Syntax for Science publishes its first post in collaboration with Pedro J. Llabrés, specialist in scientific dissemination and author of the blog Huele a Química. This is the first of a series of articles related to clinical studies. Here, the main tools to avoid bias in clinical studies are presented.
Control Group, Randomization and Blinding, Better Together
Pedro J. Llabrés, César Garcia-Rey, Juan V. Torres
In this post, the authors present three methodological aspects that have made clinical trials the most popular and accepted tool (for not saying the only one) to proof that a new treatment, device or intervention works.
A clinical trial is a prospective experiment in which the safety and efficacy of a new intervention is evaluated. The main enemy of clinical trials is bias. A biased study will provide results that cannot be reproduced and do not represent the effect of our intervention in standard conditions. Bias can be such a big problem that all our efforts to design a study, collect, validate and analyze the data might become totally useless. Bias can be minimized with a good study design, i.e., a study with a control group, randomized and blinded.