Patients with any major illness can expect to experience uncertainty about the nature of their illness, its treatment and their prognosis. Prognostic uncertainty is a particular source of patient distress among those living with life-limiting disease. Uncertainty also affects professionals and it has been argued that the level of professional tolerance of uncertainty can affect levels of investigation as well as healthcare resource use. We know that the way in which uncertainty is recognised, managed and communicated can have important impacts on patients' treatment and quality of life. Current approaches to uncertainty in life-limiting illness include the use of care bundles and approaches that focus on communication and education. The experience in communicating in difficult situations that specialist palliative care professionals can provide may also be of benefit for patients with life-limiting illness in the context of uncertainty. While there are a number of promising approaches to uncertainty, as yet few interventions targeted at recognising and addressing uncertainty have been fully evaluated and further research is needed in this area.
- EDUCATION & TRAINING (see Medical Education & Training)
- INTERNAL MEDICINE
Statistics from Altmetric.com
William Osler is quoted as saying medicine is the science of uncertainty.1 Following this, uncertainty has and will probably always remain a key aspect of medicine. Despite advances in medicine over the past century, uncertainty prevails alongside illness. As diagnostic tests, treatments and the healthcare system itself become ever more complex, patients, with any major illness, can expect to experience a great deal of uncertainty about the nature of their illness, its treatment and their prognosis.
Clinicians too experience uncertainty every day. A major part of the doctor's role is to assess and manage uncertainty in terms of the risk/benefit of deciding to operate on a patient, the value of investigating and then managing a symptom in a patient with limited prognosis or in responding to patients' difficult questions.
All of these uncertainties can be present in life-limiting illness. However, prognostic uncertainty is an additional source of patient and professional distress among those living with life-limiting disease. The age-old question ‘how long have I got, doctor?’ is one that many still struggle to answer. This article therefore aims to examine in detail the concept of uncertainty in illness and its impact on patients and professionals. We then appraise current practice and evidence in addressing uncertainty in life-limiting illness and suggest an ideal approach.
What is uncertainty in illness?
Uncertainty is not a simple or easily defined concept and a situation of uncertainty usually results from several inter-related factors. Mishel was one of the first researchers to develop an overarching theory of uncertainty in illness which aimed to explain the underlying processes governing patients' experiences of uncertainty.2 She identified four concepts that contributed to an uncertain state—complexity, unpredictability, ambiguity and lack of information.2–4 In her concept analysis of uncertainty in illness, McCormick further developed these ideas and described situations of uncertainty in terms of the probability of events occurring, the temporality of events and individuals' perceptions of their situation5 (see figure 1).
Any given situation of uncertainty may be made up of several of these factors, but equally, there are many forms of uncertainty in healthcare.6 In her discussion of uncertainty as applied to the primary care consultation, Greenhalgh defined four broad areas: uncertainty regarding the illness itself and the evidence base; uncertainty about a patient's story; uncertainty about what best to do for a patient and the uncertainty of multiprofessional collaborative care.7 A key point to add to this taxonomy is that patients and clinicians may understand and appraise uncertainty differently—some uncertainties are patient driven and others are professional.
Research contributions have considered healthcare professional responses to uncertainty over a 50-year timescale.8 ,9 Earlier research focused on the ways in which professionals ‘cover’ uncertainty to present a definite plan. For example, one paper noted that ‘physicians to be’ were ‘trained for control’ rather than trained for uncertainty.10 More recent research has examined healthcare professionals' engagement with uncertainty and the impacts of uncertainty on professionals and patients11 (see also later section).
