Improving Your Critical Appraisal Skills #1

The Randomized Control Trial

Why start this series with the RCT? RCTs are used to determine if one treatment is better than another. They are clinical experiments that provide medicine’s most reliable data. When there is a good RCT, one that can be generalized to a specific patient’s situation, there is no better guide. However, RCTs are far from perfect. Individually, their design may lead them to include important biases and not be generalizable. More generally, because of publication bias, the literature of RCTs is also biased. The conclusions that come from RCTs concern populations while our decisions, either as patients or doctors, concern individuals. It hardly needs to be stated that RCTs (let alone unbiased, generalizable RCTs) are available for the minority of decisions.

Design of Randomized Controlled Trials