Symptom Clusters and Prognosis in Advanced Cancer (414-B)

Abstract

1.Discuss the prognostic impact of symptom clusters (SCs) on survival in advanced cancer.2.Evaluate the association between known SCs and subsequent survival. Background. The prognostic impact of symptom clusters (SCs) on survival in advanced cancer is unknown. Research objectives. We evaluated the association between known SCs and subsequent survival. Methods. We did a post-hoc analysis of previously reported data which identified seven SCs. A cluster was considered present if any patient had ≥ 50% of the symptoms in that cluster. Survival was measured from the time of referral to death. Among 831 patients with survival data, Cox proportional hazards and Stepwise Cox analyses assessed the prognostic effect of SCs on survival. Kaplan-Meier survival curves were constructed for three SCs that were significant in multivariable analysis. Results. Three SCs were associated with shorter survival in both univariable and multivariable analysis (p < .03 in all cases): the fatigue/anorexia-cachexia (easy fatigue, lack of energy, weakness, dry mouth, anorexia, early satiety, taste change, weight loss), the debility (edema, confusion), and the aero-digestive (cough, dyspnea, hoarseness, dysphagia) clusters. The hazard ratios (95% CI) from the multivariable analysis were 1.4 (1.2-1.6), 1.24 (1.1-1.4), and 1.2 (1.0-1.4), respectively. This suggested that those who had these SCs had a 17% to 37% increased risk of death compared to those who did not. Among persons with none of the three clusters, median survival was 3.4 (2.5-4.4) months but 1.7 (1.6-1.9) months among people with 1-2 clusters, and 1.0 (0.8-1.3) month among those with all three. Conclusion. Three SCs had a statistically and clinically important negative impact on survival in advanced cancer. Implications for research, policy, or practice. SCs appear important not only to clinical assessment, and therapeutic management, but may also improve prognostication. Physical Aspects of Care