Data-monitoring boards

Useful practical advice on organizing the running of data monitoring boards will be found in Ellenberg et al. (2003). 

If it is considered desirable and feasible to run a double-blind trial in a given indication, the fact that a sequential trial is being run with repeated looks at accumulating data has the capacity, as O Neill (1993) puts it, to influence parts of the trial conduct such as types of patients entered, definitions of endpoints, exclusion criteria etc. (p. 606). Sometimes the data-monitoring board is kept in the dark as to which treatment is which and results are presented to it without revealing the treatment labels. This is not always possible for example, where an asymmetric stopping rule is being used, as may be the case when trials can be stopped for lack of efficacy. 

Herson J (1993) Data monitoring boards in the pharmaceutical industry. Statistics in Medicine 12 555-561. 

DSMB, Data and Safety Monitoring Board IRB, Institutional Review Board NIH, National Institutes of Health. 

Fig. 2, Laboratory setup for combustion synthesis. 1-reaction chamber 2-sample 3-base 4-quartz window 5-tungsten coil 6-power supply 7-video camera 8-video cassette recorder 9-video monitor 10-computer with data acquisition board 11-thermocouple 12-vacuum pump ...

After a mean of 7 years of follow-up, the Data and Safety Monitoring Board also recommended stopping the oral estrogen-alone arm of the study. This arm consisted of 10,739 women who had undergone hysterectomy. Estrogen-only therapy had no effect on coronary heart disease risk and no increased breast cancer risk. ... 

Another potential use for these drugs was to prevent the formation of adenomatous polyps in patients with a history of colorectal adenomas. Rofecoxib and celecoxib were both tested in this disease. It was the results of this test that prompted Merck to withdraw their drug from the market since this study clearly established the increased risk of cardiovascular events when rofecoxib is used. The results were sufficiently strong to have the safety monitoring board reevaluate the data for a similar trial with celecoxib. On the basis of these data the study with celecoxib was also terminated. Interestingly, another study in which celecoxib was used to prevent polyp formation resulted in no increase in cardiovascular events. The only difference between the two studies was that in the latter case celecoxib was given once a day whereas in the prior study celecoxib was administered at the same dose but twice a day. How this difference in treatment schedules affected the toxic outcomes is unknown at the present time. 

However, just such a situation did arise on a trial I was involved in. This was the CARDS study (Colhoun et ah, 2004) of atorvastatin referred to in section 19.1. I was the statistician on the data-safety monitoring board and we called a halt to the... 

V ftttes J, Barrett-Connor E, Braunwald E et al. 2004. Monitoring the randomized trials of the Women s Health Initiative The experience of the Data and Safety Monitoring Board. Clinical Trials 4 218-234. 

As can be seen, in each case the lower limit of the two-sided 95 % confidence interval fell below unity (1.00) and the upper limit fell above unity, hence the nonsignificant result. The Data Monitoring Safety Board for the trial considered all interim analyses and recoimnended that the trial continue. 

---