Merck, NEJM In Statistical Flap
The Wall Street Journal (WSJ) reports this morning (subscription required) that the New England Journal of Medicine has corrected a study Merck published in the journal on Vioxx titled Approve. In Merck’s original interpretation of the study, patients taking Vioxx only developed cardiovascular complications after taking the drug for 18 months. Now a biostatistician hired by NEJM has concluded that patients’ cardiovascular risk could have increased before the 18-month period was up. (Click here to view the correction.)
According to the WSJ:
“[K]ey results in the New England Journal of Medicine’s original publication of the study . . . were reached by a different statistical method than was described in the article, and that when the data were subjected to the stated method, the contention that risk increased only after 18 months didn’t hold up.â€
Earlier this year, Merck admitted to the Food and Drug Administration (FDA) that it incorrectly described the statistical method it used to analyze Approve.
The Implications
Clearly, this correction will be a boon to plaintiffs seeking damages from cardiovascular events potentially caused by Vioxx. What’s worse is that Merck will likely be accused of “gaming†the data to ensure a more favorable outcome for its embattled drug. This may lead to calls for pharmaceutical companies to analyze and publish data using a variety of statistical methods to ensure that the results are consistent.
COX-2s Vs. High-Dose Advil
In separate news, the British Medical Journal published a meta-analysis of clinical trials indicating that high-dose Advil increases cardiovascular risk almost as much as COX-2 inhibitors like Vioxx. Based on the results of this trial, the WSJ’s editorial board has asked the FDA to return Vioxx to the market.
