According to a recently conducted reanalysis, the 2001 study that paved the way for prescribing Paxil to hundreds of thousands of adolescents had serious flaws in it.

The research team conducting this reanalysis used data obtained from 77,000 pages of a document which was previous unavailable. The in-depth reanalysis of the data allowed the researchers to conclude that paroxetine, a drug marketed under the brand name of Paxil by pharmaceutical firm GlaxoSmithKline, is not more effective than placebos and is significantly more harmful than what the previous study indicated.

The previous study, which was referred to as “Study 329” started to face similar criticisms within a year of getting published. However, the reappraisal published in The BMJ, a widely read medical journal, is probably the most thorough one yet.

The two consecutive years 2003 and 2004 saw the Food & Drug Administration issue warnings regarding a rise in the rate of suicidal thinking among some adolescents and children who consumed drugs belonging to an antidepressant class called selective serotonin reuptake inhibitors; two drugs of this class were Zoloft and Paxil.

Paxil, however, can still be prescribed to teens, but, in 2004, the FDA issued an order asking manufacturers to add an additional “black box” warning on Paxil’s labels informing users about increased suicidal thinking in some young people using the medication.

Then, in 2007, the agency ordered an expansion of the warning. This time it asked the manufacturer to include individuals under 25. Five years after that i.e. in 2012, the manufacturing company of Paxil GlaxoSmithKline pleaded guilty. The company had to pay a fine of $3 billion for unlawfully marketing some drugs including Paxil.

The research team that published the new study in The BMJ this Wednesday was led by Professor Jon Jureidini of the University of Adelaide, Australia. The researchers under Prof. Jureidini analyzed the raw data and discovered that Study 329 was filled with transcription errors. Also, it also contained a range of other problems which violated GlaxoSmithKline’s own policies for statistical analyses,

The researchers said that they have failed to determine the reason behind a number of mistakes detected by them. However, they admitted that the mistakes were exceedingly egregious, which made them believe that they cannot be unintentional.