Clinical judgement in diagnostic error

Although diagnostic errors have many causes, clinical judgement is a significant contributing factor

Every day, doctors have to rely on their clinical judgement when assessing and diagnosing patients.  Their clinical judgement is informed by information gathered from patients, their observations on examination and their personal experience, knowledge and critical-thinking skills.

Clinical judgement is a complex process that involves various cognitive functions. So it is easy to understand that it features often in clinical negligence claims and medical complaints for alleged delayed diagnoses or misdiagnoses.  But why does this happen?

Cognitive errors

Many types of cognitive errors can occur during the diagnostic process and can arise from several sources, including knowledge deficits, faulty heuristics (mental shortcuts in the thought process, saving time), and ‘affective influences’/‘situativity’ (where knowledge, thinking, and learning are located in experience).

Knowledge deficits 

It might seem logical that errors occur due to inexperience or gaps in a doctor’s knowledge. Thus, a reasonable assumption is that younger, less experienced healthcare providers are at greater risk of diagnostic pitfalls than experienced clinicians. This can be true but, in our experience, other issues are more prominent features such as problems with obtaining accurate clinical information by way of patient history, test results, etc. Many diagnostic errors are associated with common diseases and conditions, suggesting that other problems with clinical reasoning – such as faulty heuristics, cognitive biases, and affective influences/situativity are more relevant than knowledge deficits. 

Faulty heuristics and cognitive biases

Given the time constraints of medical practice, doctors have to use mental shortcuts to make decisions, but this can be prone to errors. Cognitive biases occur when heuristics lead to faulty decision-making. Some common biases included anchoring, availability, overconfidence, and context effect.

Anchoring

Anchoring refers to a tendency to ‘anchor’ to, or rely too much on, a particular piece of information – often the initial information obtained, the first symptom, or the first lab abnormality. Anchoring is closely related to several other biases, including:

Under-adjustment, which is the inability to revise a diagnosis based on additional clinical data.

Premature closure, which is when a doctor stops gathering more information (eg, patient history, family history, and medication list) before all of the information is known.

Primacy effect, which is the tendency to show bias toward primary or initial information.

Confirmation bias, which is the tendency to focus on information that confirms an initial diagnosis or to manipulate information to fit preconceptions.

Availability

Availability bias can occur if a clinician considers a diagnosis more likely because it easily comes to mind. Past experience and recent, frequent, or prominent cases can all play a role in availability bias. For example, a clinician who has recently diagnosed an elderly patient with dementia might be more likely to make the same diagnosis in another elderly patient who has signs of confusion and memory loss – when, in fact, the patient’s symptoms might be indicative of another problem, such as vitamin B12 deficiency.

Overconfidence

Overconfidence bias can occur when clinicians overestimate their own knowledge and ability, which can prevent them from gathering and assessing ample information.

Framing effect

This can occur if a clinician misinterprets information or a situation based on the way in which it is presented. For example, if a patient presents with chest pain and has a known family history of heart disease, a clinician might interpret the pain as a likely symptom of a heart attack, when in fact the cause is a broken rib.

Affective influences/situativity

Whereas cognitive biases are lapses in thinking, the term ‘affective influences’ refers to emotions and feelings that can sway clinical reasoning and decision-making. For example, preconceived notions and stereotypes about a patient might influence how a patient’s signs and symptoms are viewed. If the patient has a history of substance abuse, for instance, the provider might view reports of pain as drug-seeking behaviour. Although this impulse might be accurate, the patient could potentially have a legitimate clinical issue. Doctor and patient characteristics (eg, age, gender, socio-economic status) can actively influence clinical reasoning.

Factors such as environmental circumstances (eg, high levels of noise or frequent interruptions) and a doctor’s own health/mood (sleep deprivation, stress, anxiety disorders) also affect the ability to make sound clinical judgements. 

