Human error doesn't exist in isolation: it's now time for a deeper understanding of human factors in anaesthesia

The term ‘human factors’ in medicine has been used flexibly to have a multitude of meanings. Frequently, it is used pejoratively as ‘the human factor’ when something goes wrong – something to be managed, removed or minimised [1]. Perhaps more often in health and anaesthesia, ‘human factors’ has been synonymous with the skills required for effective teamwork, also termed, ‘non-technical skills’, ‘soft skills’ or ‘teamwork competencies’ [2]. In reality, this is just a small part of human factors as a scientific discipline [3]. As its own field of study, human factors/ergonomics is the recognition that we work in a sociotechnical system, where humans interact with other humans, procedures and equipment within an imperfect environment [4, 5]. The aim of the ergonomist is to ensure that the capabilities and weaknesses of humans in the system are accounted for when designing for safety and efficiency. It is notable that the first anaesthetic guidelines to use the term ‘human factors’ in a broad, sociotechnical sense come from the traditionally technical domain of airway management – the guidelines having been co-authored by members of the Difficult Airway Society (DAS) [6]. Looking back to the 2004 DAS guidelines, there is barely a mention of the team or wider human factors aspects of airway management [7]. Similarly, the ASA guidelines for managing a difficult airway and many other documents of the time, fall into the same pattern of describing the tools, tasks and ‘toys’ used in challenging events [8]. At that time, this approach may have been warranted. There was a proliferation of devices available for airway management, with confusion about when to use each. Difficult airway trolleys were overladen with equipment. Their contents were not standardised and reflected the firmly held views of local enthusiasts, often with minimal evidence of effectiveness. A situation starkly apparent when the myriad homespun devices to manage transtracheal ventilation were reviewed [9]. In the last 18 years, two events have radically changed our view of airway management. First, our specialty and the communities we serve owe an enormous debt to Martin Bromiley, who raised the alarm of the human factors failures in airway management in the wake of personal tragedy [10]. His efforts and those of the Clinical Human Factors Group he founded alerted all health professionals to the effects of suboptimal teamwork, processes and aspects of training. These aspects were further highlighted and quantified by the Fourth National Audit Project (NAP4) in the UK which provided the impetus for systemic change internationally [11]. These events served to steer the direction of the anaesthetic community towards investigating team processes, training and readiness in airway management. The new guidelines on human factors implementation go further. They herald the start of the specialty's transition to a more comprehensive understanding of performance based on features beyond just the performance of clinicians ‘at the sharp end’ [6]. The authors recognise that the vocabulary and some of the concepts might currently be unfamiliar to many. Those that are familiar to most may be the items surrounding teamwork and non-technical skills. This is not surprising and is likely as a result of the anaesthetists non-technical skills (ANTS) framework and a focus on these aspects in training and continuing education over the last decade [12]. It is, therefore, perhaps more informative to focus on the recommendations that have solutions outside of the resources provided in the accompanying tables that focus on these teamwork aspects. Design is not in the traditional skillset of anaesthetists, yet three of the 12 recommendations focus on design issues [6]. Medical equipment is regulated in the UK by the Medicines and Healthcare products Regulatory Agency (MHRA) [13]. Similar to many regulatory organisations, it mandates a human factors, user-centred design process of prototyping and testing in accordance with international standards (IEC 62366-1:2015). This process does not have to be particularly extensive and may be of limited value in demonstrating that the new device does not pose a risk. Furthermore, the testing during the process of design might not reflect the local context of the use of a device or consider how other equipment or processes may interfere with its use [14]. Additional local testing before introduction is highly preferable as it may uncover serious latent threats of its use within the specific context. Unfortunately, the resources to do this are rarely available in a publicly-funded health system. Nevertheless, costs are in many cases justifiable in high-risk, high-volume purchases such as for example a fleet of defibrillators which may cost several millions of pounds and may be used in multiple contexts [14]. The expertise to undertake usability testing is rarely found in health settings and this must be an aspirational target. The goal of training anaesthetists as human factors leads may allow limited device testing in every department but communication between these leads and additional assistance to co-ordinate these efforts will also likely be required to avoid unnecessary replication. Arguably, cognitive aids and emergency flowcharts should also be conceptualised as medical devices as they are tools that affect the decision making of a team [15]. Their designs must also be tested to check they are suitable to be used in the context of an emergency. A department human factors lead may also have a role in the design of clinical spaces before they are built. The Canadian experience has been that mock ups and simulation of clinical spaces such as operating theatres can have substantial cost benefits over their lifetime [16]. Testing before building can minimise areas of the room that are rarely used and conversely ensure no potential high-traffic areas exist where collisions between sterile and non-sterile items can occur. Using