Article Text
Abstract
Background: The use of different forms of coordination according to situational demands plays a crucial role in teams working in complex environments. This study aimed to describe patterns of coordinative actions (CAs) as they occur during anaesthesia induction and to analyse the influence of two crucial situational factors on these patterns, namely the amount of existing standards and the level of task load.
Methods: 23 anaesthesia inductions were videotaped, and CAs of the anaesthesia teams were coded. The coding system distinguished between implicit and explicit coordination, coordination via leadership and heedful inter-relating as the individual effort to reach smooth coordination. Five phases within anaesthesia inductions were determined according to their level of standardisation and task load.
Results: Overall, 67.7% of all CAs were rated as explicit CA and 32.3% as implicit CAs. When we considered the duration of those CAs, we found the reverse tendency (coordination was explicit 40% of the time and implicit 60% of the time). In highly standardised phases, we observed less explicit coordination, less leadership behaviour and less heedful interrelating compared with less standardised phases. In high-task-load phases, we observed more heedful interrelating than in low-task-load phases.
Conclusions: The anaesthesia teams relied greatly on implicit coordination, which contrasts with findings indicating a performance benefit through explicit coordination in other work settings. Standardisation in the form of written departmental directives may have a supportive effect on coordination by partially substituting for other forms of coordination. The effect of high task load should be tested further in a simulator setting, where high task load can be induced in a more controlled fashion.
Statistics from Altmetric.com
Close cooperation between providers in anaesthesia teams is crucial, because some anaesthetic procedures can be technically demanding (eg, tracheal intubation), and unexpected events are relatively frequent.12
The induction and recovery period are often particularly challenging, because the intended smooth transition between alertness and a deeply anaesthetised state of the patient can on occasion12 be jeopardised by events such as unforeseen patient reactions to medication, unanticipated difficulty with airway management and other events of varying significance. Even though coordination strategies within teams and team adaptation to the situation are intuitively felt to be pivotal for successful coping, little is known about the relevance of such coordinative strategies in anaesthesia teams.
We analysed different types of coordination strategies, consisting of either verbal communications or silent actions.
The concept of explicit/implicit coordination34 is used to describe coordination behaviour in relation to the effort spent on the coordination activity itself. Explicit coordination (see table 1) concerns the deliberate effort aimed at managing the task through overt communication. It has been found that explicit coordination is especially appropriate in novel situations and during decision-making. During implicit coordination (see table 1) less effort is spent on overt communication, for example, by providing somebody with relevant information at the right time without the other person having to ask for that information. For instance, a team member exposed to a high task load can be relieved from the coordination activity, thereby having more cognitive capacity to carry out the task itself. Implicit coordination requires, however, mental models shared by all team members.5
In high-risk work environments, standardisation (see table 1) is often used to support coordinated action.67 Standards can support implicit coordination because they enhance the shared understanding of the task. However, the emerging need for explicit coordination in non-routine situations may be missed if teams rely excessively on standards.8–11
Grote et al investigated the effects of standardisation and explicit/implicit coordination on the performance of cockpit crews during simulator training.1213 In highly standardised work phases and in work phases with a high task load, there was significantly less explicit coordination compared with work phases with a low standardisation or low task load, respectively. However, the better-performing teams used more explicit coordination overall.
Leadership is also central in team coordination.14 Zala-Mezö et al found in an interview study15 that anaesthesia team members expected to be allowed high levels of autonomy in routine situations. During emergency situations, however, active and even directive leadership was expected. Yun et al observed similar tendencies in emergency medical teams.16 Previous studies in cockpit teams have shown that high-task-load phases should be prepared during low-task-load phases by explicit coordination and active leadership, while during the high-task-load phase, explicit coordination should be reduced.9 The use of leadership and standardisation as coordination mechanisms seems to be reciprocal. In Grote and Zala-Mezö’s study,22 work phases with high levels of standardisation were characterised by significantly less leadership behaviour. These results were in accordance with earlier research findings suggesting that standardisation served as a substitute for leadership.7
The concept of heedful interrelating (see table 1) describes individuals’ efforts to attain a good team performance based on an attitude that entails carefully following the course of action in the team and continuously matching one’s own behaviour to the behaviour of other team members. Also, the deliberate embedding of actions in a broader physical and social context is an indication of heedful action.21
To date, there are few studies that have investigated the relationships between these different coordination types in medical teams.162324 The present study aimed to investigate these based on two hypotheses.
