Effect of mental rehearsal on team performance and non‐technical skills in surgical teams: systematic review

Background Simulation‐based training in medical education has become a common method to develop both technical and non‐technical skills in teams. Mental rehearsal (MR) is the cognitive act of simulating a task in our heads to pre‐experience tasks imaginatively. It has been used widely to improve individual and collective performance in fields outside healthcare, and offers potential for more efficient training in time‐pressured surgical and medical team contexts. This study aimed to review the available literature to determine the impact of MR on team performance and non‐technical skills in healthcare. Methods MEDLINE, Embase, British Educational Index, CINAHL, Web of Science, PsycInfo and Cochrane databases were searched for the period 1994–2018. The primary outcome measure was improvement in team performance and non‐technical skills. Study quality of RCTs was assessed using the Medical Education Research Quality Instrument. The reported impacts of MR in all included studies were mapped on to the Kirkpatrick framework for evaluation of educational interventions. Results Eight studies with 268 participants were identified that met the inclusion criteria, of which there were six randomized trials, one prospective pragmatic trial and one qualitative study. Three studies found MR to be effective in improving team non‐technical skills. MR practices were varied and often poorly defined. MR benefited team non‐technical skills when it was specifically designed to do so, but was not an automatic consequence of technical MR alone. The majority of studies demonstrated benefits of MR for technical performance, but only three showed positive impacts on teamwork. Overall the studies were of low quality and lacked sufficient discriminatory focus to examine impacts on teamwork dynamics. Conclusion MR can improve technical performance, but the benefits on non‐technical skills are less clear. Future research should look at longitudinal mixed‐method evaluation designs and focus on real clinical teams.


Introduction
Successful surgical procedures are the product of a combination of sustained technical skills, effective non-technical skills (NTS) and ongoing professional education. However, this exacting balance is regularly challenged by the increasing complexity of surgical procedures, the dynamic nature of surgical team composition, changing physical and material contexts, and diverse patient safety concerns. Surgical team members often need to coordinate rapidly with other professionals with whom they may never have worked, to undertake procedures they may never have performed.
In these challenging conditions, there is a need to determine how to provide effective educational support for surgical teams, at all levels of experience. There has been a strong interest in simulation-based training for technical and non-technical skill acquisition. Surgical simulators can provide substantially more practice than traditional models of surgical education, and studies have reported positive transfer of learning to practice. However, technology-based simulation utilizing virtual reality and mannequins is expensive and resource-intensive. Mental rehearsal (MR) has been proposed as an adjunct to existing educational strategies, and has been reported to improve performance in the operating theatre [1][2][3][4][5] . MR is described as the cognitive rehearsal of a task in the absence of overt physical movement 6,7 . More broadly, it draws on our remarkable capacity to combine people, artifacts and actions in our heads in very novel ways and to pre-experience events imaginatively. This process aligns with a very down-to-earth notion of consciousness, namely the setting up and planning of future goals. MR is already well established in sports psychology, where it is widely acknowledged to improve both individual and team performance 8,9 . These findings have been attributed to a complex interplay between cognitive, motivational and motor skill functions 10 , with neuroimaging evidence indicating overlapping cortical and subcortical networks 11,12 . The suggestion that MR may contribute to improved confidence and motivation in group performance is significant. It has been demonstrated, for example, that self-efficacy strongly predicts and moderates individual perceptions of team efficacy 8,9,13,14 .
Given these relationships, it is likely that certain individual imagery and MR functions will also predict collective efficacy. The implication for surgical teams is that MR techniques are able to help both individuals and teams reach higher levels of 'shared envisioning' of a task, or foresight, and hence facilitate a more effective and safer performance [15][16][17][18][19][20][21] . This resonates with the idea of shared 'mental models' and 'situational awareness' in patient safety science. It also finds echoes in tools such as the WHO surgical checklist, which is designed to enhance risk awareness and team cohesion. The assumption is that, over and beyond improving individual technical skills, MR may improve group dynamics too. These encompass the whole panoply of well known NTS used in healthcare team training 22 .
Typically, NTS include situational awareness, communication, decision-making, teamwork, leadership, and the management of stress and fatigue. There is substantial evidence that poor teamwork is a key contributor to preventable errors in healthcare 23,24 .
A number of recent studies have examined the impact of MR on the acquisition of surgical motor coordination skills, and some focused on non-technical aspects of performance such as individual stress reduction and coping strategies. However, the use of MR to develop team skills has not been a primary focus. The aim of this study was perform a systematic review of evidence for the impact of MR on team performance and NTS in surgical teams.