the actual equipment in simulations can confirm that there are no visual obstructions to monitors and emergency call systems or physical obstructions to doors or pendants [16, 17]. It is not uncommon for safety risks to be built into the clinical setting, with staff often responding with ‘if only they'd asked us first’. Nowadays, no operating theatre should be commissioned for use without first undertaking in-situ simulation testing of the area. Some interventions, such as drug labelling, packaging and storage, are problems that persist despite ongoing efforts [18]. Standardisation and interventions that may improve outcomes, such as red barrelled syringes and ampoules with red caps marked ‘paralysing agent’ for neuromuscular blocking drugs must be communicated and adopted with a collaboration of other agencies and manufacturers [19]. Other interventions that have not worked but have become ritual such as the use of Tall Man lettering, where selective parts of the medication name are capitalised (eg. tRAMadol) should be phased out as they only serve to give the impression of meaningful action on safety [20]. The role of the proposed human factors lead in the new guidelines must be a diverse one including usability testing of equipment and protocols, evaluation of new clinical spaces and identification of medication hazards. It will be a complex task and one that will require training from specialists that work outside of healthcare. Clinical knowledge and context of usability testing has to date been lacking and the suggestion of embedding an anaesthetist in this role is potentially revolutionary in moving to a proactive patient safety approach. ‘Human error’ is an attribution in hindsight of a person's actions that they should have done something differently. This value judgement is made without fully understanding the complexity, options or cues that they were faced with at the time [21]. What we think of as an error is merely a retrospective judgemental label that represents an old view of thinking. Unfortunately, this way of thinking permeates our culture and current investigation system of adverse safety events. Clinicians rarely come to work planning to make the wrong decision in a set of circumstances; they are led to that decision by training, previous experience and the setting they find themselves in. The overriding aim of any investigation must be to understand those facets in all their complexity. This can only be done by using an appropriate tool that allows for the drafting of meaningful recommendations that move beyond just more training, regulation or new protocols. Root cause analyses see a sequence of events as simple and linear rather than complex and interdependent and the fixes put in place as a result are often weak or ineffective [22]. While the Swiss cheese model has served us well in starting our journey into safety science, it is time to discard it and reach for more detailed, complex and nuanced models to understand the world as it is. Newer methods such as the Systems Theoretic Accident Model and Process (STAMP), AcciMAP and Functional Resonance Analysis Methodology (FRAM) look at interdependencies and interactions between components of the system. These methods are dynamic, observing how individuals might act under different circumstances and how different levels of the system such as regulatory bodies and management as well as production pressures might change the circumstances [23]. Patterns may be found that will inform future clinicians and managers as to how best adapt to the system or re-engineer it, rather than finding a simple, convenient but ultimately unhelpful root cause. Through this dynamic, systems modelling lens, Dekker argues there is no such thing as human error except as a symptom of a deeper problem [21]. Adverse events do not have just one cause, the causes are almost never the fault of one individual and the factors are multiple, interrelated and cannot be easily explained with oversimplified methods. In our traditional way of thinking, now termed ‘Safety 1’, improvements can only be made by finding out what went badly when outcomes were poor. Complex, systemic root causes such as staffing levels are either rarely found or not amenable to an easy remedy. Instead it is far more likely that human error will be the cause and a scapegoat found. Along with moving beyond ‘human factors’ as an accusatory term for the failings in healthcare, we also need to recognise the human as a source of potential strength. More commonly than causing harm, harm is avoided by the quick thinking of a health professional adapting to unique circumstances they faced that would otherwise have been hazardous. Humans are not the inconvenient weakness in the system, but the powerful component that recognises patterns pointing to danger and adapts to avoid it. Removing or minimising the human in a complex sociotechnical system such as health and urging them only to follow the agreed protocol in all circumstances would be a mistake [24]. To understand how we avoid so many potential incidents we need to fully comprehend why things go well. We need to fully appreciate the cues that teams and individuals attend to, and the fixes, adaptations, and decisions in those moments. This is the core of what is termed the ‘Safety 2’ approach. Furthermore, as events such as airway disasters become more infrequent, we arguably have fewer lessons we can learn from them and by necessity must look to the successes to advance our knowledge [25]. The new guidelines on implementation of human factors into anaesthetic practice mark the beginning of a new phase of safety science in anaesthesia. A phase where we recognise the fallacy of the error label and reach towards a deeper understanding of the complex sociotechnical system we work in. Possibly the most complex of sociotechnical human endeavours ever attempted. SM is an Associate Editor of Anaesthesia and a director of Vital Capacity Pty Ltd, a human factors safety consultancy for healthcare organisations.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.