Statistical Analysis

All data are expressed as mean ± standard deviation. We first undertook a bivariate relationship using linear regression analysis between pain intensity NRS (day 8) as the dependent variable and expectation of pain decrease NRS (day 1) and the following covariates as an independent variable: patients’ characteristics such as age, gender, PS (day 1), genotype of COMT, pain intensity NRS (day 1), and total scores of HADS (day 1). We also tested the association between pain intensity NRS (day 8) and cancer types (pancreatic cancer or not) because pancreatic cancer patients had significantly lower expectation of pain decrease NRS (day1) than the other cancer patients. Then, multiple linear regression analysis was performed using forced entry methods with pain intensity NRS (day 8) as the dependent variable and expectation of pain decrease NRS (day 1) and the covariates as independent variables that had a p value <0.05 in simple linear regression analysis, which means that the required level for p values was <0.00625 with Bonferroni correction for multiple tests in eight simple regression models. A two-sided significance level of 0.05 was used for multiple regression analysis. All statistical analyses were performed using SPSS software (version 19.0; SPSS Japan Inc., Tokyo).