(By Joseph W. Koletar) The Economist Magazine of March 26, 2016 carried the following article on p. 85: “Clinical Trials: For my next trick…”
It recounts the experience of drug company GlaxoSmithKline (GSK) with a drug called Paxil. Coming to market in 1992, it was presented as a treatment for depression. Some years later it was earning $2B a year, but when British authorities took another look at its effectiveness in 2003 they found Paxil was not only marginal in treating depression, it actually had negative side-effects, to include suicidal ideation. In 2012 American authorities imposed the largest fine ever on a pharmaceutical firm – $3B.
The issue is, in academic terms, called “outcome switching.” In the case of GSK, the article reports the initial trial of the drug was meant to measure eight indications of effectiveness in treating depression. None, per the article, showed it was any better than a placebo (“sugar pill.”) As a result, the researchers came up with nineteen new measures of effectiveness, most of which indicated no significant benefit. Four however did, and these were presented to the medical community and public as if they had been the original focus of the effectiveness study. Thus, did authorities come to levy an unprecedented financial penalty on GSK.
Such issues hardly disappeared with the GKS case. The article notes many promising discoveries in various areas of medicine suffer from what is called “replication crisis.” The concept is simple: If I know how you did what you claim you have done, I should be able to replicate the results. If I cannot, something is amiss.
The article notes the Centre for Evidence-Based Medicine at Oxford University is in the process of doing this, at least for some medical “breakthroughs” with a program named COMPARE. A later study by others indicated that 31% of clinical trials did not adhere to the measurements they initially promised to utilize. Yet another study of 137 medical trials found 18% altered their criteria of success mid-way during the testing period, and 64% had altered their evaluations to emphasize secondary, less impressive, indications of success.
COMPARE, for example, examined 67 clinical trials, which had been reported in well-respected professional journals. Of these, nine were found to be excellent, in that they either achieved their initial goals, or explained why they had altered their initial measurement process. The other 58 seemed to move the goalposts a bit during the measurement process. The studies had 300 outcomes that should have been reported, but were not. At the same time 357 other outcomes were added, although they had not been specified at the beginning of the respective studies.
Such matters were brought to the attention of the journals in question that published the original test results. Seven published COMPARE’s comments; 16 responded to the effect that COMPARE’s observations were trivial or the result of inadvertent error. The other 44 seem to be in a state of suspended animation, even though all the journals involved had previously adopted guidelines to be attentive to “outcome switching.”
For mere mortals, are we left with a situation wherein the hunter can proclaim?
“I missed the deer, but wounded the tree!”
Again, we return to issues of human motivation. These are smart, apparently well-intentioned people. Is it pride of ownership? A sense that great effort deserves at least some recognition/success? Or, is it something else?
In our book, The A.B.C.’s of Behavioral Forensics, we explore the Apple-Bushel-Crop dimensions of organizational misconduct. Is it one person, a group of people, or the entire organization? Is it the CEO forcing subordinates to do improper things? Is it a section of the organization that flies by its own rules? Or has the entire organization gone off course? Is it the familiar defensive refrain of “industry standards” or “accepted/common practice?”
It was once common practice to collude with competitors to fix prices, but that was outlawed; or was it? News reports indicate U.S. federal authorities have recently begun an investigation of airlines that seem to have a pattern of having the same prices to fly from point ”A” to point “B.” In an industry that is supposed to be competitive, is this not odd?
In our book we use “C” to refer to an organization, but is it possible it can also refer to an entire industry segment?
In matters both medical and, increasingly, corporate and financial how are we to know what is authentic? Or, have we slipped into the increasingly-common world of “virtual reality?”
Time, our health, and our finances will tell.
© Joseph W. Koletar, 2016
BEHAVIORAL FORENSICS GROUP
The Behavioral Forensics Group is a team of professionals with vast experience in detecting fraud, understanding why it occurs, and in recommending steps to mitigate fraud incidence within the corporate workplace, particularly within higher-level (and therefore more costly to the enterprise) executives. The fields of investigation, organizational psychiatry, accounting and behavioral forensics, and law enforcement are represented within the Behavioral Forensics Group. Acting in synergy to help organizations prevent, find, and/or reduce fraud, BFG is a premier, pioneering practice in this field.
We are blogging at: http://www.bringingfreudtofraud.com