Competing Risks
Susan Halabi, Stefan Michiels in Textbook of Clinical Trials in Oncology, 2019
It is worth noting that the Bonferroni method would always be statistically significant at level α = 0.05 if one of the individual tests has a p-value less than 0.025. However, even in this case, joint inference could provide further insight into the effects of variable or condition by examining the joint confidence region of a pair of hazard ratios of interest. For example, Figures 21.4 and 21.5 give 95% confidence regions for the CSH ratio and ACH ratio, and the CSH ratio and OCH ratio, based on their joint Wald test statistics. It is seen that for (CSH, ACH), the areas of confidence region from the chi-square test are much smaller than the Bonferroni and maximum tests. Furthermore, the chi-squared joint confidence region reveals that a higher CSH ratio is associated with a higher ACH ratio, as in the previous example.
Selected Statistical Topics of Regulatory Importance
Demissie Alemayehu, Birol Emir, Michael Gaffney in Interface between Regulation and Statistics in Drug Development, 2020
In association analyses involving many markers, one needs to control the possibility of false positives. Traditional approaches, such as the Bonferroni method tend to be too strict and may lead to many missed findings. The false discovery rate (FDR), defined as the expected proportion of incorrectly rejected null hypotheses among the declared significant results, was introduced by Benjamini and Hochberg (1995) as an attractive alternative to the more conservative traditional methods for simultaneous inference. Subsequent enhancements of the FDR include the positive false discovery rate (pFDR) and the q-value, which is a measure of significance in terms of the FDR rather than the usual false positive rate associated with traditional p-values (Storey and Tibshirani 2003).
A Brief Review of Multiple Testing Problems in Clinical Experiments
Albert Vexler, Alan D. Hutson, Xiwei Chen in Statistical Testing Strategies in the Health Sciences, 2017
Bretz et al. (2010) described the common underlying theory of multiple comparison procedures through numerous examples, including the Bonferroni method and Simes’s test. The book also depicts applications of parametric multiple comparisons in standard linear models and general parametric models, Dunnett’s test, Tukey’s all-pairwise comparison method, and general methods for multiple contrast tests for standard regression models, mixed-effects models, and parametric survival models. There is a detailed discussion and analysis regarding many multiple testing techniques, of which many examples are implemented in R.
Lessons learned from routine, targeted assessment of liquid biopsies for EGFR T790M resistance mutation in patients with EGFR mutant lung cancers
Published in Acta Oncologica, 2019
Sebastian Mondaca, Michael Offin, Laetitia Borsu, Mackenzie Myers, Sowmya Josyula, Alex Makhnin, Ronglai Shen, Gregory J. Riely, Charles M. Rudin, Marc Ladanyi, Helena A. Yu, Bob T. Li, Maria E. Arcila
Fisher exact test was used to identify significant associations between categorical variables. Time to treatment discontinuation (TTD) was used as a surrogate for progression free survival (PFS) [22,23]. TTD was defined as the time from start of EGFR-TKI to last dose administered. TTD and overall survival (OS) were estimated from treatment start date using the Kaplan–Meier method. Log-rank test was used to evaluate for associations between categorical variables and survival. Patients without complete survival data were censored at date of last follow-up. For patients with both plasma ctDNA testing and tissue testing for EGFR T790M, concordance along with an exact 95% confidence interval was evaluated. All statistical tests were two-sided, and a p value of less than .05 was considered statistically significant. The Bonferroni method was used to adjust for multiple comparisons. Statistical software (SAS version 9.3) was used for statistical analyses.
Analysis of small bowel angioectasia in asymptomatic individuals depending on patients’ age and gender
Published in Scandinavian Journal of Gastroenterology, 2019
Li-Hong Teng, Tao Yang, Jia-Wei Lu, Wei-Li Liu
We divided the age into three groups as follows for comparation: elderly ( > 60 years), middle-aged (40–59 years) and young adults (<40 years). At the same time, we also evaluated the differences between different gender groups. All the data analysis was performed with SPSS software (version 20, IBM, Chicago, Illinois, USA). Continuous variables were presented as means and standard deviations and were compared by independent t-test. Qualitative statistics were manifested as percentages. Statistical description and chi-square tests were suitable for analysing the prevalence and particular relationships of some positive findings. Logistic regression analysis was performed to identify some relevant factors of angioectasia and P2 lesions after regulating for age group, gender and bowel preparation. Odds ratio (OR) and 95% confidence interval (CI) were reported for each variable. The difference was statistically significant when p<.05. In this study, Bonferroni method was used to adjust P values of multiple comparisons.
Effects of a brief problem-solving intervention for parents of children with cancer
Published in Children's Health Care, 2018
Jennifer Lamanna, Matthew Bitsko, Marilyn Stern
Analyses of Variance (ANOVAs) were conducted to examine the effects of the intervention on PTSS, caregiving stress, and problem-solving ability. Outcome data were analyzed separately based on time since diagnosis and psychosocial risk. In total, 12 separate analyses were conducted on these data. In order to reduce the likelihood of Type I error, the Bonferroni method was used to correct for multiple comparisons and the significance value was set at 0.004 (0.05 ÷ 12). No significant differences were found for condition (between-subjects effects), time (within-subjects effects), or for the interaction, indicating that neither the intervention nor the passage of time had a significant effect on the outcome variables. All effect sizes were below Cohen’s standard of 0.10 for a small effect size (1988).
Related Knowledge Centers
- Bonferroni Correction
- Type I & Type II Errors
- Multiple Comparisons Problem
- Family-Wise Error Rate
- Uniformly Most Powerful Test
- Type I & Type II Errors
- False Positives & False Negatives
- P-Value
- Statistical Significance
- Closed Testing Procedure