Drug Companies Flagged for Biased Reporting

The studies examined how well gabapentin—originally approved to help control seizures and sold as Neurontin—treats other neurological ailments such as migraines and bipolar depression. Internal documents from Pfizer and Parke-Davis indicated that the companies changed the yardsticks they used to measure the drug's effectiveness during the course of the studies, an ethical no-no. In the studies’ planning phases, the studies listed the main criteria they would use to judge the effectiveness of gabapentin, such as relieving certain symptoms. These criteria are known as primary outcomes. Many studies also list secondary outcomes, less important criteria to judge success.

But according to the NEJM article, Pfizer and Parkes-Davis later swapped some primary and secondary outcomes, dropped some criteria altogether when the drug did not treat a symptom, and introduced new outcomes halfway through when the drug treated a symptom unexpectedly. None of these changes was mentioned in final, published scientific papers. "All the changes that took place ... led to a more favorable presentation in the medical literature of gabapentin’s efficacy,” the NEJM report says.

Drug companies are hardly alone in rigging their research results in this way. A report earlier this week in the Annals of Family Medicine found similar chicanery with many clinical trials: In 110 studies in top medical journals, primary outcomes had changed 30% of the time, secondary outcomes 70% of the time.

Even critics of this practice are not immune, it seems. MedPage Today included this biting aside in its story about the Annals study: “The researchers [who conducted the Annals study] noted—whether with intentional irony is unclear—that the relationship between studies’ funding status and alterations in outcome measures was not one of their original research questions, but they provided some analysis of it anyway.”