Methods
For the purposes of this review, MR was construed as an inclusive concept, including mental practice, mental imagery and mental simulation, that is associated with strategies to enhance learning and performance.
PRISMA guidelines 25 for reporting of evidence in systematic reviews were applied. Studies that examined the impact of a broad range of MR interventions in surgery on performance of NTS were identified.

Locating systematic reviews
Bibliographic databases (PubMed, British Educational Index, Educational Resources Information Center (ERIC), Cumulative Index to Nursing and Allied Health Literature (CINAHL), PsycINFO, Cochrane Library) and the internet (Google Scholar, Web of Science) were searched. Search strings comprising a variety of synonyms for MR ('mental imagery', 'mental practice', 'mental time travel') were combined using Boolean operators (OR,AND) with 'surgical teams' or 'medical teams', before further combinations with 'non-technical skills' and associated terms 'teamwork' and 'team performance'. The literature search was conducted between November 2017 and September 2018.

Selection criteria
The following selection criteria were applied to titles and abstracts of provisionally identified papers to identify relevant reviews: studies investigating the impact of MR on physicians or surgeons or members of their respective teams; studies that examined impacts on NTS, including stress reduction, communication, teamwork and sense of self-efficacy or confidence; studies in English, published between 1994 and 2018; and full study available. A study had to meet all four criteria to be included.
At the second stage of screening (after reading full papers) these selection criteria were reapplied. Reference lists of included studies were scrutinized for additional papers.
The initial search for literature was conducted by two authors independently. Results were then compared and exchanged, and the remaining titles, abstracts and full texts were reviewed for eligibility and relevant information was extracted.

Data extraction, analysis and synthesis
Studies that met the selection criteria were coded for relevant details about contexts, methods, and results or outcomes. A narrative synthesis of each study was completed.  This narrative involved summarizing and combining the descriptive and contextual outcome information from the included papers. Two instruments were used to judge the quality of the papers reviewed. The Medical Education Research Study Quality Instrument (MERSQI) 26 was used to measure the methodological quality of the RCTs. The MERSQI consists of six domains (study design, sampling, type of data, validity of evaluation instrument, data analysis, outcomes), each of which carries a maximum score of 3. Five domains have a minimum score of 1, with a possible total score of 5-18. The Kirkpatrick Impacts on Learning Outcomes Framework was also used to evaluate the impact of the educational interventions 27 (Table 1).

Results
A total of 574 articles were identified, resulting in 404 abstracts being reviewed after duplicates were removed. From the 404 abstracts, 11 studies were included for full-text review. Three further studies were excluded as the full text was not available after contacting the author for one, and two others focused on mental models rather than MR (Table S1, supporting information). Thus eight studies [28][29][30][31][32][33][34][35] were included in the review (Fig. 1).

Demographics
Of the included studies, six were prospective blinded RCTs [28][29][30][31][32][33] ; the other two studies were a qualitative semistructured interview study 34 and a prospective two-part 'real world' study 35 . A total of 268 participants were enrolled, with 134 being randomized to the MR group. One paper 34 did not state the number of participants. Participants' medical experience varied from medical student to consultant, the most common being a postgraduate doctor at a mid-point through a specialty training programme. Of the studies where MR was used as an intervention, five 29-33 utilized a simulated setting with the other two 28,35 using real surgery. Demographic characteristics of each study can be found in Table S2 (supporting information).
The primary outcome for four trials [28][29][30]35 was a measure of teamwork or team performance. In three studies [31][32][33] , the primary focus was another aspect of medical performance (such as technical skills or stress), with teamwork as a secondary outcome. The final trial 34 used thematic analysis of consultant interviews, focusing on preoperative preparation. Teamwork was measured using validated NTS observational tools, time taken to perform procedures, and 'errors' (incorporating delay and danger). Four studies 29-32 also used the Mental Imagery Questionnaire, a tool designed to quantify an individual's ability to perform a mental practice task.

