A heuristic is a rule, strategy, or similar mental shortcut that one can use to derive a solution to a problem. A heuristic that works all of the time is known as an algorithm. Consider the following scenario: you get lost in a maze, what can you do to escape?

The solution is to place your hand on the outer edge of the maze and keep walking until you find the exit. Heuristics can help us make sense of the world in a reliable way and reduce our mental load. However, they aren’t always perfect and there is also a tendency to use them inappropriately to form beliefs. A systematic error that results from the use of a heuristic is called a cognitive bias, although a more accurate definition would be that a cognitive bias is a systematic mistake that derives from limits that are inherent in our capacity to process information.

Since the 1970s, models of cognitive thinking and reasoning were proposed that consisted of 2 separate processing systems: the current evolution of these models has come to be known as dual process theory. The first, ‘intuitive’ reasoning system uses pattern recognition and association to arrive at a rapid conclusion when presented with a collection of data; simplistically, this is Emergency Room thinking. The second processing technique is characterised as ‘analytical’ as it relies on interpretation of data using rules, deduction, and active reasoning; again, simplistically, this is ICU thinking. Heuristics are the mental shortcuts employed in the first processing system, enabling rapid decision making when presented with familiar problems; however, they are not infallible, and blind adherence to them can lead to cognitive and clinical errors.

In experienced clinicians, both systems are used interchangeably and variably, often throughout the clinical journey of an ECC patient.

Cognitive Load

If heuristic shortcuts can be prone to error, why use them at all?

Cognitive load theory proposes the existence of 3 distinct types of memory: sensory, long-term, and working. Sensory and long-term memory both preserve information over prolonged periods for later recall on demand and have an almost infinite capacity. Working memory utilises new/current information and interprets it in light of previously stored data. Working memory, however, has a finite capacity and can be easily overloaded if it is required to process too many pieces of information at one time. The information presented to the working memory to deal with is known as the cognitive load. In clinical settings, 3 factors influence cognitive function: intrinsic load, extrinsic load, and germane load. Intrinsic load is made up of cognitive demands that are related and essential to the task being performed; as other lectures have discussed, prior technical skills training can reduce intrinsic load over time. Extrinsic load is information not essential to the task being performed; robust working practices aligned to the 8 core principles of crisis resource management can greatly assist in reducing extrinsic cognitive load. Germane load relates to information, derived from the task being performed, that contributes to the sensory and long-term memory associated with the task; it is likely that non-technical aspects, such as incivility from colleagues, influence this long-term memory framework as significantly as do the motor activities involved.

In order to mitigate the massive amounts of information presented to the human brain at any time, it has evolved to look for patterns in order to create cognitive shortcuts—heuristics. Although modern life has removed the evolutionary drivers behind this process (for most people), our neurophysiology cannot be denied and so attempts to apply heuristic solutions to other context, such as clinical decision making.

Healthy Heuristics

Heuristics are simple, intuitive decision-making tools that are effective in many clinical situations. They are formed by clinical experience, training, and intuition and in the setting of ‘omnipresent time pressures and [limited] resource availability’ can be extremely useful decision tools in high-pace, high-risk settings such as the ER or ICU.

In these contexts, heuristics can be extremely useful, rapidly identifying potentially life-threatening injuries quickly, as well as enabling the processing of multiple patients in a timely fashion. The use of heuristics in this way also enables the reduction of some information ‘white noise,’ reducing extrinsic cognitive load.

Research carried out in human general practitioners suggests a high degree of accuracy (98%!) in those diagnoses able to be made within a short period of time, and evidence from experienced human emergency physicians suggests that such a heuristic approach can function more successfully than a logistic regression model.

However, as well as our focus tends to be drawn to the most prominent (perceptually salient) information, which may not be a problem as indicated above, we also tend to equate the ‘most prominent’ with the most causally influential stimuli. This, again, may not be too disastrous in the ECC setting, when much of the urgent focus tends to be on critical body systems. Accurate heuristic pathways also take time to develop and can be prone to uncontrolled reinforcement.

One effective way of bypassing the time needed to create heuristic shortcuts in an individual clinician can be the use of checklists for key procedures and transitions in care (e.g., handover forms, anaesthetic/surgical safety checklists), as well as algorithmic approaches to multidisciplinary clinical problems. However, it must be remembered that unconsidered adherence to generic guidelines can be just as error prone as the unfiltered use of type 1, intuitive, reasoning. For this reason, it is essential that any such systems put in place are subject to clinical audit and review, with the results fed back into a loop of ongoing training, systems development, and improvement.

Cognitive Biases

Cognitive bias, or ‘cognitive dispositions to respond’ describe the inevitable tendency of individuals to respond to a situation in a particular manner or predictable way. Cognitive biases are not inherently negative in their nature, but can be when they lead to adverse outcomes or cognitive errors. A large number of cognitive biases have been described, with a great deal of overlap and some confusing terms, but we shall explore those with the greatest relevance to clinical ECC practice.

*, ^, ∼ all refer to interlinked heuristic traps

Various techniques to attempt to mitigate heuristic traps have been proposed, with little convincing evidence of the value of any particular strategy. Therefore, the author would suggest a multifaceted approach:

• Employ ‘cognitive speedbumps’ (10 every 10)
• Apply metacognition (thinking about thinking)
• Apply good Crisis Resource Management techniques, with particular attention to the negative impact of extrinsic mental load and HALTS on cognitive function. Encourage a flat hierarchy within teams, with clinical decision making able to be challenged.
• Judicious use of checklists and guidelines to prevent common errors
• Have Residents/Interns/Students around to make you justify your decision-making!