The additional uncertainties of advanced life-limiting illness
At all stages of life-limiting illness, the difficulty in predicting prognosis represents an additional source of uncertainty.13 A growing body of research has been undertaken into the illness trajectory and prognostication in life-limiting illness, yet some degree of prognostic uncertainty usually remains, even when life expectancy is short.14 ,15
Despite this, we know a lot about prognosis and illness trajectories. We know that the trajectory of decline in terminal cancer is different to that of organ failure, different to renal failure and different again to frailty and dementia.16–18 We can also identify poor prognostic features of many illnesses, and in some situations, very well-validated scores (such as the Acute Physiology And Chronic Health Evaluation (APACHE) 2 score19) exist to help predict outcome. But the task of prognostication has its limits and analysis has demonstrated that a proportion of deaths are sudden and unexpected.20 Prognostication remains difficult in the last 6 months of life: A review of prognostic indicators found that they have insufficient sensitivity or specificity when predicting prognoses shorter than 6 months in non-cancer.21 ,22 A study of 149 deaths on medical wards in an acute hospital identified that 38% of deaths were not expected on admission—they either occurred suddenly or following an unpredictable dying trajectory.23 Another study of prognostication by expert clinicians observed only one out of every five prognostic estimates was within 33% of actual survival (n=468).13 ,24 While prognostication may be refined further in future, it remains to some extent an art rather than a science,25 and prognostic uncertainty is likely to remain a common feature in advanced illness.
Some early approaches in the uncertainty literature argued that the clinician's role is to reduce or remove uncertainty,8 ,10 ,26 ,27 but it is now considered that as an abstract concept, uncertainty is neutral cognitive state.28 ,29 It is rather the appraisal of uncertainty that can lead to its protective or harmful effects. Mishel acknowledged that uncertainty can be experienced positively as ‘opportunity’ or negatively as ‘danger’, but also suggested that the role of ‘structure providers’, including clinicians, is to reduce uncertainty4 (see figure 2).
As we have seen, elimination of uncertainty in advanced illness is unlikely to be achievable. It is also known that some patients hold on to uncertainty in a protective manner to shield themselves from a poor prognosis30 and that overexpression of uncertainty can lead to poorer decision satisfaction for patients.31 Knowing this, it is clear that any approach to uncertainty needs to be nuanced and patient specific, particularly because the level of uncertainty experienced in advanced illness is likely to fluctuate over time.11 ,32 ,33
Why does uncertainty matter? How does it impact on patients, families and professionals?
Uncertainty matters because it affects patients with life-limiting illness, their family and carers at a profoundly emotional level.32 Qualitative studies of patient experiences in a wide range of illnesses, including cancer, heart failure, chronic lung disease, multimorbidity and hospital inpatients, have elucidated that uncertainty significantly impacts on patients' lives.34–39 Patients who are uncertain about their futures can become preoccupied with this uncertainty and sometimes overwhelmed to the extent that their sense of self is impacted.40 A majority of patients wish to discuss this uncertainty.41 ,42 Research suggests this rarely happens.43
It has been shown that if uncertainty is not appropriately addressed, this can result in worse psychological outcomes for patients.44–46 Uncertainty is linked to anxiety in its association with fear of future.44 ,47 There are additional potential impacts on the patient–professional relationship if uncertainty is imperfectly expressed; Politi et al31 found that greater communication of clinician uncertainty during decision making about breast cancer treatment led to poorer decision satisfaction for patients. Blanch et al48 found that medical students who make more expressions of uncertainty during their interactions with patients are less well regarded.
Uncertainty also impacts on clinician's practice and their confidence; clinicians frequently struggle with uncertainty which can result in overtreatment or overinvestigation,49 increased costs45 ,50 and lack of communication with patients about their future.51 ,52 Some have gone further than this, arguing that intolerance of uncertainty is leading to a culture of chronic disease where every abnormality is classified as pathology.53 Further, it could be argued that training for clinicians to manage uncertainty is lacking—less than 20% of current UK postgraduate medical training curricula contain detailed recommendations and curriculum goals relevant to dealing with uncertainty.54
It is not possible to eliminate prognostic or other uncertainty in life-limiting illness, and so a paradigm focused purely on uncertainty reduction is inappropriate.32 But as we have shown, unrecognised or poorly addressed uncertainty can lead to negative patient and healthcare outcomes. It is therefore paramount to appropriately address uncertainty in patients with life-limiting illness. The complexity of the concept means that an individualised and nuanced approach is likely to be most beneficial.55
Addressing uncertainty: current strategies
There is a range of current practice in addressing uncertainty in life-limiting illness, but three main approaches are evident.