Techniques of mental rehearsal
There was a wide variety in the methods of application of MR. For each study, there were differences in the focus of the MR, the delivery of MR, the length of time for which it was performed, and the aids used to facilitate delivery ( Table 2).
Two studies 30,35 intervened specifically to promote collective MR performed as a team. One 30 encouraged pairs to discuss an upcoming trauma simulation, and the other 35 paused a surgical procedure before a critical part of surgery and the lead consultant performed a verbal run-through of the upcoming steps. In the remaining six studies, participants were encouraged to visualize on their own, rehearsing their own upcoming performance and actions. MR was oriented to a physical task and gave 'visual, kinaesthetic and cognitive' cues to perform a successful operation. Examples of these included 'grasping the bowel only where I can see it' (visual) and 'I feel where the bowel wants to go' (kinaesthetic).
In three studies [29][30][31] , the primary focus of MR was teamwork and NTS. Participants were asked, for example, to imagine how they would 'interact with team members' and who would 'perform which task'. In one study 31 , MR was part of a larger 'stress management training'. A variety of aids were used to facilitate MR, including written MR scripts and videos. Three studies 28,31,32 also employed the use of an MR 'trainer', who was trained specifically in the delivery of MR.
The amount of time performing MR for participants varied greatly between studies, from 5 min to more than 90 days of repeated individual sessions. Three 29,30,35 of the prospective trials only used one session lasting 20 min or less, but in some studies 28,32 it was unclear how much time was spent.

Effect of mental rehearsal
Of the seven prospective trials, three 30,33,35 displayed significantly improved teamwork in the MR group, with no significant difference found in the other four 28,29,31,32 ( Table 3). The single qualitative study by Ibrahim and colleagues 34 identified MR as a recurring theme and key part of every consultant's preoperative preparation. However, this study described acquired expertise by senior surgeons rather than educational impact.
There was significant heterogeneity in terms of methodology and assessment, making direct comparison between studies difficult. The primary outcome and tasks assessed were varied, and there was no shared measurement of teamwork across the studies.

Theoretical perspectives employed
Four papers [28][29][30]32 referred explicitly to theory to provide explanatory frameworks for how MR enhances learning and performance. A further study 34 described MR as a key part in a larger theoretical educational construct. The other three studies 31,33,35 made no reference to underlying concepts ( Table 2).
'Dual-coding theory' 36 , which is a psychological theory of cognition, was referred to in two studies 29,30 . Dual-coding theory postulates there are two methods to represent information, verbal association and visual imagery, and that, when combined, these reinforce learning. The hypothesis is that MR within teams may enable participants to share their imagery and associations, building an improved and detailed mental model of the procedure they are about to perform. In this way, cognitive load is reduced, increasing the amount of available working cognition to focus on more complex problem-solving.
Two further studies 28,32 describe 'neuroplasticity' 38 and the proposition that learning is reinforced when the brain activates neuronal pathways to simulate or 'rehearse' physical actions. This assumption is supported by physiological evidence that neuroplastic and synaptic changes occur during MR, imitating the changes that occur when physically performing the task 37 .
For Ibrahim et al. 34 , the impact of MR is best described through the lens of Actor Network Theory (ANT) 39 . ANT provides a sociological perspective to explain how, within a given situation, people, ideas, objects and processes interact with one another on an equal basis to produce certain outcomes. Rather than single out the impact of MR for separate analysis, the authors describe how a surgeon's use of MR is inherently linked to preoperative preparations and fits into a complex interactive 'web' of tools, policies and agents that mediate and shape individual and collective learning and performance.

Level 4c
Lorello et al. Values indicate the number of outcomes of that type and the level reported in the study. * A qualitative study with self-reported outcomes from expert informants and no independent analysis. 12 or above correlates with 'quality' research, publication and funding 40,41 . Only three 28,29,32 of the included studies scored at or above this threshold.

Reported impact on learning outcomes
Reported outcomes in each study were mapped on to the Kirkpatrick framework (Table 5). Seven studies [28][29][30][31][32][33]35 provided details of the impact of MR activities on outcomes. One qualitative study 34 was included for comparison and did not refer to the impact of an educational intervention, rather to the impact of acquired expertise on preoperative routines. The majority of outcomes were at level 1, 2a and 2b. Two studies 34,35 reported outcomes at level 3 and the qualitative study 34 at levels 4a to 4c. How participants react to a particular educational strategy is often the first layer of evaluation, but this was not reported in three studies 29,30,32 . Louridas and co-workers 31 reported that eight of ten respondents said they would transfer MR to their clinical practice but, like Wetzel et al. 33 , provided few details of how reactions were collected or interpreted. None of the RCTs reported outcomes in terms of skills, attitudes or behaviours that were transferred beyond the experimentation.
Of the two studies 34,35 reporting outcomes involving transfer to practice, only one 34 treated the adoption of MR itself as a key practice outcome. Patel and colleagues 35 described how surgical errors during real surgery were reduced in combined open/endovascular arterial procedures following MR. It also was the only study to acknowledge the number and complexity of other interacting and interdependent components (such as WHO checklists, local cultures) as factors contributing to the bolstering of group foresight through MR and planning, but did not report any follow-up investigation of organizational changes to practice that incorporated MR. The qualitative analysis of expert surgeons by Ibrahim et al. 34 described a wide variety of self-reported outcomes deriving from individual, collective and embedded organizational practices. These were subsumed in routine practices of mental imagery, collective planning, written scripts, prebriefings, individual and group reflexivity.

