Infinitely Meta – a 10,000-foot view of clinical trials


By Betsy Raymond Stevenson, posted October 1, 2010

By Sally Church

There has been quite a lot of noise in the media recently over the results of meta analyses of various clinical trials following the FDA committee meeting on Avandia in diabetes.

In modern medicine large scale studies are often limited in that they are designed to prove that a drug is generally or broadly safe and effective in order to gain approval in a wider population. We must remember that patients in the trials tend to be more motivated and have a better performance status than those in the general population, so extrapolating data from one to the other is fraught with difficulties.

One of the challenges is that sometimes a relatively rare (<5%) but fatal side effect may not show up in a short term trial, but will eventually emerge with more chronic use in a broader population. Of course, a relationship between two variables does not imply causation, therein lies the rub – how do we differentiate between spurious correlations and a real effect? Read on …

  Newer Entries »

RSHC copyright © 2009