Thursday, July 20, 2006

IFA conference: RCTs

I have just started writing the article for my upcoming IFA presentation: "Lies, damned lies, and random control trials". The title might be a bit strong, but in today's world you need to get out strong message to get people's attention! The presentation is about the use of random control trials (RCTs) in stuttering. In the last year, two main treatment studies have used RCTs: the Pagoclone and the Lidcombe study. I want to show that RCTs cannot just be blindly applied to study the efficacy of treatment studies. This is especially the case for early-childhood dysfluencies.

Here is what I say. This is typically the result of months long subconscious brain storming and discussions with others.

First, I will explain what an RCT is. I have to admit that I have never been formally trained to know what an RCT is, so I have to make it up and looking at a few sources. What I understand as the standard form of RCT is the following: You have a group of people affected by a condition (high blood pressure, AIDS, worm infection, etc). You have a medication which you administer in form of a pill, and you want to see whether the treatment is effective. You split the group into two subgroups: a control group, and the treatment group. You have to do this in such a way as to create the same type of group; for example you select them randomly, and possibly control for the age or gender in each group. You give a pill to both groups, but only the treatment group receives the true medication and the control group a non-effective substance. The level of the condition is measured in both group before and after the test phase.

... I need to drive to the airport now .... still havent written the talk... i'll do it on my laptop... I will fly to Estonia first visiting a friend and doing a day visit to Russia! Then I am going to Dublin for IFA.

No comments: