Objectives To improve the ability of clinical staff to recognise end of life in hospital inpatients dying as a result of cancer and heart failure, and to generate new hypotheses for further research.

Methods This mixed-methods study used decision theory as a theoretical basis. It involved a parallel databases-convergent design, incorporating findings from previously published research, with equal priority to study groups and synthesis by triangulation. The individual arms were (1) a retrospective cohort study of 102 patients with cancer and 81 patients with heart failure in an acute trust in the North of England, and(2) a semistructured interview study of 19 healthcare professionals caring for the same patient groups.

Results The synthesis of findings demonstrated areas of agreement, partial agreement, silence and dissonance when comparing the cohort findings with the interview findings. Trajectories of change are identified as associated with poor prognosis in both approaches, but based on different parameters. Management of patients has a significant impact on decision-making. The decision process requires repeated, iterative assessments and may benefit from a multidisciplinary approach. Uncertainty is a defining characteristic of the overall process, and objective parameters only have a limited role in predicting end of life.

Conclusions The role of uncertainty is important as a trigger for discussions and a defined stage in a patient’s illness journey. This is consistent with current approaches to recognising irreversible deterioration in those with serious illness. This study contributes ongoing evidence that these concepts are vital for decision-making.