Question

URGENT HELP WITH THESE 3 ARTICLES (A, B, C)!!!

A. Goodman SN. Rashoman revisited: two views of monitoring the Women’s Health Initiativetrials. Clin Trials 2007;(4)205-206.

B. Wittes J, Barrett-Connor E, Braunwald E, Chesney M, et al. Monitoring the randomized trialsof the Women’s Health Initiative: the experience of the data and safety monitoring board. Clin Trials 2007;(4)218-234.

C. Anderson GL, Kooperberg C, Geller N, Rossouw JE, et al. Monitoring and reporting of theWomen’s Health Initiative randomized hormone therapy trials. Clin Trials 2007;(4)207-217.

1. In general, what are the 3 courses of action with regard to the conduct of the trial that a DSMB can recommend upon reviewing interim results?

2. Why did the DSMB in collaboration with the CCC choose the global index measure to monitor the outcomes evaluated during the course of the trial? Please explain the ethical as well as methodological issues that guided this decision.

3. After reading the section titled ‘Ending the PERT study’ (Wittes et al., pgs 227 – 228), tell me what YOUR thoughts are about the DSMB’s decision to end the study in May 2002 rather than in November 2001.

Answer 1

Question no:1

1. 3 technical issues were raised in this regard

1. Un weighted and weighted test statistics
2. Acknowledging multiple comparisons
3. Ignoring the asymmetry in risk and benefit evaluation

2. Un weighted and weighted test statistics:

1. The protocol defines the specified weighted log rank statistics
2. Un weighted COX regression analysis was utilized.
3. Un weighted analysis was particularly preferred for reporting of the CHD result
4. It does not rely on the assumption of lag time to full prevention effect.
5. Same methodology was used for reporting for all outcomes, even though this provided a more conservative statistic compared to the corresponding weighed test.
6. The weighted test significance level was also presented but only in the description of the trial termination.

3. Acknowledging multiple comparisons:

1. multiple comparison adjusted confidence intervals
2. discouraging over-interpretation
3. Width adjustments were made by an appropriate transformation of the monitoring boundaries.
4. involvement of early stopping decisions of trials involving
5. It plays a minor role in the early stopping decision so that the nominal confidence interval at that time can be expected to be fairly accurate.

4. Ignoring the symmetry in risk and benefit evaluation:

• it requires stronger statistical evidence of benefit
• it incorporates the asymmetry into confidence intervals

                                                                                                                                                                                                               

Question no:2

1. global index was used in clinical outcomes subject for the formal monitoring in the WHI hormone therapy trial component
2. overall benefit was defined as the global index comparison
3. global index tool was not intended to be used as a standalone monitoring tool
4. global index introduced additional conservatism into the monitoring boundaries
5. It organizes the guidelines for the conservatism.
6. Concept of a global index of risks and benefits was one particularly useful aspect of the monitoring plan.