Exploring anaesthetists’ work as part of a sociotechnical system: A human factors perspective

Exploring anaesthetists’ work as part of a sociotechnical system:  A human factors perspective

Exploring anaesthetists’ work as part of a sociotechnical system:  A human factors perspective

Provided on behalf of the Chartered Institute of Ergonomics and Human Factors (CIEHF) from Dominic Furniss, Erik Berndt and Ann Blandford, University College London

There has been a growing interest in human factors in healthcare to improve clinical performance and to address issues of safety (e.g. NHS England, Human Factors in Healthcare). Human factors aims at: “enhancing clinical performance through an understanding of the effects of teamwork, tasks, equipment, workspace, culture and organisation on human behaviour and abilities and application of that knowledge in clinical settings” [1]; however, the concept of human factors has been poorly understood – for example, it has been equated with training, non-technical skills, crew resource management and other strategies that are intended to change human behaviour [2]. More broadly, human factors seeks to improve the design of systems to better aid people, i.e. it is focused on system changes (rather than people changes) to improve performance.

Ball & Frerk (2015) discuss the reactive approaches to safety in anaesthesia as part of Safety 1, and the proactive approaches as part of Safety 2 [3]. These can be seen as two sides of the same coin. Even within human factors, Safety 1 has been the dominant perspective – for example, people review incidents that have occurred to address the causes that may reduce the likelihood of recurrence [4–6]. Safety 2 looks more at the positive side of maintaining safety on a daily basis so that abilities can be enhanced. For example, it seeks to understand how professionals compensate for poor circumstances and poor systems to maintain safety and performance. This includes analysing sociotechnical systems during ‘normal’ performance in the absence of anything going wrong. This article reports results from the application of two techniques to explore anaesthesia practice: contextual inquiry and distributed cognition for teamwork (DiCoT).

Contextual inquiry

Contextual inquiry involves observations of, and interviews with, people as they work [7]. This allows one to focus on understanding the details of normal work, with the opportunity to ask about the details of that work while it is being done (or shortly after if the person should not be interrupted). Analysing this data involves constructing five models, each of which highlights a different aspect of the work:

flow model – shows the overall organisation and coordination of the workflow;

sequence model – gives a description of tasks;

artefact model – represents objects used for work;

cultural model – provides a schematic overview of informal and broad influences on actors; and

physical model – illustrates the workplace’s physical structure.

Collectively, these five models build an understanding of ‘context’, and as they are developed issues and areas for improvement can be identified.

Distributed cognition for teamwork (DiCoT)

The second technique, DiCoT [8,9], gathers data in a similar way to contextual inquiry and builds on the idea of constructing different models of work; however, it builds in theory and principles from distributed cognition (DCog) [10]. DCog was proposed in the 1990s as a contrast to traditional notions of cognitive psychology. Rather than restrict accounts of cognition to faculties inside the skull, DCog removes these limits, so accounts of cognition can include artefacts and tools in the external world. Here, a blind person’s stick becomes part of the cognitive system, and a shopping list can be considered a form of memory. Seminal works involving DCog include analysing how people, processes and equipment work together in complex cognitive systems such as the bridge of a ship [10] and an aircraft cockpit [11]. DCog has also been used in different healthcare settings, e.g. communication in surgery [12], infusion pump use in the ICU [13], and blood glucose meter use in an oncology ward [14]. Fioratou and co-workers (2010) use DCog to argue for the relevance of ‘distributed situation awareness’ for thinking about safety and training in anaesthetic practice [15]. This study differs in that it is an empirical analysis of anaesthetic practice; we do not focus on non-technical skills, situation awareness or training; and we use DiCoT to shape our analysis.

DiCoT was developed as a more structured approach to help apply DCog. It essentially combines the structure of contextual inquiry with the theoretical perspective of DCog. DiCoT has five models with associated DC principles (see [9] for a summary) – Table 1 describes the principles from the information flow model for illustrative purposes.

Information flow model – shows how information flows between actors in tasks. Its principles include: information movement and transformation; information hubs and buffers; communication bandwidth; informal and formal communication; and behavioural trigger factors.

Artefact model – focuses on the structure of tools and representations. Its principles include: mediating artefacts, creating scaffolding; representation-goal parity; and coordination of resources.

Physical model – focuses on the layout of the environment, e.g. a desk or a room. Its principles include: space and cognition; perceptual principle; naturalness principle; subtle bodily supports; situation awareness; horizon of observation; and arrangement of equipment.

Social model – focuses on the social relationships, responsibilities knowledge and goal sharing between individuals in the system. Its principles include: social structure and goal structure; and socially distributed properties of cognition.

Evolutionary model – shows how the system has changed over time. Its principles include: cultural heritage and expert coupling.

Contextual inquiry and DiCoT are complementary approaches; one builds on the other. We applied both to (i) investigate the challenges of learning and applying each approach, and (ii) provide insight into the process and outcomes of using each approach. These results are reported in [16]. In this article we report a summary of the issues found after using both approaches.


