Managing ambiguity and uncertainty in clinical decision-making

Red flags

As paramedics gather a history and assess a patient, they should be alert to red flags. Red flags are clinical signs or symptoms that should alert them to the potential presence of a significant disease process (Maselli et al, 2022).

Although publications and resources can vary in their consideration of what constitutes a red flag (Verhagen et al, 2016), there is typically a consensus regarding common conditions such as low back pain, infections and headaches.

If a red flag is present, that should alert paramedics to obtain further assistance in how best to manage the patient even if they cannot establish what the diagnosis may be.

For example, if a patient has acute low back pain and they are found to have had recent unintended weight loss, this is a red flag that may indicate the presence of a malignancy. At that point, a diagnosis will not have been established, so there is still diagnostic ambiguity, but the attending clinicians know how to manage the patient next; this person needs to be referred under a 2-week rule pathway where possible or referred back to their GP.

Imminent threats to life

Another practical strategy to manage ambiguity, which is closely linked to establishing any red flags, is to consider whether the likelihood of a patient having a condition that may pose an imminent threat to life can be discounted.

For example, when a patient presents with an acute infection but is still passing urine regularly, is cognitively intact and has observations within normal parameters, the patient does not currently show indicators of a dysregulated immune response that would suggest sepsis (Cecconi et al, 2018). It does not mean that they will not go on to develop sepsis, so an element of ambiguity remains but, at that point, their risk can be considered low. Obviously, paramedics must consider any individual variances that may alter that risk, such as drugs or other pathologies that alter the effectiveness of the body’s immune response.

Conversely, if paramedics cannot obtain enough information to safely eliminate a potentially life-threatening event, then they need to pass the process up the line where more information can be gathered. In a patient with cardiac-sounding chest pain—even if they have no obvious cardiac risk factors, have a normal ECG and normal textbook observations—a significant cardiac pathology cannot be excluded without additional information from blood tests.

Differential diagnoses

Next, having a range of differential diagnoses to work from helps protect against cognitive biases. When people are uncertain, there can be a temptation to jump to an answer that is supported by a cognitive bias. Practitioners can mitigate against this by considering whether there is another explanation for the potential diagnosis they are considering.

Having a range of differential diagnoses is crucial to doing this. Paramedics can also voice aloud their reasoning to themselves, a colleague or even the patient and their family as this can highlight flaws in their reasoning (O’Sullivan and Schofield, 2018).

Determining appropriate differential diagnoses can be difficult. Sometimes, patients can present with identifiable, single pathologies, but when they have vague systemic symptoms that could arise from various body systems, it can be harder to determine the appropriate differential diagnoses (Randall and Feather, 2011). This is where practical resources help. Symptom Sorter (Hopcroft and Forte, 2020) is an excellent resource written by GPs and designed to provide a range of potential ‘common, occasional and rare’ causes for a wide range of symptoms. There are also various online resources such as For Medical Professionals (Patient, 2023), an area of a website that provides helpful lists of differential diagnoses for various conditions and presentations.

Ruling in and ruling out

Once paramedics have a range of differential diagnoses and are aware of any red flags, they can use two key processes alongside any tests: ruling in and ruling out problems (Danczak et al, 2016). These require clinicians to use signs, symptoms, tests and potentially trials of treatment (depending on the individual’s area of clinical practice) to either exclude or include a specific disease process.

The ruling-out process should be restricted and not involve using every test available to rule out the likelihood of a particular illness.

Thompson et al (2009) demonstrate this in the context of two case studies of febrile children, one of whom was found to have meningitis and one who did not. Not every child with symptoms of an upper respiratory tract infection can or should be subjected to invasive and traumatic testing for meningitis but, when paramedics are aware of key factors such as the frequency of meningitis and the course of the disease itself, restricted ruling-out can be used as an effective strategy in safely managing diagnostic uncertainty.

This approach of ruling in and ruling out also requires an awareness of the limitations that tests have. For example, where a 69-year-old woman is complaining of dysuria, frequency and urgency, a paramedic may assume that a urine dipstick will help rule in the presence of a urinary tract infection (UTI). Dipsticks alone, however, can be notoriously poor at ruling in the presence of a UTI in elderly patients because of the high prevalence of asymptomatic bacteriuria (Froom and Shimoni, 2018). It may well be that the patient has an alternative diagnosis such as urethral syndrome or atrophic urethritis with vaginal atrophy. In this situation, urine cultures are needed in addition to a dipstick to rule in or rule out a UTI.

Safety netting

Finally, the inherent risk of diagnostic ambiguity should be managed with safety netting (Jones et al, 2019). Safety netting is also a way of communicating uncertainty to patients. A useful mnemonic to guide safety netting is SAFER (Edwards et al, 2022):

The significance of potential risk should determine the threshold for safety netting. All safety netting should be specific and tailored to the individual patient in both the depth of information given and the format offered (Jones et al, 2014). Providing the patient with written information to support verbal safety netting is useful, particularly if there is a risk they have not taken all the information in (Greenhalgh et al, 2020).