Disengagement or minimisation of uncertainty—evidence indicates that uncertainty is commonly either unrecognised or recognised and unaddressed.51 Systematic reviews of communication in chronic obstructive pulmonary disease (COPD) and heart failure observed that discussions of end-of-life care (EOLC) and prognosis rarely occurred.41 ,56 Multiple barriers to discussion of prognosis in advanced illness are evident, including lack of time, lack of expertise, fear of taking away hope by removing protective uncertainty, disagreement within teams about the level of disclosure41 ,42 and perhaps this accounts for why in many settings uncertainty remains underserved.
The care bundle approach—in the hospital setting, tools such as the Assessment, Management, Best practice and Engagement when Recovery is uncertain (AMBER) care bundle or the Psychosocial Assessment and Communication Evaluation (PACE) tool have been developed with the intention to aid recognition and management of clinically uncertainty situations.57 ,58 The AMBER care bundle was developed for patients who are deteriorating, with uncertain recovery and who are at risk of dying in the next 1–2 months.57 It comprises an identification phase, followed by the development of a clear medical plan for patients identified as fulfilling the criteria in conjunction with the patient and their family. The individual patient-centred plan is then reviewed daily. The care bundle, which includes roles for the entire medical team emphasises clear communication and planning. This approach is promising and recent research suggests that there may be some benefits; awareness of prognosis appeared to be higher among patients supported by the AMBER care bundle.35 However, a full evaluation of the efficacy of this approach is needed. Moreover, a recent single-centre study identified that rather than being used as a tool to identify patients with an uncertain recovery, the care bundle was principally used when it became certain that patients would not recover.23
Education/communication approach—communication is key to all patient interactions and this is no different when addressing uncertainty. Approaches thus far have focused either on patient education or on training of healthcare professionals. Cognitive behavioural therapy-based interventions directed at patients to improve their resilience and ability to cope with uncertainty have shown some success.59 ,60 Related are patient activation interventions aimed at encouraging patients to engage more closely with long-term illness, though there is less evidence in advanced or life-limiting disease.61 Others have suggested the use of mindfulness-based practice, a way of self-reflection in the present moment,62 as a possible helpful tool in changing how uncertainty is appraised by patients, though as yet little empirical evidence supports the use of mindfulness in this context.44 ,63
Education of health professionals to cope better with and communicate more effectively the uncertainties of advanced illness is the other approach. This is consistent with the suggestion that all healthcare professionals dealing with patients with advanced life-limiting disease are providing ‘generalist palliative care’, while specialist palliative care providers have a role in supporting other professionals in providing this care.64 Courses such as the Sage and Thyme approach65 or Transforming End of Life Care66 provide multidisciplinary training for non-specialist staff in communication, holding difficult conversations and dealing with clinical uncertainty in life-limiting illness. Other disease-specific training programmes focused on communication have shown to aid discussion of prognosis and preferences for EOLC in COPD67 and to alter interview style for physicians.68 These programmes may empower health professionals to recognise and address uncertainty in life-limiting illness. All warrant evaluative scrutiny.
Addressing uncertainty—the way forward?
While several different approaches exist to manage uncertainty in life-limiting illness, there is not sufficient evidence to identify a ‘best method’. There is evidence that if addressed poorly or unaddressed, uncertainty is detrimental to patient and family care, but there is little evidence supporting the use of any of the approaches outlined above, and no comparative studies exist. Further research and evaluation of these approaches are needed. In the meantime, there are some principles which can guide our approach.
