Your Handy Neurontin Study Odds Calculator!

(MoneyWatch)  Pfizer (PFE) frequently changed the stated primary endpoints of its published Neurontin studies in order to cherry-pick favorable results, according to a study by Kay Dickersin of Johns Hopkins University in Baltimore, to be published in the New England Journal of Medicine.

Dickersin's study is complicated. So complicated that Pfizer believes Dickersin has misunderstood what she read:

"We believe the review suffers from significant bias, insufficient data, poor methodology, and cannot pass the threshold of credible scientific research," Pfizer said in a statement.

Here's a handy primer for calculating the odds of a published Pfizer Neurontin study actually being accurate, according to Dickersin.

• In 12 published trials, eight showed the definition of the primary study goal differed between the original and published documents, according to Bloomberg. Odds of a Pfizer Neurontin trial accurately stating its primary goal: 33 percent.

• For 21 primary study objectives: six weren't included in published reports and four were reported as secondary goals, according to Bloomberg. Odds of a specific primary goal staying the same from start to finish: 52 percent.

• [In] seven of the nine trials published as full-length research articles ... the outcome differed between the published account and the original documents. Odds of a trial's reported end outcome differing from start outcome: 78 percent.

• Researchers reviewed internal documents from Pfizer and its Parke-Davis unit on 20 gabapentin trials, which became public as a result of litigation against the firms. Eight of the studies were never published, according to MedPage Today. Odds of a P-D/Pfizer Neurontin study never being published: 40 percent.

• In the four trials for which there was full agreement, three reported a statistically significant finding in gabapentin's favor for the primary outcomes. Odds of an unmanipulated Neurontin study being positive: 75 percent.

• In the eight trials with disagreement, five reported favorable results for the new primary outcomes. Odds of a manipulated Neurontin study being positive: 63 percent.

Image by Flickr user stoneflower, CC.