Towards a conceptual model of affective predictions in palliative care

Abstract Context Being diagnosed with cancer often forces patients and families to make difficult medical decisions. How patients think they and others will feel in the future, termed affective predictions, may influence these decisions. These affective predictions are often biased, which may contribute to suboptimal care outcomes by influencing decisions related to palliative care and advance care planning. Objectives This study aimed to translate perspectives from the decision sciences to inform future research about when and how affective predictions may influence decisions about palliative care and advance care planning. Methods A systematic search of two databases to evaluate the extent to which affective predictions have been examined in the palliative care and advance care planning context yielded 35 relevant articles. Over half utilized qualitative methodologies ( n  = 21). Most studies were conducted in the U.S. ( n  = 12), Canada ( n  = 7), or European countries ( n  = 10). Study contexts included end of life ( n  = 10), early treatment decisions ( n  = 10), pain and symptom management ( n  = 7), and patient-provider communication ( n  = 6). The affective processes of patients ( n  = 20), caregivers ( n  = 16), and/or providers ( n  = 12) were examined. Results Three features of the palliative care and advance care planning context may contribute to biased affective predictions: 1) early treatment decisions are made under heightened emotional states and with insufficient information; 2) palliative care decisions influence life domains beyond physical health; and 3) palliative care decisions involve multiple people. Conclusion Biases in affective predictions may serve as a barrier to optimal palliative care delivery. Predictions are complicated by intense emotions, inadequate prognostic information, involvement of many individuals, and canceru0027s effect on non–health life domains. Applying decision science frameworks may generate insights about affective predictions that can be harnessed to solve challenges associated with optimal delivery of palliative care.