Antidepressant Drug Effects and Depression Severity
I'm reading this paper but i don't quite understand it.
A good bit of the method is excluding studies with a placebo washout
period. This seems pointless for clinical practice since placebo
responders can't be detected. The fact is most GPs and psychiatrists
don't have the time to do a proper interview anyway. Diagnosis in
clinical practice is based on subjective judgements and the science
often gets left behind.
There's an interesting note about how many people who self-report
depression fall below the diagnostic criteria for the use of
antidepressants.
"In fact, a recent survey of depressed, treatment-seeking outpatients
found that 71% of the 503 patients assessed had HDRS scores less than 22."
Zimmerman M, Posternak MA, Chelminski I. Symptom severity and exclusion
from antidepressant efficacy trials. J Clin Psychopharmacol.
2002;22(6):610–614.
(22 and above is very severe depression but antidepressants can work for
people who have scores of 19+ which is the threshold for severe depression.)
This isn't what I'm interested in though. It's the value of the
so-called mega-analysis. A single study presents the results as averages
and stuff like that. A meta-analysis takes lots of similar studies and
averages them. It's sort of getting the percentage of men in London by
averaging (in a sophisticated way but it's essentially a mean average)
the percentages from each borough. A mega-analysis works by taking the
raw data from each borough to make the average rather than taking the
averages from each borough to get the average for the whole of London.
Here's what the authors of the paper say,
"Unlike the data analyzed by Kirsch et al and Khan et al, which
contained information only at the level of treatment group and thus
could support only standard meta-analytic procedures, the databases from
the 6 studies included in the present investigation provided data for a
patient-level meta-analysis, also known as a mega-analysis. This
approach is more appropriate and more powerful than a standard
meta-analysis when original data are available and a fine-grained
multivariate analysis is desired."
And they reference
Steinberg KK, Smith SJ, Stroup DF, et al. Comparison of effect estimates
from a meta-analysis of summary data from published studies and from a
meta-analysis using individual patient data for ovarian cancer studies.
Am J Epidemiol. 1997;145(10):917–925
http://jama.ama-assn.org/cgi/ijlink?linkType=ABST&journalCode=amjepid&resid=145/10/917
The paper won't load on this shitty computer with a shitty operating
system which fucks up even in Safe Mode. All I can think is whether
patient-level meta-analysis without a funnel plot s a valid research
method because publication bias would still affect "mega-analysis" but
what seems to be even stranger (in meta-analysis terms) is the variety
of the studies and treatments included. 3 SSRI trials and 3 tricylic
trials. 1 study included which was on minor depression. Authors were
contated to ensure a washout period hadn't been used? What the fuck?
At this point I'd consider this study a total waste of time but I still
don't know why they bothered doing this nor what the advantage of a
mega-analysis is over a meta-analysis, especially given the failures in
the fundamental design of this statistical review of trials.
Either I'm stupid or this study is stupid. I'll go with the former. I
dont get this study, why it was done and why I've bothered to keep
reading. Can't be arsed to read any more but i will. I can understand
how patient-level analysis into effects and stuff to provide
individualised, works-firt-time treatments might be better if all the
factors of the biopsychosocial model could be recorded but the method
and science here....well it baffles me.
I'm surprised this paper could pass peer review. It stinks of what's
typical in psychiatric research. Nothing useful has been determined and
there's no attempt to find useful information relevant to clinical
practice. With so many people presenting with self-reported symptoms of
depression (of any order of magnitude) and so little a practising
physicians can do but trial and error...fuck....frankly patients could
do a better job than any of this evidence-based bollocks.
No comments:
Post a Comment