Hazards in obstetrics leading to harm.
Vignette: due to staff sickness the levels required on a particular shift are less than adequate, and the skill mix may also be inappropriate for the busy workload (perhaps too many relatively inexperienced midwives, with more high-risk patients in labour than expected). Three high-risk inductions were started during the afternoon shift. A patient has developed pyrexia in labour but the recently updated sepsis guideline has not been uploaded so is not available. Her vital signs confirm a rising early warning score (EWS), but the significance is not recognized and therefore not acted on with appropriate escalation. By the time the consultant is called, the patient has developed signs of septic shock and an acute emergency has arisen, compromising both maternal and fetal safety. In the subsequent critical incident review (CIR), the report only focused on the staff member who did not recognize or act on the deteriorating EWS.
Inadequately conducted CIRs will contribute to continuing risk to patients, particularly if systems errors are ignored while concentrating on single errors by individual practitioners. The most important objectives in a high-quality approach to risk management in obstetrics are to delineate all contributory factors (individual, team and system) and to produce appropriate ‘actions’ that are implemented and disseminated widely, in a timely fashion [15].
Errors and Violations
Reason published a useful taxonomy which describes the various types of human error and violation which may compromise patient safety [13]:
Errors
Execution errors: these include slips, lapses, trips or fumbles which can occur even in the presence of an appropriate plan. Error in these cases results from an action failure.
Mistakes: these occur because a plan was inherently faulty or inadequate. Error in these cases results from a higher level failure involving poor planning or judgement, and occurs even when execution of specific actions is done correctly.
All errors are associated with some form of deviation away from the norm.
Execution Errors
Slips tend to be associated with attention errors, either failure to select correctly or failure to recognize when undertaking a task. These often occur during routine procedures when the individual (or team) is distracted or perhaps working in an unfamiliar environment. An example on the labour ward would include surgical error during abdominal entry for an elective third repeat caesarean section, when unexpected patient discomfort/movement leads to distraction and inadvertent bladder damage. Slips also occur when operating in inadequate settings, for example a second theatre in a converted delivery room – cramped, with backup equipment and an inexperienced assistant. Lapses relate to failures of memory or attention and again lead to error during execution of a task. Both slips and lapses tend to occur when there is attentional capture – i.e. when the individual is distracted and/or preoccupied or when changes are made to a pre-defined plan.
Mistakes
Mistakes are either ‘rule-based’ or ‘knowledge-based’ and occur after a problem has been confirmed. In medicine, these often occur with unexpected complications that require a change in approach or deviation away from the agreed plan. ‘Rule-based’ mistakes occur when an individual uses a management approach with which they are familiar, in an inappropriate circumstance (e.g. giving intravenous ergometrine during a large postpartum haemorrhage to a patient with underlying severe pre-eclampsia). Alternatively, they may approach a specific complication incorrectly (e.g. giving 2 l of crystalloid rapidly for reduced urine output in severe pre-eclampsia, leading to fluid overload and pulmonary oedema). ‘Knowledge-based’ mistakes occur in new situations which are unfamiliar to the individual or team, often in the absence of agreed or available guidance to direct care. This type of mistake is less likely to occur with direct input from an experienced clinician. Managing these cases requires critical assessment so that a clear ‘mental model’ is formed which will help to clarify the most appropriate action. However, if the available information is incomplete, incorrect decisions may be made. Any subsequent error may be compounded by various biases. In unfamiliar, safety-critical situations, the tendency is to look for those bits of information that support a pre-determined cause and action. This can lead to a ‘blinkered’ view, with the individual or team excluding information that may be vital in reaching a correct plan of action. This is known as confirmation bias and leads to a loss of situational awareness. Mistakes can result from: relative inexperience (but not exclusively), inadequate challenge from within the team, poor communication or decision making, lack of senior supervision, tiredness and fatigue. These factors will be explored in subsequent sections.
Violations
Deliberate violations imply an intended action with deliberate deviation away from evidence-based guidelines, safe operating procedures (SOPs), etc. ‘Routine’ violations occur when guidance is only partially followed, often to save time or effort. ‘Optimizing’ violations may result from boredom, where moving away from defined guidance is a form of ‘risky’ behaviour undertaken by an individual to compensate. ‘Situational’ violations occur when the action proposed, although not in keeping with agreed procedure, is felt to be a better (or possibly the only) option to follow. Deliberate violations may be related to underlying motivational or team issues which can stem from organizational issues such as poor leadership, inadequate support, failure to correct poor performance, etc. Senior clinicians may deviate from a labour ward guideline based on prior experience and the individual patient’s circumstance. This may be viewed by some as a violation, but can be justified and the reason for deviating from guidance should be clearly recorded. A ‘protocol’ is defined as ‘a system of rules that explain the correct conduct and procedures to be followed in formal situations’. By implication, a protocol does not imply flexibility and violations should not occur – in the labour ward situation this would encompass drug regimens (e.g. magnesium sulphate and intravenous antihypertensive treatment in eclampsia).
