Factors associated with distinct prognostic‐awareness‐transition patterns over cancer patients’ last 6 months of life

Abstract Background Cancer patients may develop prognostic awareness (PA) heterogeneously, but predictors of PA‐transition patterns have never been studied. We aimed to identify transition patterns of PA and their associated factors during cancer patients’ last 6 months. Methods For this secondary‐analysis study, PA was assessed among 334 cancer patients when they were first diagnosed as terminally ill and monthly till they died. PA was categorized into four states: (a) unknown and not wanting to know; (b) unknown but wanting to know; (c) inaccurate awareness; and (d) accurate awareness. The first and last PA states estimated by hidden Markov modeling were examined to identify their change patterns. Factors associated with distinct PA‐transition patterns were determined by multinomial logistic regressions focused on modifiable time‐varying variables assessed in the wave before the last PA assessment to ensure a clear time sequence for associating with PA‐transition patterns. Results Four PA‐transition patterns were identified: maintaining accurate PA (56.3%), gaining accurate PA (20.4%), heterogeneous PA (7.8%), and still avoiding PA (15.6%). Reported physician‐prognostic disclosure increased the likelihood of belonging to the maintaining‐accurate‐PA group than to other groups. Greater symptom distress predisposed patients to be in the still‐avoiding‐PA than the heterogeneous PA group. Patients with higher functional dependence and more anxiety/depressive symptoms were more and less likely to be in the heterogeneous PA group and in the still‐avoiding‐PA group, respectively, than in the maintaining‐ and gaining‐accurate PA groups. Conclusions Cancer patients heterogeneously experienced PA‐transition patterns over their last 6 months. Physicians’ prognostic disclosure, and patients’ symptom distress, functional dependence, and anxiety/depressive symptoms, all modifiable by high‐quality end‐of‐life care, were associated with distinct PA‐transition patterns.

Abstract licence: CC BY

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.