Prognostic significance of symptoms of hospitalised advanced cancer patients

Purpose: To assess the prognostic value of symptoms in hospitalised advanced cancer patients. Patients and methods: A prospective analysis was performed of 181 hospitalised patients referred to a Palliative Care Team. Comprehensive symptom questionnaire, functional status, estimated life expectancy and survival were assessed. Using a Cox regression model, a predictive survival model was built. Results: Median survival: 53 d. Median number of symptoms: 4; 20 symptoms occurred in >= 10%. Multivariate analysis showed nausea, dysphagia, dyspnoea, confusion and absence of depressed mood as independent prognostic factors for survival (p < 0.05) with relative risks of dying of 1.96, 1.81, 1.79, 2.35 and 1.79, respectively. Patients with 2, 3 or 4 of these factors at the same time had a relative risk of dying of 2.7, 2.1 and 9.0, respectively. Conclusion: A cluster of factors comprising nausea, dysphagia, dyspnoea, confusion and absence of depressed mood may be used to accurately predict survival in hospitalised advanced cancer patients. (c) 2006 Elsevier Ltd. All rights reserved.