If paramedics feel that there is sufficient uncertainty regarding the progression of the patient’s illness (i.e. they suspect there is a considerable risk of deterioration), they should give the patient an extremely low threshold to contact the emergency services again, to seek other help, or they should schedule a review of the patient with themselves or another clinician (clinical setting dependent).

Use the ‘Can you sleep tonight?’ test

Discomfort in the face of ambiguity should be embraced and used as a guide to what clinicians should do next (Ilgen et al, 2021).

When faced with two options, and one of those leaves a practitioner so uncomfortable that they suspect they may not sleep that night from worrying about the patient, that is probably a good indicator that it is not the optimal choice.

Leaving an 88-year-old at home with a chest infection, antibiotics from the GP and a relative who can check in on them later is vastly different from leaving a completely isolated 88-year-old with a chest infection at home having told them to contact the GP themselves the following morning.

Shared decision-making

Another means of effectively addressing management uncertainty is to use the principle of shared decision-making.

This begins with sharing decision-making with the patient themselves. Ironically, management uncertainty may arise through the use of the principle of shared decision-making. Paramedics can share what they feel is the optimal treatment plan but the patient may then introduce ambiguity by giving reasons why they do not feel that option is a good one. Whether practitioners feel there is an optimal solution that the patient discounts or whether there are several options, they should be involving patients in making decisions. This can be done by outlining the various risks and benefits of which they aware and by using statistics wherever possible in a way that the patient can understand and considering their preferences (Stiggelbout et al, 2012). This will often help reduce ambiguity as to what to do next.

There are usually options to extend shared decision-making beyond the immediate clinicianpatient relationship. Where possible, particularly when practitioners are faced with complex management decisions, decision-making should involve an interprofessional team (Légaré and Thompson-Leduc, 2014). It may be difficult at times to achieve this, particularly if working in an emergency setting outside normal hours but, even if shared decision-making cannot happen in real time, there may be options to request others’ input into the management plan at a later point. Depending on where they are working, paramedics may have direct access to peers or more highly trained clinicians to discuss options, and this will often help clarify what the optimal decision is.

Action versus inaction

Finally, paramedics should be aware that not doing something can sometimes be preferable to acting.

When clinicians are not sure what to do, they often act, even if there is a lack of evidence to support their action. This is a form of bias known as commissioning bias (Yuen et al, 2018). Unfortunately, this can have negative effects on the patient because unnecessary and unwarranted interventions can have negative effects on the patient.

Given the significant consequences of medical error and frequency of medication side-effects, paramedics should always consider whether a particular intervention is required.

Individual

Longer-term strategies can be implemented as individuals and collectively.

As individuals, paramedics need to focus on developing capability over competence. Healthcare education takes a competency-based approach (Englander et al, 2013), in which various skills, values, attitudes and knowledge are required to reach a particular professional standard. Competency alone, however, is not enough to assist practitioners in ambiguous situations. To manage this, capability is needed: this is the ability to apply competence in unique and flexible ways (Kaslow et al, 2022). This skill requires clinicians to become creative thinkers. By constantly challenging thinking, looking for analogies between different situations and even by developing positive emotions, thought processes can become increasingly adaptable and innovative (Isen, 2001; Munir and Awan, 2022).

Next is building personal resilience. Uncertainty has the potential to significantly increase a clinician’s level of stress (Iannello et al, 2017), and resilience can protect against this. Resilience interventions have shown some potential (Cleary et al, 2018), but practitioners do not have to rely upon employers or outside sources to provide these. There are numerous ways in which people can build their own resilience such as by improving their work-life balance as much as they can and by incorporating mindfulness practices into daily life (Brennan, 2017). Granted, these can be difficult to implement given the demands of working life, particularly in the NHS at this time, but even minor changes can have positive results.

Collective

Finally, the ability to manage ambiguity and uncertainty can be improved by collective actions, beginning with changing culture.

The historic emphasis upon individual blame and perfectionism within healthcare and medicine accentuate the fear of ambiguity and uncertainty because of the accompanying perceived risk of error and resultant consequences (Robertson and Long, 2018). Although lip service is paid to human fallibility, cultures often behave very differently, with a lack of self-compassion and openness (Shanafelt et al, 2019). Therefore, people need to work collectively to improve workplace culture.

Supervision needs to be a priority. Effective clinical supervision provides a safe space that provides time for reflection and review of clinical practice and unusual cases with an aim of improving practice (Johns, 2007; Afsar and Masood, 2018; Reynolds and Mortimore, 2021). Supervision is particularly important for individuals in advanced practice roles where ambiguity is more likely to arise at greater levels (Health Education England, 2021).

Research in the social work profession suggests that supervision can be useful in enabling individuals to manage uncertain, complex situations (Bourn and Hafford-Letchfield, 2011), so should be a priority for organisations wishing to equip their staff to manage ambiguity and uncertainty more effectively.