According to the Washington Post:
There is almost no scientific evidence to back up the U.S. intelligence community's use of controversial interrogation techniques in the fight against terrorism, and experts believe some painful and coercive approaches could hinder the ability to get good information, according to a new report from an intelligence advisory group. [emphasis added]
This isn't exactly news, of course, but this study (pdf), sponsored by the Defense Intelligence Agency and the Pentagon's own Counterintelligence Field Activity, and published by the National Defense Intelligence College, makes the point clear.
In the report:
... experts find that popular culture and ad hoc experimentation have fueled the use of aggressive and sometimes physical interrogation techniques to get those captured on the battlefields to talk, even if there is no evidence to support the tactics' effectiveness. [emphasis added]
And the result is as predictable as it is horrible. According to Robert A. Fein, chairman of the study:
"This shortfall in advanced, research-based interrogation methods at a time of intense pressure from operational commanders to produce actionable intelligence from high-value targets may have contributed significantly to the unfortunate cases of abuse that have recently come to light."
So "extreme measures" are taken for uncertain results, in ignorance, and based on fiction. A sad example of life -- and death -- imitating art.