Improving communication of prognostic uncertainty in advanced illness: a realist review

Prognostic uncertainty is common in advanced illness, driven by multimorbidity, treatment complexities and unpredictable disease trajectories. When inadequately acknowledged or discussed, it can lead to patient and family distress and inhibit decision-making.We aimed to build a programme theory of how, why and for whom communication of prognostic uncertainty 'works' (or not) in advanced illness and palliative care contexts.

Realist systematic review, reported according to RAMESES criteria.

Embase, MEDLINE, CINAHL, Cochrane Database of Systematic Reviews, SCI via Web of Science, PsycINFO, SSCI and multiple grey literature sources. We included evidence examining communication of prognostic uncertainty in adults with advanced illnesses.

We coded papers qualitatively to identify Context-Mechanism-Outcome Configurations (CMOCs) relevant to prognostic uncertainty communication. Supported by clinical and lay stakeholder consultation, we refined these into a programme theory.

From 4634 search results, we included 205 studies with 46 579 participants across 11 countries. The final programme theory identified contexts of communication: socio-cultural norms and values, patient and family preferences, communication models and interventions, clinician uncertainty tolerance, training and experience. Mechanisms of communication were mediated by the care relationship, the timing and initiation of discussion, the content of communication and explicit versus implicit communication. Relevant outcomes were emotions, hope, distress, decision making and care planning. There was no single model of effective communication as this varied based on individual characteristics and preferences, but openness was usually viewed positively.

This programme theory provides a new understanding of the contexts and mechanisms, such as explicitness of discussion, that support effective communication of prognostic uncertainty across advanced illness contexts to reduce distress and aid decision making. Sociocultural factors are important, but they have the most limited supporting evidence, especially in low- and middle-income countries. The model presented here may support professionals in individualised and compassionate communication of prognostic uncertainty.