Editors' ChoiceBiostatistics

Bio-Creep in Clinical Trials

See allHide authors and affiliations

Science Translational Medicine  08 Dec 2010:
Vol. 2, Issue 61, pp. 61ec191
DOI: 10.1126/scitranslmed.3001991

Is "bio-creep" a problem in your clinical trial? Clinical trial designs vary from traditional superiority trials, in which the goal is to show that a new treatment works better—is more efficacious—than the existing treatment, to noninferiority trials, in which the goal is to show that a new treatment is at least equivalent to an existing treatment. Equivalency is defined as the difference between two treatments being within an acceptable margin, which might include ranges in which the new treatment is slightly less efficacious than the existing treatments. Thus, a new, slightly less efficacious treatment might become a new standard of care with which compare other new treatments. After multiple trials in which this occurs, it is possible for new treatments to be statistically equivalent to existing treatments but not efficacious when compared with a placebo—a phenomenon known as bio-creep. Until now, it has been unclear whether bio-creep is merely a theoretical concept and rare in practice, or something that should be actively considered in most noninferiority trials.

Everson-Stewart and Emerson performed a simulation study to investigate several factors that might influence the probability that bio-creep occurs, including the distribution of the new treatment’s effect, the influence of study-to-study differences in the treatment effects, and trial design parameters such as choice of the existing treatment (for example, using the most effective existing treatment as the comparison treatment) and the measure of variation of the existing treatment effect. They found that bio-creep was overall a rare occurrence when appropriate noninferiority analyses were performed and the effect of the existing treatment was constant over time. Bio-creep was more prevalent when the effect of the existing therapy was changing over time because of changes in the populations being studied. In these cases, ineffective treatments were more likely to be declared effective when there were many existing treatments and a long chain of studies had been performed, with the estimated effect of the existing treatment coming from the primary clinical trial that established efficacy.

S. Everson-Stewart, S. S. Emerson, Bio-creep in non-inferiority clinical trials. Stat. Med. 29, 2769–2780 (2010). [Abstract]

Navigate This Article