The role of technology in avoiding bias in the design and execution of clinical trials

Correspondence: Hazel Goodale Perceptive Informatics Inc, Lady Bay House, Meadow Grove, Nottingham, NG2 3HF, United Kingdom Tel +44 115 844 3962 Fax +44 115 955 7555 Email hazel.goodale@perceptive.com Abstract: There are many documented instances in which bias has had an adverse effect on the results of clinical trials. This has led to a number of design techniques being developed that can be implemented in clinical trials in order to reduce bias. Sources of bias referring to published case studies are reviewed and discussed. The potential uses of technology to alleviate bias are outlined, particularly the use of centralized interactive response systems to randomize patients and manage medication in such a way as to limit the risk of bias caused by knowledge of either a patient’s current treatment or future treatment assignments. Potential sources of bias include selection bias, accidental bias, assessment bias, observer bias, and operational bias. These can arise through inadequate randomization and concealment methods during the trial. The blind may be broken by individual code breaks or through deduction in studies with frequent dose adjustments; there is scope for deduction in adaptive trials that might also introduce bias. Technology can reduce or eliminate the potential for bias in a variety of manners including central randomization and secure methods to protect the blinding and trial integrity. However, if the separation of randomization and dispensing, made possible by the use of technology, is not applied correctly then new unblinding scenarios can be introduced.