Written directives can be understood as knowledge shared by the team. In this sense every team member knows what he or she has to do. Any explicit coordination in the form of clarifying, planning or leading the situation is unnecessary. It is sufficient to occasionally follow the work processes of the other team members to check if everything is working out.
Hypothesis 1
In highly standardised work phases, we expected more implicit coordination, less leadership and less heedful interrelating compared with low standardisation phases.
If a team member has a delicate task, like inducting the tube into the trachea, they have less capacity to listen and to talk, since they are absorbed by the task. The person is experiencing a high task load. In such a situation, implicit coordination can be undertaken by another, less busy, team member by silently giving them a hand.
Hypothesis 2
In high-task-load situations, we expected more implicit coordination, less leadership and more heedful interrelating compared with low-task-load phases.
Studying these hypotheses also helps to answer the following very practical questions:
Which coordination form is the most appropriate in a given situation?
Which particular shifts in coordination forms are required as changes in situational demands occur?
The answers to these questions can also guide the development of specific team training content for medical personnel.
METHODS
Anaesthesia inductions were analysed because they offer the opportunity to study coordination within anaesthesia teams without much interference from other teams. Cases were considered eligible if a general anaesthetic without invasive monitoring lines was planned (inductions with tracheal intubations). Informed consent of the patient and staff was obtained, and 23 anaesthesia teams were videotaped. The composition of the teams is described in table A1.
Taping was started when the patient was brought to the induction room and stopped at the time of transfer to the adjacent OR. A Panasonic NV-DS27 camera attached to a mobile stand was placed so as to obtain an overall view of the anaesthesia team. An external microphone was mounted close to the team to enhance the acoustic recording quality. The videotapes were analysed using the software ATLASti,25 which facilitates the analysis of large quantities of video data.
An experienced and human-factors-trained anaesthetist defined five main phases of anaesthesia induction and assessed the level of task load for each work phase. Work phases were classified as low, moderate and high standardisation phases according to the number of written rules26 pertaining to the respective phase. The rules analysis was carried out for four departmental directives relevant to anaesthesia induction: standard anaesthesia, residents’ duties; shift organisation; and anaesthesia protocols (see table 2).
The coding scheme for CAs included four main types of coordination (implicit, explicit, leadership, heedful interrelating; see tables 1, A2) with overall 27 subcategories derived as appropriate from existing behavioural marker systems line operation safety audit (LOSA)27 and non-technical skills (NOTECHS)28 and from descriptions in the literature of explicit/implicit coordination,4 task-oriented leadership29 and heedful interrelating.5 The categories of explicit and implicit coordination were mutually exclusive and were applied to all task-relevant utterances. The categories related to leadership and heedful interrelating were coded additionally only for selected utterances according to the definitions of these categories. Almost all leadership subcategories referred to CAs which were also coded as explicit coordination, while heedful interrelating CAs concerned both implicit and explicit coordination. The inter-rater reliability (two independent raters: two of the authors of this article) of coding was 84% in a test sample of eight cases out of the 23. Differences in coding were discussed, and codes were reassigned based on the agreements reached.
To test differences in the occurrence of the different coordination forms between work phases, we used the non-parametric Pearson chi square test (SPSS, Chicago). In our analyses, we focused on frequencies of use of different coordination forms across the 23 teams because we were interested in adaptiveness regarding use as such, not in temporal use patterns.
RESULTS
Table 3 shows the frequencies of types of CAs across the five work phases. Overall, 67.7% of the CAs were explicit, and 32.3% were implicit. Leadership and heedful inter-relating occurred rarely, with a mean across the five work phases of 14.6% and 9.3%, respectively. There were several differences in frequencies between teams (implicit/explicit CAs: SD 6.1; leadership CAs: SD 5.5; heedful inter-relating CAs: SD 4.4).
The mean duration of a CA was 7.8 s. Implicit CAs tended to take longer, as 60% of the time overall was spent on implicit and 40% on explicit coordination (see Appendix A).
