Posted 10 September 2013 - 12:48 PM
Unequivocally better than placebo. Somewhere between 20% & 40% percent more effective. In the case of less severe/chronic depressive episodes they are unlikely to be very effective. In fact NICE guidelines clearly indicate first line for presentation of mild depression be anything but pharmacological - something I emphatically agree with - except in the case of Chronic disorders or people with a history of more severe disorder.
STAR*D is the study to extrapolate from here. It clearly showed effect and didn't limit intake to the study based on only Depressive Episode as a diagnosis. A massive proportion of the depressed population have Psychiatric co-morbidity (ignoring non psychiatric co-morbidity) and pharmacological management of the depressive episode may well also be having an impact on other psychopathology. 1/10 inpatient admissions at least have a physical co-morbidity that may be effecting mental status. If these facts were controlled for, which they rarely are satisfactorily in studies an improvement in effect is likely.
The older studies show a much more significant effect of ADs (in general i.e. Mirtazapine, Meclobremide and the like may be exempt) and some have interpreted this as shoddy research. The reality is that studies from 30 years ago were using TCA's, MAOI's and the like. SSRI's are clearly less effective (and the most commonly studied now) and there use should be limited to first line agents (in less severe or non-sub-typed cases), for polypharmacy and when tolerance is a big issue. ADs are commonly prescribed not on the basis of symptoms but on general efficacy which is also a mistake.
I remain quite convinced that if rational pharmacy was studied in a large trial (atypical depression treated with Meclobremide, Melancholic with venlafaxine or Mirtazapine, 'typical' with SNRIs/Burpropion/Mirtazapine and then a rational evidence based progression including adjuncts, combinations etc) a much larger effect would be seen.
If you read a lot of the non-published studies they tend to have major methodological flaws (not always by any means).
Its also worth mentioning that the most effective and possibly worst tolerated of the treatment for depressive disorder remains non-selective irreversible Monoamine oxidase inhibitors. Adjuncts can be used to improve efficacy further and manage side effects when used safely and effectively. If a G.P can manage use of high potency opioids then there is no reason they can't follow a well set out algorithm for MDD including the use of MAOI's, which are seriously under used. Only when there is a significant risk of suicide, severe self-harm, complex co-morbidity, level four treatment resistance, psychosis or need for specialist psychotherapy and case management does a referral to Community mental health teams and Psychiatric input really need to be made.
I may have gone slightly off topic. In short my views are:
With the use of ADs as they currently are they are still more effective than Placebo and do reach significance.
A massive improvement in response is likely if a more rational algorithm is used, based on long standing evidence both controlled and from case studies.