Specificity profile of paroxetine in major depressive disorder: meta-regression
of double-blind, randomized clinical trials.
Author(s): Serretti A, Gibiino S, Drago A.
Affiliation(s): Institute of Psychiatry, University of Bologna, Italy.
alessandro.serretti@unibo.it
Publication date & source: 2011, J Affect Disord. , 132(1-2):14-25
BACKGROUND: Meta-analyses and reviews are powerful tools to inform clinicians on
overall effects of their therapeutic choices, but do not provide practically
useful clinical profiles for each drug optimal efficacy. Therefore clinicians in
everyday practice have to rely mainly on personal or anecdotic experience. The
aim of present study was to define the most suitable sociodemographic and
disease-related profile for the use of paroxetine as an antidepressant treatment.
METHODS: MEDLINE, EMBASE, CINAHL, and PsycINFO were searched for randomized
controlled trials (RCT) in English focused on "paroxetine" and "depressive
disorder" or "major depression". We also considered reviews and meta-analyses
focusing on paroxetine. Fifty-five total unique RCTs were included and
sociodemographic and clinical data as moderator of efficacy measures
(standardized mean difference based on Hamilton Depressive Rating Scale and
Montgomery-Asberg Depression Rating Scale) were investigated via meta-regression
analysis.
RESULTS: Paroxetine was significantly characterized by better response in females
and in Caucasians, whilst for patients who have been ill for a longer time before
treatment, the smaller was the antidepressant effect. Other disease-related
variables were not found to be significant moderators in clinical outcome.
LIMITATIONS: Meta-regression may lack sufficient sensibility to detect specific
subtle features, so a failure to find significant effect is not definitive
evidence of a lack of effect.
CONCLUSIONS: Our results, though significant, were generally observed with small
estimate values, their clinical relevance is subtle since each feature is
expected to influence marginally the whole outcome, and probably a more
pronounced effect could result only by analyzing very large samples.
|