ReCoMuTe) checklist Assessment informing (tailored) design, deployment and evaluation § Clear and aligned (e.g.organizational) strategies and justi cations § Speci cation of local problems/challenges C2 Engagement & endorsement § Motivation, engagement and readiness of the organization and participants (activities and assessments related to, for example: sense-of-urgency; sense-making; shared understanding; coalition formation; resistance reduction; incentives) § Endorsement and support (e.g., management; executives; medical staff) § Clear communication and information (e.g., on program objectives) * § Participants (e.g., characteristics; selection; recruitment; enrolment); team(s) (e.g., authority differentiation; temporal stability; tasks and skills differentiation) § Content and materials* § Detailed planning activities* (e.g., timing; duration; location; frequency; timeline visualization) § Required resources (e.g., time; nances; materials; location)* § Anticipated causal relation objectives vis-à-vis content and activities* § Metrics on progress and outcomes monitoring* § Theories and evidence (e.g., effectiveness) underpinning the rationale for the design, deployment and evaluation D2 Facilitation § Planned facilitation strategies, tactics and processes (during and outside planned sessions) § Main facilitator's (e.g., master trainer) characteristics (e.g., background; position; selection; gender) § Additional facilitation (e.g., individual leadership coaching) § Informal facilitation (e.g., support to and from champions) § Resistance/'change fatigue' reduction § Implementation communication (e.g.progress; milestones; updates; materials, such as: pens, badges, (online) nudges) D3 Sustainability § Sustainment strategies and activity (e.g., integration in organizational quality and educational cycles; periodical evaluation) § Post-intervention (e.g., refresher or new-comers' training) analysis § Methodology for (process) evaluation (including theoretical rationale) § Descriptions and/or selective narratives re ecting topics relating change management to organization behavioral and other social structures (What happened?)§ Impact mechanisms (e.g., anticipated versus unexpected) § Mediating (expected and unexpected) factors, pathways and consequences § Strengths and limitations of intervention's components § Lessons learned § Possible (causal) explanations of activities' impact on outcomes § Description of the control group and conditions (if applicable) (*at various (organizational) levels) Face Validity Testing

Table 2 .
Study characteristics of the included (n=27) studies reporting on I implementation, including their application of Kirkpatrick's levels of training

Table 3 .
Scoring of the publications on TeamSTEPPS implementation (n=27) with the ReCoMuTe checklist components, including summated scores (Lee et al., 2017)2;Marshall & Manus, 2007)ance of describing sustainment, or maintenance activities, since team training effects at various levels (i.e., reactions, learning, transfer, and outcome) often diminish over time(Arthur et al., 1998;Wierenga et al., 2014).Reports on TS implementation note a decline in training effects within six months to a year(Forse et al., 2011;Thomas & Galla, 2013).Strategies such as regularly planned competency refresher training can help mitigate such declines(Sonesh et al., 2015).Further, concurrent workplace-based coaching and mentoring of staff, as well as post-training sustainability activities, can be applied(Morey et al., 2002;Marshall & Manus, 2007).Strategies and activities focusing on sustaining the long-term effects of healthcare team training implementation have been suggested to be an underexplored part of implementation and research(Lee et al., 2017).Since "team training is not a one-day or single-session event" (Gillespie et al., 2010, p. 655), the sustainment phase of TS and similar interventions is imperative (AHRQ, 2014).However, only 15 of the 27 articles reviewed here reported such sustainment activities.
(Kitson et al., 2008)Salas et al., 2002)(s) characteristics are essential, such as task types, the differentiation of authority across the team(s), and their stability over time(Hollenbeck et al., 2012; Wildman et al., 2012).Thus, effective reporting comprises detailed descriptions of strategies, tactics, and processes used by an individual (i.e., facilitator) to help others improve their knowledge, skills, or attitudes and thereby improve the intervention's likelihood of success, including speci cations of the facilitator's quali cations and subject matter expertise(Kitson et al., 1998;Salas et al., 2002).Furthermore, contextual information can be relevant since, for example, small facilities can lack implementation strength, which has to be addressed and anticipated (e.g., with more appropriate facilitation)(Kitson et al., 2008).

Table 4
. reported as (a) facilitating or (b) impeding TeamSTEPPS implementation in the 27 publications (Clancy & Tornberg, 2007))ensus on the 'what-and-how' when reporting evaluations of TS or similar programs.Effective reporting might not be within the scope of those who report (Mohler et al., 2010;Golub & Fontanarosa, 2015).Arguably, this is due to the relative novelty and complexity of the eld for these types of healthcare team training interventions and the young age of one of the eld's pioneering programs (i.e., TS curriculum)(Clancy & Tornberg, 2007).Promisingly, al., 2011; Brodsky et al., 2013; Thomas & Gala, 2013; Spiva et al., 2014; Amaya-Anyas et al., 2015; Sonesh et al., 2015; Lisbon et al., 2016), it is also explicitly mentioned that physician participation had a similar impact (Jones et al., 2013a; Sheppard et al., 2013; Thomas & Gala, 2013; Wong et al., 2016).Di culties in scheduling around on-call duties resulted in poor physician participation and a need to adjust programs into abbreviated, physician-only versions (Forse et al., 2011; Sheppard et al., 2013).Interestingly, one study mentioned the bene cial effects of peer-to-peer training among physicians (Thomas & Galla, 2013).
Simera et al., 2010;Pinnock et al., 2015;Bragge et al., 2017).An imaginable selection bias, instilled by a possibly incomplete list of sources, was mitigated by our stepped development of the iteratively adapted consecutive versions of the checklHoffmann et al., 2014) extensive scholarly and practical experience in TT, organizational change, medical leadership, and healthcare transformation within our group, as well as including an initial validity test, also mitigated bias.Follow-up work could comprise expert surveys and consensus meetings, including Delphic methods, which could build on the checklist presented by the present study(Möhler et al., 2010b;Hoffmann et al., 2014).Since the TT eld is certainly expanding rapidly and in various dimensions, the ReCoMuTe checklist will need, like most reporting guidelines, revision or re nement.