Maxfield et al. assessed the frequency of four safety concerns among labour ward teams in the USA: dangerous shortcuts (violations), missing competences, disrespect and performance problems [16]. In total, 3282 participants completed surveys, with 92%, 93% and 98% of physicians, midwives and nurses respectively reporting at least one concern in the previous 12 months. They reported: serious patient harm by a dangerous shortcut/violation (8.4%); performance problems which undermined patient safety (55.3%). It is clear that human error is as relevant in labour ward practice as it is in other safety-critical environments.
Managing Error in Health Systems
Historically, the ‘person approach’ process has been used to manage error in medicine. This often led to blame levelled against one or more individuals. The ‘system approach’ accepts that error is an inevitable consequence of the human state, even in the most expert and aware organizations. ‘High-reliability’ organizations aim to minimize risk and use a system approach for handling error. They acknowledge that ‘to err is human’ and accept that while ‘error is inevitable, harm is not’. Nuclear power plants, nuclear-powered aircraft carriers and air traffic control are three examples [14]. These organizations have developed ‘resilient systems’ and maintain error reduction by addressing issues with individuals, teams, systems, equipment, environment and the totality of the organization. These safety-critical environments, like the labour ward, must maintain reserve capacity to deal with significant and unpredictable peaks of activity. Keogh confirmed that understanding the causes of high mortality in the NHS was not about finding a rogue surgeon, but required a ‘whole-system’ approach [8].
In the following sections we will explore NTS in more detail, highlighting the fallibilities of human cognition and interactions. We will discuss methods which can be used to minimize error to benefit patient safety.
Non-Technical Skills: Cognitive Components
The cognitive components of NTS are situation awareness and decision making. Human cognition (‘thinking’ or ‘reasoning’) describes the mental processes dealing with information gathering, interpretation and use of knowledge. Cognition relies on complex brain interactions involving three types of memory: sensory storage, short-term (or working) memory and main (or long-term) memory. Between 70% and 80% of human errors relate to problems with cognition. In medicine we train in technical skills but the importance of cognitive and non-technical skills training is only now being realized.
A simple ‘dual process model’ can be used to describe the differences between ‘automatic’ (unconscious) and ‘analytic’ (conscious) cognitive processes [3]. These are at play in every aspect of our lives and clinical practice; we switch between them without conscious effort. Undertaking an elective caesarean in a primiparous patient of low BMI is usually a relaxed affair for the middle-grade registrar and conversation will often take place between the operating team and with the patient and partner – ‘automatic’ cognition is in use. However, the situation changes if there is an atonic haemorrhage following delivery of the placenta. The registrar and team focus on the problem, considering options and outcomes as well as appropriate management. Noise levels in the operating room drop to allow the surgeon, anaesthetist and team to concentrate, with a seamless switch to ‘analytic’ thinking. Distractions may lead to error and conversation will be minimal and focused. This is similar to expectations in the aviation industry when an emergency occurs. The ‘silent cockpit’ refers to an environment where unnecessary or irrelevant conversation is prohibited. In many airline disasters, ‘black box’ recordings reveal episodes of distracting conversation, resulting in loss of situation awareness with few routine checks and cross-checks occurring. The development of NOTECHS (non-technical skills training) led to improvement in pilots’ cognitive skills and reduction in air disasters [17].
The ‘automatic’ part of the system is intuitive and does not involve working or short-term memory. Overall active ‘awareness’ is low and actions/skills are undertaken in a reflexive way using honed skills. There is a high degree of ‘automaticity’, with fast thought processes and activity conducted speedily using minimal effort. The process is relatively ‘unreliable’ because of the high degree of automaticity, and the potential for error can be quite high and susceptible to biases. The ‘analytic’ system that we switch to in crisis or emergency situations uses short-term memory, requires conscious effort and awareness, and a deliberate/rule-based approach. It is ‘slow’ and effortful, but less liable to bias with relatively few errors. We use both systems and neither is necessarily better or worse than the other. The automatic part is more primitive biologically, but both systems are required for efficient and effective function, dependent on the clinical scenario at the time. We will often use a more ‘automatic’ approach in busy antenatal clinics, switching to an analytic approach when dealing with a complex obstetric case where there are multiple risk factors. Errors in both elective and emergency situations can arise if the switch from automatic to analytic cognition does not occur at the correct point.
Situation Awareness (or Situational Awareness)
Situation awareness (SA) is a cognitive skill and is the way we maintain awareness of our environment. Endsley defines SA as ‘the perception of the elements in the environment within a volume of time and space, the comprehension of their meaning, and the projection of their status in the near future’ [18]. Three levels are described:
Level 1 is perception of information from the environment using all of our senses. ‘Sensory storage’ will only hold information very briefly –between 0.5 s for visual and 2.0 s for auditory memory [2].