Data gathering was conducted in operating theatres of a busy hospital in London. The focus was on how anaesthetists use infusion devices as part of a broader sociotechnical system. Ten elective abdominal operations were observed with anaesthetists who have a special interest in liver transplants, and one emergency operation totalling 40 hours. Sessions were conducted during the morning (the normal time for operations), and each took 3–5 hours. In total, 11 different anaesthetists were interviewed and observed.

The analysis took place in three parts: part one applied contextual inquiry; part two applied DiCoT; and part three was a reflection on applying both approaches. Contextual inquiry was learnt and applied first because DiCoT builds on contextual inquiry’s techniques and models. The application of both methods involved three sub-parts: learning the method, data gathering and evaluation. Parts one and two took four to five weeks each; part three took three weeks.


We have organised our results into seven themes which highlights how work is distributed over multiple people and devices in this workspace.

Table 1

Communication and coordination between anaesthetic team

Two anaesthetists form the anaesthetic team. Both are responsible for maintaining the homoeostasis of the patient and for supervision of the three subsequent steps of anaesthesia (inducement, maintenance and ending of anaesthesia). The interaction between both anaesthetists is usually face-to-face and includes formal communications about the patient’s state and future therapy and also informal communication about prior experiences and work-unrelated topics. There is often only one anaesthetist present, however, making the use of a phone necessary; this reduces the communication bandwidth between the anaesthetists significantly. Furthermore, only having one anaesthetist present reduces the opportunity for double-checking for errors, a shared memory and the ability to process information and absorb interruptions. During informal shift changes, following a swap of anaesthetists, information might be omitted. In rare cases, both anaesthetists may leave the room at the same time, which may lead to further communication and coordination issues.

Communication between anaesthetists and surgeons

Communication between the anaesthetists and surgeons is critical but it can be hindered because it takes place across a barrier that separates the anaesthetic area from the surgical space (see Figures 1 and 2). There is also a barrier in terms of the surgeons wearing masks, which additionally decreases non-verbal communication through facial cues; the arrangement of equipment can also be a barrier that reinforces their separate working spaces.

Figure 1: Photo of the workspace of the anaesthetists showing the barrier between the anaesthetic area and surgical space

Figure 1: Photo of the workspace of the anaesthetists showing the barrier between the anaesthetic area and surgical space

Fig 2

Communication with the patient

Direct communication with the patient was impossible during the operations we observed because they were all unconscious. Their vital parameters are measured and transformed into a graphical representation on screen so the anaesthetists can monitor their state. The patient’s face may give indications about the state of the anaesthesia but this channel is blocked because the face is covered during the operation.

Interruptions from other healthcare professionals

The operating department practitioner (ODP) prepares equipment and helps with tasks to facilitate the anaesthetists’ work. Besides communications with them and the surgical team, which could have different content, ranging from patient-related topics (“We have a severe blood loss”) to work-related (“Could you please adjust the height of the operating table?”), interruptions could come from outside. For example, there may be external persons entering the workflow by asking for any kind of information that are not related to the current operation. These might disrupt the anaesthetists’ work.

Interruptions from external devices

People are not the only source of interruptions as external devices, like phones and laptops, are sometimes used during the work and may distract anaesthetists from the actual work.

Monitoring devices

Observing monitoring devices is essential for tracking the state of the patient and diagnosing problems. Interpreting figures and abstract representations of the patient’s state is a complex process the anaesthetist is skilled at. The arrangement of equipment may impede effective monitoring. For example, attaching many syringe drivers to one pump stand may mean that some pumps obscure information from others (see Figure 3).

Infusion pumps_figure 3 lo-res

Figure 3: Showing a stack of pumps where the higher pump obscures information from lower pumps

Controlling devices

The functioning of monitors and infusion devices could be adjusted by, sometimes cumbersome, manual input. This turns the anaesthetist’s attention to controlling the devices rather than delivering patient care. For example, in one case the epidural pump’s battery was put in the wrong way round, which caused a delay before the problem was properly diagnosed. This is despite many batteries being designed so they can only be inserted the right way round eliminating this problem in common consumer electronics. The epidural device also needed a special physical key to access almost all of its functions (see Figure 4) and some needed a password in addition to this, which was difficult to memorise. Some anaesthetists stored the password on their smartphones. They also found it more convenient to calculate dosages on their phones – showing how they use external artefacts and adapt their behaviour to support their work. One of the anaesthetists gave a bolus manually because the integrated feature on the epidural pump was so hard to use. Another alluded to the poor usability of their equipment: “They are probably specifically designed to be awkward – I can’t believe that someone designed them who normally uses them. They might be reliable mechanically, but the interface…”

Fig 4


We found that using these human factors approaches facilitated thinking about anaesthetists’ work as part of a sociotechnical system. Contextual inquiry was easier to learn, but DiCoT’s principles encouraged deeper reflection and insight [16]. DiCoT emphasised how information was propagated and transformed in the system, including communication issues, interruptions, and monitoring and controlling devices. The usability of devices was a source of frustration. DiCoT also provided insights into some of the strengths and weaknesses of the system in the absence of anything going wrong, which is more in keeping with Safety 2 rather than Safety 1. In many cases, anaesthetists will proactively organise their (cognitive) environment to foster quality and safety. For example, more junior anaesthetists are provided more support and less complex cases; more critical drugs are within easier reach for monitoring and control; and interruptions are not tolerated during or approaching complex periods.