The relationship between coordination forms and level of standardisation was tested by comparing CA frequencies in phase 1 (low task load and high standardisation) and 5 (low task load and low standardisation) because these two phases differed considerably in standardisation and had the same task load level. In accordance with hypothesis 1, there was a higher share of implicit coordination and less leadership and heedful inter-relating in phase 1. These results were statistically significant (Pearson chi-square test values of 17.05, 43.05 and 32.62, respectively, p<0.001 in all three tests).
To test the relationship between coordination pattern and task load, we compared phases 2 (moderate standardisation and moderate task load) and 3 (moderate standardisation and high task load). They had the same level of standardisation but varied regarding the level of task load. We found significant differences only in the category heedful inter-relating in accordance with hypothesis 2: more heedful inter-relating was observed in phase 3 (Pearson chi-square test value: 5.02, p<0.01). Differences concerning explicit coordination and leadership were non-significant.
DISCUSSION
Our video-based analysis found that anaesthesia teams coordinated implicitly during 60% of the time during routine inductions. Explicit coordination (40% of the time) however was characterised by a higher density of coordinative actions (CAs), with explicit CAs representing 67.7% of all CAs. In highly standardised phases, we observed lower frequencies of explicit coordination, leadership behaviour, and heedful inter-relating compared with less standardised phases. These findings support our first hypothesis, suggesting that standardisation may ease implicit coordination and may serve as a substitute for leadership, as has already been shown in other work domains.12
In high-task-load phases, we observed more heedful inter-relating than in low-task-load phases and so found less evidence to maintain our second hypothesis (more implicit coordination and heedful inter-relating but less leadership in high-task-load situations). This could be due to only moderate observed differences in task load in our study setting. Hence, this issue could be better addressed in a simulator setting allowing artificial creation of high-task-load situations.
Methodologically, we demonstrated the feasibility of recording data in a live clinical setting, using mostly basic, easily accessible equipment, and developed a theory-based, reliable rating system for coordination behaviours useful also for future analyses of team coordination.
The association of work phases involving higher standardisation with more use of implicit coordination, less leadership and less heedful inter-relating has been reported in other work domains. However, we are not aware of any comparable reports referring to the specific setting of anaesthesia. Given the overall rather low level of standardisation in anaesthesia in comparison with other high-risk work environments, the generally high level of implicit coordination in our results was unexpected. It may be related to the immediacy of common action in a shared visual field in anaesthesia, and to a comparatively high degree of familiarity in anaesthesia teams, which may compensate for the lack of common ground through standardisation.30
This study is limited by its observational design allowing only the detection of associations and not of underlying causality. An established system for observational categories is also missing. A less detailed categorisation would allow a more economical way of analysing this kind of behavioural data. In addition, our document analysis may have missed elements of standardisation of other origins, like action standards acquired by team members during their medical training, continuing education and independent study of literature.
In summary, our results show a high occurrence of implicit team coordination during anaesthesia inductions and indicate that less leadership and heedful interrelating was used in highly standardised work phases. It can be speculated that in the setting of routine anaesthesia inductions, standardisation may be an economical way to facilitate workflow by substituting for leadership and heedful interrelating in certain phases. Considering written rules as part of the shared knowledge about the hazards connected to work procedures has important safety implications for medical practice. However, standards should be chosen carefully so as not to reduce the degree of freedom for adaptive action in these highly uncertain working environments. Further research is needed to establish the presumed relationships using an interventional, controlled design and to define reliable measures for team performance to relate coordination patterns with relevant, preferably patient- and safety-related outcomes. If future research results in defining a relationship between coordination patterns and relevant outcomes, implications for a variety of fields can be derived. These range from better founded principles for team training or focused training of standard procedures in anaesthesia to leadership directives dealing with the shift from implicit to explicit coordination during unexpected events.
Appendix
REFERENCES
Supplementary materials
web only appendix 18/2/127
Files in this Data Supplement:
Footnotes
Competing interests: None.
Ethics approval: Ethics approval was provided by the ethic commission of the University Hospital of Zurich.
Patient consent: Obtained.
▸ Additional appendix is published online only at http://qshc.bmj.com/content/vol18/issue2