Discussion
This systematic review analysed RCTs and prospective trials assessing the impact of MR on teamwork and NTS in surgical education and medical team training. Of the eight studies included, three reported positive impacts of MR on teamwork and NTS. Five studies reported improved technical performance after MR, but no significant effects on teamwork, and one study linked MR to improved coping strategies.
In surgical education, with few exceptions, MR before performing a surgical task has typically been designed for individual technical performance. However, surgical tasks are often undertaken by multidisciplinary teams, and cognitive and affective states that emerge as a result of team member interactions can affect overall performance 42 . The potential role of MR in priming NTS to improve team performance was acknowledged in all included studies, but was not theorized sufficiently to produce adequate tools to prime collective performance.
The single components of each learning activity are important. How educators 'constructively align' learning goals to activities designed to achieve them, and criteria to assess them, is a key component of instructional design 43 . Rao et al. 4 performed a large meta-analysis of MR and concluded that effective use was characterized by 'being directed toward the task'. However, of the eight included studies, only two 29,30 specifically focused the MR activities on team interactions and NTS. In the remaining studies, the MR script was based solely around the physical actions involved in a surgical procedure. Without purposeful priming it is unlikely this would stimulate participants to imagine collectively effective NTS interactions with their colleagues. Numerous team NTS tools are available and could potentially be adapted to aid collective MR 44,45 .
The MERQI analysis revealed a number of methodological weaknesses in the RCT designs, such as small sample sizes, high risks of cross-contamination between control and intervention groups, multiple confounding interventions with MR, and large time lapses between intervention and testing, making it difficult to determine causal mechanisms of change. Geoffrion and colleagues 28 observed that, with an innate human attribute such as MR, the control group may have been performing it 'unknowingly' anyway, and inadvertently subverting the design intention of the study -one of the limits of an RCT in this educational setting 46 .
The first level of analysis in the Kirkpatrick framework concerns the participants' views. Nevertheless, three studies 29,30,32 neglected to report people's reactions to the use of MR. In the study by Lorello and co-workers 30 , the primary outcome was the 'acquisition' of NTS behaviours, but normal team composition was not reproduced in the task. The use of actors in studies to 'simulate' the team is contrived and makes generalizations regarding the impact of MR on team NTS questionable.
Few educational impacts were reported beyond the immediate experimentation period, which in most cases was very short. Seven studies reported positive changes in either psychomotor or social skills (level 2) as an immediate result of the experimentation with MR; however, only two studies reported transfer to practice of beyond level 2b.
Only one study made reference to educational frameworks to evaluate more detailed and longitudinal impacts. Outcome data concerning NTS were not derived in any of the studies from rigorously developed, independent data sources. Mixed methods and longitudinal studies in which the unit of analysis is the genuinely multidisciplinary team would be more suitable to study the impact of MR on team NTS.
For a number of years, there have been calls in medical education for research publications to make their theoretical bases explicit [47][48][49][50][51] . Although there is no universally agreed theory behind MR, numerous explanatory constructs have been proposed. Only four studies in this review referred to underlying concepts of MR. Regardless of the paradigm chosen, being clear about the explanatory lens through which inquiry is conducted and the theoretical assumptions that underlie research adds value to it 52 .
Such conceptual frameworks can guide researchers to look at problems in particular ways and are thus crucial in the linkage between theory and empirical data. A combination of neuropsychological and sociological frameworks used by different authors provides interesting directions for future translational research.
To develop a more accurate picture of the relationship between MR, teamwork and collective efficacy, more appropriate measurement criteria and evaluation models are essential. Recent research in elite sports and sociology has emphasized the need for a multilevel approach to examine group constructs [53][54][55][56] . Future research should explore not only the immediate effects on skill demonstration, but broader notions of acquisition and, importantly, application of MR practices among users over time to enhance their performance.
Leadership, contextual and organizational factors shape the success of MR as a routine team and safety practice 34 , and a similarly broad view should be used to understand how MR is embedded in workplace practices that affect safety in high-risk contexts such as surgery.
Preoperative MR has the potential to provide a free, quick and widely accessible tool to augment team performance in theatre, potentially decreasing the number of surgical errors and improving patient outcomes and safety.

Disclosure
The authors declare no conflict of interest.