The feasibility of patient accrual is evaluated at the same time. The recommended cluster size might be adjusted accordingly. Data analysis Due to the cluster structure, comparisons of different outcomes between treatment arms will be analyzed by mixed models. For endpoints with continuous values, linear mixed model may be applied. For endpoints with categorical or binary values, nonlinear mixed model or generalized estimating equations may be applied. The data will be stored and analyzed at the SAKK Coordinating Center using SAS software,

Version 9.2 of the SAS System for Windows (SAS Institute Inc., Cary, Inhibitors,research,lifescience,medical NC, USA) and the open source R statistical software package (http://www.r-project.org/). All statistical tests will be done two-sided at a significance level of 0.05. P-values will be corrected for multiple testing

where appropriate. Descriptive statistics will be done by median and range for continuous variables. Categorical data will be reported using absolute and relative frequencies. For the primary Inhibitors,research,lifescience,medical endpoint, selected influential Inhibitors,research,lifescience,medical variables (education, tumor type, predominant symptom, anxiety, complexity, hospitalisations) and the baseline G-QoL value will be included in the analysis model as covariates. For the primary analysis, only evaluable patients will be used. As a sensitivity analysis, non-evaluable patients will be included Inhibitors,research,lifescience,medical if possible. For instance, the difference between baseline and 3weeks will be analyzed including patients who are evaluable at week 3 but non-evaluable at week 6. Several pre-defined subgroup analyses are foreseen: The difference in G-QoL will be compared between both arms in sub-groups of patients having a) a tumor size response (SD, PR, CR) or not (PD), b) basic education or additional education c) one of the main tumor types defined as LY317615 research buy composing>= 20% of the evaluable Inhibitors,research,lifescience,medical study patients. d) a predominant symptom, if composing>= 20% of the evaluable study patients (expected based on symptom epidemiology data: pain, anorexia and/or fatigue [both through predominant vs. other symptoms or alone vs. other

symptoms], anxiety and/or depression [both predominant vs. other symptoms or alone vs. other symptoms], nausea, shortness of breath). e) anxiety <6/10 or>=6/10 f) complexity less than 3 symptoms above threshold vs. >=3 symptoms above threshold (fatigue and anorexia>=9/10, other symptoms>=6/10). All subgroup analyses will include baseline G-QoL as covariate. The study population will be described separately by institution (study center)-, oncologist-, and patient-related factors. The study center will be described with regards to actual procedures of symptom and syndrome assessment at the participating institution and local available interventions for multidimensional symptom and syndrome management.