First, the way in which patients' understand and process uncertainty is key to its impact. In some situations, uncertainty may be maintained for its protective effect (eg, not seeking to find out about prognosis to avoid bad news), but in others it may have deleterious effects (eg, when waiting for a potentially life-threatening diagnosis). Any approach to uncertainty therefore needs to take an individual view based on the situation and a patient's experience. Until now, no approach has done so. Recent qualitative research suggests that patients' response to uncertainty in life-limiting illness may usually be explained by three main factors, and perhaps it is these that should form the basis of future uncertainty-addressing interventions: (1) level of engagement with illness and treatment, (2) information needs and preferences and (3) temporal focus (present or future).69
Second, uncertainty is experienced differently and at different times by patients and professionals. Patient–professional differences in experience and understanding of uncertainty can be expected to result in differences in agendas and increase the risk of patients' concerns going unaddressed (as has been noted in other settings70). Interventions addressing uncertainty by targeting this patient–professional gap might be expected to improve outcomes (figure 3).
Third, evidence indicates that careful discussion with all stakeholders is needed to ensure the best communication of prognosis and prognostic uncertainty, especially in complex situations.71 Poor communication was a major failing identified by the report into the use (and misuse) of the Liverpool Care Pathway.72 Research indicates that careful communication regarding prognosis can be beneficial in life-limiting illness.73 It is therefore reasonable to expect that the additional support and experience in communicating in difficult situations that specialist palliative care professionals can provide may lead to benefits for patients with life-limiting illness in the context of uncertainty. By prioritising clear communication and care planning, the palliative care approach is promising in providing support for patients with life-limiting illness experiencing uncertainty and is worth evaluating in this context.
Clinicians can and should not eliminate uncertainty in life-limiting illness. Nor is it their role to blindly discuss all uncertainties with every patient. But uncertainty, particularly prognostic uncertainty, requires greater thought and should be more frequently considered in order to achieve the best outcomes for patients with life-limiting illness and their families. Appropriate communication in these often complex situations is paramount. There is scope for care bundle-based or education-based interventions to provide benefit although these require rigorous evaluations. In addition, the palliative care approach promotes in-depth individual assessment of each patient's concerns, and this is potentially very helpful to the process of addressing uncertainty. Further research in this area is needed to evaluate the impact of existing approaches and those in development.
Uncertainty will always be present in life-limiting illness, and if poorly recognised and managed, can impact adversely on patient outcomes.
When managing uncertainty, professionals should consider what their patients' may prioritise in situations of uncertainty.
Interventions exist to assist in the management of uncertainty, but need further evaluation. Promotion of communication between patients and professionals, education interventions and integration of palliative care are promising approaches.
Current research questions
Do currently used interventions targeted at addressing uncertainty improve outcomes for patients?
What is the longer term impact of ‘protective uncertainty’ on patient outcomes and should this uncertainty ever be challenged?
Can the use of the palliative care approach itself affect outcomes in the context of uncertainty?
How can palliative care be integrated into care for patients with life-limiting illness and uncertain prognosis?
Mishel MH. The measurement of uncertainty in illness. Nur Res 1981;30:258–63.
McCormick KM. A concept analysis of uncertainty in illness. J Nurs Scholarsh 2002;34:127–31.
Lunney JR, Lynn J, Foley DJ, et al. Patterns of functional decline at the end of life. JAMA 2003;289:2387–92.
Barclay S, Momen N, Case-Upton S, et al. End-of-life care conversations with heart failure patients: A systematic literature review and narrative synthesis. BrJ Gen Pract 2011;61(582):e49–62.
Murtagh F. Can palliative care teams relieve some of the pressure on acute services? BMJ 2014;348:g3693.
Self assessment questions
Please answer true or false to the below.
Prognostic uncertainty should always be discussed with patients with life-limiting disease.
Care bundles have been shown to improve quality of life in situations of uncertainty.
Any approach to uncertainty needs to be patient specific and depends on patients' level of engagement with their illness.
Patients with chronic obstructive pulmonary disease (COPD) and heart failure have similar end-of-life disease trajectories.
Professionals' recognition and management of uncertainty has potentially large cost implications.
Contributors SNE contributed to the article concept, information gathering, drafting and approved the final draft. JK contributed to the article concept, drafting and approved the final draft.
Competing interests None declared.
Provenance and peer review Commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.