An excerpt from the item I just posted: 4.4.1.1
Post# of 154118
(Total Views: 684)
Posted On: 08/03/2020 6:35:25 PM
An excerpt from the item I just posted:
4.4.1.1. Monitoring for Effectiveness
In studies with serious outcomes, all parties would wish that any major
treatment advance be identified and made available as soon as possible. It is
critical, however, that the study yield a valid and definitive result. Thus,
tensions between ethical and scientific considerations may arise. Consider,
for example, a placebo-controlled trial of a new product for a serious illness or
condition for which there is no standard treatment. If the emerging data
suggest that those receiving the treatment are doing better, one might expect
that a DMC would consider whether the study should be terminated earlier
than planned. Estimates of treatment effect, however, will be unstable at early
points in a study, and the chance is substantial of observing a nominally
statistically significant benefit (e.g., p<0.05) at one of multiple interim
analyses during a study of an ineffective product (see Section 4.4.2). A DMC,
guided by a pre-specified statistical monitoring plan acceptable to both the
DMC and the study leadership, will generally be charged with recommending
early termination on the basis of a positive result only when the data are truly
compelling and the risk of a false positive conclusion is acceptably low.
A second type of consideration is whether the hypothesized benefit is likely
ultimately to be achieved. If the interim data suggest that the new product is
of no benefit—that is, there is no trend indicating superiority of the new
product—or that accrual rates are too low or noncompliance too great to
provide adequate power for identifying the specified benefit, a DMC may
consider whether continuation of the study is futile and may recommend early
termination on this basis. In this case, false negative conclusions are of
concern; statistical procedures are available to guide such determinations (see
Section 4.3.2).
4.4.1.1. Monitoring for Effectiveness
In studies with serious outcomes, all parties would wish that any major
treatment advance be identified and made available as soon as possible. It is
critical, however, that the study yield a valid and definitive result. Thus,
tensions between ethical and scientific considerations may arise. Consider,
for example, a placebo-controlled trial of a new product for a serious illness or
condition for which there is no standard treatment. If the emerging data
suggest that those receiving the treatment are doing better, one might expect
that a DMC would consider whether the study should be terminated earlier
than planned. Estimates of treatment effect, however, will be unstable at early
points in a study, and the chance is substantial of observing a nominally
statistically significant benefit (e.g., p<0.05) at one of multiple interim
analyses during a study of an ineffective product (see Section 4.4.2). A DMC,
guided by a pre-specified statistical monitoring plan acceptable to both the
DMC and the study leadership, will generally be charged with recommending
early termination on the basis of a positive result only when the data are truly
compelling and the risk of a false positive conclusion is acceptably low.
A second type of consideration is whether the hypothesized benefit is likely
ultimately to be achieved. If the interim data suggest that the new product is
of no benefit—that is, there is no trend indicating superiority of the new
product—or that accrual rates are too low or noncompliance too great to
provide adequate power for identifying the specified benefit, a DMC may
consider whether continuation of the study is futile and may recommend early
termination on this basis. In this case, false negative conclusions are of
concern; statistical procedures are available to guide such determinations (see
Section 4.3.2).
(2)
(0)