Jump to content

Property:Specifications on analyses

From CAMIH

This is a property of type Text.

Showing 20 pages using this property.
B
C
3 weeks of receiving placebo/intervention, 1 week washout, 3 weeks of receiving intervention/placebo Consort diagram shows, that drop-out was excluded from final analysis, even though an ITT-analysis was mentioned in the method section  +
Baseline characteristics were compared among the three study arms using analysis of variance methods (ANOVA) for continuous variables and chi-square tests for categorical variables. To test the two primary hypotheses to detect a decrease in the mean changes in IGF-1 (or COX-2) gene expression between the lycopene (or fish oil) and placebo arms a t statistic was used with significance set at a probability <0.025 to adjust for the two comparisons. The same method was used to test for a decrease in IGF-1R with the lycopene supplement without any adjustment for multiple testing. Fisher’s exact test to determine whether a greater proportion of patients on the supplement arm achieved at least a 2-fold decrease in IGF-1 (or COX-2) expression when compared with the placebo group was used. An exploratory analyses was conducted using 2-way ANOVA methods to investigate the change in gene expression (pre to post-intervention) in DCT on the log2 scale due to the study arm (supplement or placebo), baseline tomato/fish intake (high or low) or their interaction. If statistical significance was observed, the Newman-Keuls post hoc test was used to identify which subsets were significantly different. Mean baseline and change from baseline results were presented on the log2 scale.  +
Analysis of covariance (ANCOVA) was used to assess between-group differences. Baseline values, chemotherapy status, and/or change in fat mass were used as covariates where appropriate. To assess differences in perceived energy between diet groups, sub-group analyses were conducted on the basis of chemotherapy status due to the prevalence of chemotherapy-related fatigue. Paired t-tests were used to assess within-group differences. Pearson correlation analyses were used to estimate associations between questionnaire results and serum β-hydroxybutyrate concentration, a biomarker of circulating ketones. The statistical analysis plan included an investigation for outliers, and the final analyses excluded them in order to lessen the influence of extreme observations on the overall results. One outlier (>3 standard deviations above the mean) was excluded from food craving analyses involving starches and high-fat foods. In addition, two participants in the KD group only partially completed the SF-12 at the follow-up visit; two participants in the KD group also did not complete all items on the VAS. Accordingly, these participants’ responses have been excluded from the corresponding analyses.  +
ITT-Analyse n=29 Per Protocol Analyse n= 27 An exploratory analysis was performed in which the data (all 17 patients who started on L-carnitine plus 10 patients who started on placebo, inkluding two protocol violators) was supplemented by the scores recorded at the end of the open-label phase. The pattern mixture analysis used three measurements over time and accounted for missing data.  +
Repeated measures analyses of variance (ANOVA), estimated with a linear mixed model, enable us to test the effect of time and treatment on continuous outcomes, while a generalized linear mixed model is used as a logistic regression for dichotomous outcomes. The P-values that are presented are for the interaction term of these models. All the analyses were also carried out while adjusting for site, treatment, and tumor size.  +
D
Unclear when the two attritions took place, as according to the tables all were included in the analysis.  +