Case study

A 40-year-old male presented to his GP with sternal pain after lifting a boat in his backyard. The pain increased when the patient raised his arms. An ECG was ordered and the results were negative. The doctor decided that muscle strain was the cause of the patient’s symptoms. A few days later, the patient died from a heart attack.

Discussion: This case offers a good example of anchoring bias. Knowing that the patient had recently lifted a boat, the doctor honed in on muscle strain as the likely cause of the patient’s pain. The negative results from the ECG reinforced the narrow diagnostic focus. As a result, the doctor failed to order further testing and prematurely terminated the data-gathering process. Further investigation of the patient’s history would have revealed that the patient was a heavy smoker and drinker. He also had a family history of cardiovascular disease and both his father and grandfather died in their early 40s. An affective influence also might have been at play in this case; the doctor might have considered a cardiac condition less likely because of the patient’s young age.

More research is needed to determine effective approaches for addressing cognitive errors

Proposed solutions

Numerous solutions can be proposed to try prevent errors in clinical judgement, including implementing strategies to improve teamwork, adjusting processes and workflows, using diagnostic aids, and exploring debiasing techniques.  Other proposed solutions would be to emphasise clinical reasoning and decision-making in medical education, including a strong focus on heuristics and biases. Implementing and following clinical guidelines, checklists, and clinical pathways to support the reasoning and decision-making processes are also important. It remains to be seen if artificial intelligence will assist in supporting diagnostic decision-making and potentially reducing errors. Doctors can practise techniques such as perspective-taking, emotional regulation, and partnership-building, which can help reduce bias and promote empathy, humility, and patient-centred care. In reality, there are many obstacles to using these techniques in busy, under-resourced clinical environments.

The following list offers suggestions for managing these risks in both GP settings and hospitals, where possible:

Update and review patients’ medical histories, problem lists, medication lists, and allergy information regularly to ensure the records reflect their most recent information.

Always ‘safety net’ patients giving information on red flags, timelines, and actions to take. Remember to record the safety net advice.

Consider using a checklist or template to guide-taking each patient’s medical history and performing a thorough physical exam. This can help ensure consistency and prevent oversights.

Perform complete patient assessments, including establishing differential diagnoses, considering appropriate diagnostic testing (where feasible), and carefully reviewing test results.

Engage patients and their families in the diagnostic process.

Ensure awareness of clinical care pathways and determine how best to implement them into workflow patterns.  Incorporate a diagnostic review process into the workflow pattern.

Develop a written policy that outlines how disagreements in diagnosis and care among the diagnostic team will be managed, including the appropriate chain of command for escalating conflicts.

Formalise procedures peer reviews of diagnostic tests and imaging and the use of diagnostic guidelines. Ensure all relevant information is included in handovers, referrals, and transfers.

Be aware of common cognitive biases and affective influences and how they might negatively affect clinical judgement. It is important for doctors to also look after their own health and wellbeing ensuring they can confidently exercise clinical judgement. 

Consider using structured tools or approaches to identify the types of diagnostic errors occurring in the practice and the root cause of the errors. Use this information to educate the clinical team and develop countermeasures to improve quality-of-care.

Consider group educational opportunities that allow members of the diagnostic team to explore cognitive biases and develop solutions together.

Although diagnostic errors have many causes, clinical judgement is a significant contributing factor. The complex nature of clinical reasoning and decision-making makes it vulnerable to error, which can subconsciously lead to lapses in judgement and in turn can cause delayed or misdiagnoses, resulting in patient harm. More research is needed to determine effective approaches for addressing cognitive errors. However, various strategies – such as improving teamwork, increasing cognitive awareness, and using clinical decision support systems, clinical pathways, checklists, and debiasing techniques show promise. By considering how to implement these strategies in clinical tasks and processes, doctors can take proactive steps toward managing diagnostic risks.

This article has been adapted, with permission, by Ms Dee Duffy, from an article, written by Ms Laura M Casella, Medical Writer/Editor and Publications Lead, MedPro group.  The original article is available at: www.medpro.com/documents/10502/2820774/Article_Clinical+Judgment.pdf