Duncan & Miller Strike Out...Again!

If you have been following these baloney watch pages, you know that our last effort to write a letter to the editor in response to the never-ending supply of fresh off the presses baloney, met with a rebuff from the APA Monitor ( click here.)

Undaunted, we gave another shot at the Psychotherapy Networker after we read two pieces, about very important topics, that significantly misinformed readers. Our letter addressed Jay Lebow’s article called “A Look at the Evidence,” which should have been titled: A Look at the Evidence from the Perspective of Model Maniacs Who Ignore Forty Years of Contradictory Findings.” Unfortunately, our response to this empirically bankrupt pining over model and technique never saw the light of day—it was cut from our letter. The part of our letter that did make it to print addressed Garry Cooper’s research column which quoted a drug company spokesperson (or researcher depending on your point of view) pooh-poohing the risk of Prozac for adolescents while trumpeting its effectiveness. Cooper’s response to our letter did not address our points at all (a sure sign of the veracity of our comments) but rather discussed issues we did not address and would tend to agree—yes, it is not a perfect world, and yes, client preferences are key to the choice of any treatment strategy. In a “perfect world,” statements by drug company stooges would not be accepted without scrutiny/fact checking, or at least disclosure so that readers might know to look at the research themselves (see Sparks, Duncan, & Miller, 2006.)

Here is the letter as we submitted it:

Lebow’s “Look at the Evidence” promotes gross misunderstandings about the research supporting specific approaches for specific “disorders,” perpetuating the field’s love affair with model/technique. What is not mentioned is that model/technique accounts for the smallest portion of the variance of outcome (meta-analysis suggests as low as 1%). Consider the NIMH study of depression (TDCRP) which reported that only 2% of the variance was attributable to model/technique, while 20% of noted changes were accounted for by the alliance. What is also left out in the accolades to specific models is that NONE of the mentioned approaches have ever proven themselves better than any other systematically applied therapy. They have only demonstrated efficacy over NO treatment or treatment as usual (TAU)—which is nothing to write home about. Any study that investigates the pet approach of the researcher and picks the therapists, trains them, and provides special supervision (usually with reduced caseloads) will outperform TAU. The exorbitant cost of implementing these models could be better spent on reducing caseloads, and supporting therapists with supervision, feedback about outcome, and training in the models of their choice.

Cooper’s enjoyable column also misinforms. Prozac is proclaimed to be effective, but what is not said is that even after weeding out placebo responders, Prozac has not been shown to be better than placebo on any primary or client-rated measure in any study. The risk of suicide, no matter how low, combined with such underwhelming “effectiveness” makes Prozac a bad first choice for troubled adolescents—which is the conclusion of the recent APA working group on kids and psychotropic drugs. And also what is not said is that the psychiatrist extolling the virtues of Prozac has received recent funding from guess who—Eli Lilly who happen to make Prozac.