Tag Archives: meta analysis

The antidepressant you are on is probably OK.

Posted on by

A little bit of evidence based medicine. There has been a meta analysis of all the newer antidepressants in the Archives of Internal Medicine. (Medline link). This is important as (a) the older antidepressants are hardly ever used because they have lots of side effects (or lots more, because the newer ones have side effects and (b) some of the newer antidepressants are now available from generic manufacturers.

I’m going to show you two forest plots. For there to be a signficant difference… the confidence intervals must not touch the vertical line.

SNRI vs SSRI, Arch Int Med 2011;155:772-85

Serotonin and Noradrenaline uptake inhibitors versus Serotonon only uptake inhibitors.

As you can see, there are virtually no differences between the serotonin and noradrenaline (neurotransmitter) medications and the serotonin alone group.

This is the SSRI (Serotonin alone) group.

SSRI vs SSRI Arch Int Medci 2011;155:772-85

Serotonin only antidepressants against each other.

Now, most trials are sponsored by drug companies. About 20 years ago, the companies under powered each trial (which showed the medications were equivalent , which allowed registration). Nowdays, they need to show superiority… which requires more participants, and more expense.

But the data here is similar to clinical lore. Most modern antidepressants work well enough. You select them by looking at side effects, tolerability, and what did not work last time.

Oh, it must be Wednesday.

Posted on by

Which means work.

At present I’m trying to work out how to do a meta analysis of the correlations or odds ratios in a nested confirmatory factor analysis. Probably not needed, as there are descriptive differences and I simply cannot compare apples with gorillas. Problem is there are two papers with almost identical results. But not quite.

WIll need to look at more than OVID for this — there were only 17 hits with the first search so I’ll change words and look again (I was hitting multiple databases at the same time though, so only five papers that looked useful). May not be much else

In the meantime, CRAN let me find this:

Authors: Ken Kelley
(2445)Confidence Intervals for Standardized Effect Sizes: Theory, Application, and Implementation
Reference: Vol. 20, Issue 8, Feb 2007
Submitted 2006-10-01, Accepted 2007-07-30
Type: Article
Abstract:

The behavioral, educational, and social sciences are undergoing a paradigmatic shift in methodology, from disciplines that focus on the dichotomous outcome of null hypothesis significance tests to disciplines that report and interpret effect sizes and their corresponding confidence intervals. Due to the arbitrariness of many measurement instruments used in the behavioral, educational, and social sciences, some of the most widely reported effect sizes are standardized. Although forming confidence intervals for standardized effect sizes can be very beneficial, such confidence interval procedures are generally difficult to implement because they depend on noncentral t, F, and x2 distributions. At present, no main-stream statistical package provides exact confidence intervals for standardized effects without the use of specialized programming scripts. Methods for the Behavioral, Educational, and Social Sciences (MBESS) is an R package that has routines for calculating confidence intervals for noncentral t, F, and x2 distributions, which are then used in the calculation of exact confidence intervals for standardized effect sizes by using the confidence interval transformation and inversion principles. The present article discusses the way in which confidence intervals are formed for standardized effect sizes and illustrates how such confidence intervals can be easily formed using MBESS in R.

I am such a stats geek that this is my bedtime reading. Pathetic.