While these approaches and results show promise, the observations were only carried out in one hospital in the UK, and the data gathering was predominantly done with anaesthetists specialising in liver operations; however, DCog seems relevant for understanding the complexities of anaesthetic practice more generally [15]. In common with other methods, contextual inquiry and DiCoT will emphasise some factors and de-emphasise others.

This can be perceived as a blessing in terms of leading the analyst towards issues, and a curse in terms of biasing the analyst’s perspective away from other noteworthy issues. Alternative human factors methods could produce different results. Understanding normal anaesthetists’ work as part of a sociotechnical system could lead to system enhancements in the longer term.


  1. Catchpole (2010), cited in Department of Health Human Factors Reference Group Interim Report, 1 March 2012, National Quality Board, March 2012. Cited in NHS England, Human Factors in Healthcare – A Concordat from the National Quality Board, www.england.nhs.uk/wp-content/uploads/2013/11/nqb-hum-fact-concord.pdf (accessed 4 January 2016)
  2. Russ, A.L., Fairbanks, R.J., Karsh, B.T., et al. (2013). The science of human factors: separating fact from fiction. BMJ quality & safety, bmjqs-2012
  3. Ball, D.R. & Frerk, C. (2015). A new view of safety: Safety 2. Br. J. Anaesthesia 115(5), 645–647
  4. Cooper, J.B., Newbower, R.S., Long, C.D. & McPeek, B. (1978) Preventable anesthesia mishaps: a study of human factors. Anesthesiology 49(6), 399–406
  5. Chopra, V., Bovill, J.G., Spierdijk, J. & Koornneef, F. (1992) Reported significant observations during anaesthesia: a prospective analysis over an 18-month period. Br. J. Anaesthesia 68(1), 13–17
  6. Arbous, M. S., Grobbee, D. E., Van Kleef, J. W., et al. (2001). Mortality associated with anaesthesia: a qualitative analysis to identify risk factors. Anaesthesia, 56(12), 1141-1153
  7. Beyer, H. & Holtzblatt, K. (1998) Contextual Design: Defining customer-centered systems (reprint edn) Morgan Kaufmann, San Francisco
  8. Furniss, D. & Blandford, A. (2006) Understanding emergency medical dispatch in terms of distributed cognition: a case study. Ergonomics 49(12–13), 1174–1203
  9. Blandford, A. & Furniss, D. (2006) DiCoT: a methodology for applying distributed cognition to the design of teamworking systems. In Interactive systems: Design, specification and verification (pp. 26–38), Springer Berlin Heidelberg
  10. Hutchins, E. (1995a) Cognition in the Wild. MIT press
  11. Hutchins, E. (1995b) How a cockpit remembers its speeds. Cognitive Science 19(3), 265–288
  12. Hazlehurst, B., McMullen, C.K. & Gorman, P.N. (2007) Distributed cognition in the heart room: how situation awareness arises from coordinated communications during cardiac surgery. J. Biomedical Informatics 40(5), 539–551
  13. Rajkomar, A. & Blandford, A. (2012) Understanding infusion administration in the ICU through distributed cognition. J. Biomedical Informatics 45(3), 580–590
  14. Furniss, D., Masci, P., Curzon, P., et al. (2015) Exploring medical device design and use through layers of distributed cognition: how a glucometer is coupled with its context. J. Biomedical Informatics 53, 330–341
  15. Fioratou, E., Flin, R., Glavin, R. & Patey, R. (2010) Beyond monitoring: distributed situation awareness in anaesthesia. Br. J. Anaesthesia 105(1), 83–90
  16. Berndt, E., Furniss, D. & Blandford, A. (2014) Learning contextual inquiry and distributed cognition: a case study on technology use in anaesthesia. Cognition, Technology & Work 1–19


We would like to thank the participants for the time they gave to this study. This work was supported by the UK Engineering and Physical Sciences Research Council [EP/G059063/1] requirements engineering and usability.


Dominic Furniss

Dominic is a senior research associate at University College London. He works in the area of human computer interaction (HCI) and human factors applied to healthcare. His research includes qualitative approaches to understanding sociotechnical systems and evaluating medical device design and use. He has an interest in usability, distributed cognition, resilience engineering and patient and public involvement (PPI).

Ann Blandford

Ann is professor of human–computer interaction at University College London and director of the UCL Institute of Digital Health. She is an expert on the design and use of interactive technology in healthcare delivery, and particularly on how to design systems that fit well in their context of use and for their intended purposes. She is principal investigator on an NIHR grant, ECLIPSE, studying infusion pump design. She has published widely on the design and situated use of interactive health technologies, and on how technology can be designed to better support people’s needs.

Erik Berndt

Erik is a human factors engineer in industry and a visiting researcher at University College London. His scientific work focuses on human–computer interaction (HCI) in complex work systems. He applies qualitative and quantitative approaches to understand technology use and to inform design improvements of future technology. He has an interest in distributed cognition, automation psychology, requirements engineering and usability.

Categories: ARTICLES