Level 2 is assimilation and interpretation of the information we receive. This occurs in short-term memory which is where we are ‘consciously aware’. Storage in working memory is limited, holding up to seven ‘bits’ of information only and is easily overloaded. External distractions or interruptions can interrupt cognitive processing and lead to error. Over time, ‘expert tasks’ become automatic and are embedded in long-term memory (e.g. managing operative vaginal delivery). This is an advantage for the more experienced accoucheur, as it frees up ‘bits’ within working memory, allowing better SA.
Level 3 is the ability to predict the ‘near future state’. This relies on experiential learning which is stored in long-term memory and uses ‘pattern recognition’. This is a primitive but well-developed process in humans which happens automatically, with minimal conscious processing. However, it is prone to error if the pattern is misinterpreted. If no pattern is recognized then automatic cognition does not function effectively and the process can be distorted by pre-existing biases. More experienced practitioners build up a series of ‘mental models’ or ‘schema’ which are stored in long-term memory and consist of specific patterns or sensory cues which are associated with a particular meaning. This anticipatory function feeds directly into decision making and is prone to error in the presence of distractions or interruptions.
The three levels of SA are shown in Figure 33.2; the mnemonic ‘CIA’ may be a useful reminder.
The importance of maintaining SA cannot be over-emphasized in safety-critical environments. Between 80% and 90% of all aviation disasters involving human error are caused by SA errors and 70–80% of these were level 1 (perception) errors. These errors can affect both individuals and teams. When allocated a specific task that requires concentration and focus, it is possible to miss what may be a very obvious deterioration in the overall situation. Task focus can be likened to putting on a pair of ‘blinkers’ which restrict one’s wider vision. In their classic psychology experiment, Chabris and Simons arranged two teams of students in white or black T-shirts. Each team passes a basketball to each other, while the audience is asked to count the exact number of times the white team passed the ball, ignoring the black team passes. Despite this occurring in a space no larger than a small labour room, half of the subjects missed an adult in a full-sized black gorilla suit crossing the room! [19]. This error of perception is known as ‘inattentional blindness’ and demonstrates the need to learn how to maintain SA, particularly in the presence of multiple distractions and interruptions.
In a situation in which an obstetric team are managing a large atonic postpartum haemorrhage after a forceps delivery, the registrar will be focused on managing the atonic uterus. While applying fundal and possibly bimanual pressure, he or she will also be examining the genital tract to exclude trauma. Their SA will be compromised and they may miss the fact that the patient has developed cardiovascular instability. This requires someone to ‘step back’ from all the other required tasks (e.g. airway management, checking vital signs, siting IV access, sending bloods to labs, etc.). They need to take the lead, keeping a careful eye on the patient’s overall condition and watching for the ‘gorilla in the room’. This may be a senior midwife, particularly if the anaesthetist is occupied with airway management or securing IV access. Effective teams do not rely on hierarchy but ensure a nominated person is keeping an eye on the bigger picture. In an emergency, all team members will have specific roles and tasks to perform, but the ability to watch for overall deterioration using any spare cognitive capacity should be encouraged throughout the team. ‘Cross-checking’ is a tool that can facilitate safety. Each member should feel able to challenge others confidently and constructively, checking that planned interventions are correct. This was one of three NTS identified by Bahl et al. which were specific to conducting operative vaginal delivery (the two other procedure-specific skills were maintaining professional behaviour and a professional relationship with the woman) [5,20].
The following vignette demonstrates how SA errors can arise in labour. A primiparous patient with a twin pregnancy was admitted in labour at 37 weeks; both presentations were cephalic. Monitoring was via fetal scalp electrode and abdominal transducer. Four hours into labour the trace on twin 1 was pathological and a fetal scalp pH was undertaken which was normal. As the trace did not improve, two further samples were taken before full dilatation. The registrar and consultant reviewed regularly. Progress in the second stage was rapid – twin 1 delivered normally in good condition, followed within five minutes by twin 2. Twin 2 was asphyxiated and required prolonged resuscitation: arterial cord pH 6.95, base excess –16 mmol/l. No-one in the team had noticed that the transducers were crossed – the pathological CTG had been from twin 2 for the last four hours. This is a classic level 1 SA error – once missed, the error was not picked up by any of the experienced team.
Various methods are described for maintaining SA, but the following summary could be usefully adopted in obstetric practice [2]:
Consider fitness to work (health, tiredness, fatigue).
Start with a structured, relevant briefing.
Minimize interruptions or distractions.
Prohibit all non-essential activities (maintain a ‘sterile cockpit’).
Regularly update (compare real clinical data with the ‘mental model’).
Monitor (self and others, and vice versa).
Encourage good followership (ask others to cross-check and ‘challenge’).
Good time management (leaves maximum time for critical decisions).
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
